Transport Planning and Modeling Introduction

CONTENTS
1. INTRODUCTION ....................................................................................................6
1.1 Transport planning and modeling ....................................................................6
1.1.1 Overview ..................................................................................................6
1.1.2 Models and their Role ..............................................................................7
1.2 Characteristics of Transportation Problems.....................................................8
1.2.1 Characteristics of Transportation Demand ...............................................8
1.2.2 Characteristics of Transportation Supply ..................................................8
1.2.3 Equilibration of Supply and Demand ........................................................9
1.3 Modeling and Decision Making .....................................................................10
1.3.1 Decision making Styles ...........................................................................10
1.3.2 Choosing Modeling Approaches .............................................................11
1.4 Issue in Transportation Modeling ..................................................................12
1.4.1 General Modeling Issues ........................................................................12
1.4.2 Aggregate and Disaggregate Modeling ...................................................13
1.4.3 Cross Section and Time Series ................................................................14
1.4.4 Reveal and Stated Preferences ...............................................................14
1.5 The structure of the Classic Transportation Model ........................................14
1.5.1 The Explicit Demand Models ..................................................................14
1.5.2 The Implicit Demand Models..................................................................15
1.6 The Urban Transportation Planning Process..................................................17
1.6.1 Formulation of goals and objectives .......................................................18
1.6.2 Inventory................................................................................................18
1.6.3 Analytical methods.................................................................................19
1.6.4 Forecasting.............................................................................................19
1.6.5 Formulating of alternative plans.............................................................19
1.6.6 Evaluation ..............................................................................................19
1.6.7 Implementation .....................................................................................19
1.6.8 Information needed ...............................................................................20
1.6.9 Study area ..............................................................................................21
Transportation Systems Planning (ECIV 540) | INTRODUCTION 1

1.6.10 Zoning .................................................................................................21
1.6.11 Network ..............................................................................................23
2. DATA AND SPACE ...............................................................................................24
2.1 Basic Sampling Theory...................................................................................24
2.1.1 Statistical Considerations .......................................................................24
2.1.2 Conceptualization of the Sampling Problem ...........................................26
2.1.3 Practical consideration in sampling ........................................................28
2.2 Error in Modeling and Forecasting ................................................................29
2.2.1 Different Types of Error ..........................................................................29
2.3 Basic Data Collection Methods ......................................................................31
2.3.1 Practical Considerations .........................................................................31
2.3.2 Type of Survey........................................................................................32
2.4 Stated Preference Surveys.............................................................................33
2.4.1 Introduction ...........................................................................................33
2.4.2 Contingent Valuation (CV) ......................................................................33
2.4.3 Conjoint Analysis (CA) ............................................................................34
2.4.4 State Choice Methods ............................................................................34
3. TRIP GENERATION ..............................................................................................36
3.1 Types of trips.................................................................................................37
3.1.1 By trip purpose .......................................................................................37
3.1.2 By time of day ........................................................................................38
3.1.3 By person type .......................................................................................38
3.2 Trip generation by expansion factors ............................................................38
3.3 Cross-Classification Models ...........................................................................40
3.4 The FHWA-Simplified Trip-Production Procedure ..........................................42
3.5 Regression Analysis .......................................................................................44
3.5.1 Forms of Regression ...............................................................................44
3.5.2 Correlation .............................................................................................45
3.5.3 Hypothesis tests for a .............................................................................46
4. TRIP DISTRIBUTION ............................................................................................50
4.1 Definitions and notations ..............................................................................50
4.2 Growth factor methods .................................................................................51

4.2.1 Uniform growth factor ...........................................................................51
4.2.2 Singly constrained growth factor ............................................................52
4.2.3 Doubly constrained growth factor ..........................................................53
4.2.4 Average factor........................................................................................54
4.2.5 Detroit....................................................................................................54
4.2.6 Advantages and limitations of growth factor model ...............................55
4.3 The Fratar Method ........................................................................................55
4.4 The Gravity Model .........................................................................................58
4.4.1 Gravity Model by Drew's Technique .......................................................61
5. MODAL SPLIT......................................................................................................65
5.1 Overview .......................................................................................................65
5.1.1 Factors Affecting Modal Split..................................................................65
5.1.2 Methods.................................................................................................66
5.2 Types of modal split models ..........................................................................66
5.2.1 Trip-end modal split models ...................................................................66
5.2.2 Trip-interchange modal split models ......................................................66
5.2.3 Aggregate and disaggregate models .......................................................66
5.3 Diversion Curve .............................................................................................67
5.4 The Stratified Diversion-Curve Model............................................................67
5.5 Probabilistic Models.... Logit Model ..............................................................74
5.5.1 Utility and Disutility Functions ................................................................74
5.5.2 Binary logit model ..................................................................................74
5.5.3 Multinomial Logit Model ........................................................................76
5.6 Behavioral Model ..........................................................................................77
6. TRAFFIC ASSIGNMENT ........................................................................................80
6.1 Overview .......................................................................................................80
6.2 Link cost function ..........................................................................................80
6.3 All-or-nothing traffic assignment ...................................................................81
6.4 Diversion .......................................................................................................83
6.5 User equilibrium assignment (UE) .................................................................85
6.6 System Optimum Assignment (SO) ................................................................88
6.7 Incremental assignment ................................................................................89

6.8 Stochastic user equilibrium assignment ........................................................89
6.9 Dynamic Assignment .....................................................................................89
6.10 Capacity-Restraint Techniques ......................................................................90
6.11 Summary .......................................................................................................91
7. URBAN PUBLIC TRANSPORTATION .....................................................................92
7.1 Transit planning.............................................................................................92
7.2 Definitions of transit service ..........................................................................92
7.2.1 General function definitions ...................................................................93
7.2.2 Technological definitions ........................................................................93
7.2.3 Service definitions ..................................................................................96
7.3 Bus Transit Systems .......................................................................................97
7.3.1 Priority treatment of buses ....................................................................98
7.4 Para Transit ...................................................................................................99
7.5 Light Rail Transit Systems ............................................................................ 100
7.6 Rapid Rail Transit Systems ...........................................................................101
7.7 Automated Guideway Transit (AGT) ............................................................ 102
7.8 Transit Costs ................................................................................................ 103
7.9 Cost Allocation Model ................................................................................. 105
7.10 Bus Service Cost Models.............................................................................. 105
7.11 Costing Process ........................................................................................... 107
7.12 Transit System Design and Operation .......................................................... 112
7.13 Rail Capacity ................................................................................................ 115
7.14 Operational Design ...................................................................................... 115
7.14.1 Rail Operation Design........................................................................115
7.14.2 Bus Operation Design ........................................................................119
7.15 Capacity of Urban Transit ............................................................................ 121
7.16 Route Schedules .......................................................................................... 122
7.16.1 Headways ......................................................................................... 123
7.16.2 Constructing Timetables ...................................................................125
7.16.3 The Scheduling Algorithm .................................................................128
8. DATA COLLECTION ........................................................................................... 134
8.1 Overview .....................................................................................................134

8.2 Survey design .............................................................................................. 134
8.3 Information needed .................................................................................... 134
8.4 Selecting Your Sample ................................................................................. 135
8.5 Survey Methods .......................................................................................... 136
8.5.1 Personal Interviews .............................................................................. 136
8.5.2 Telephone Surveys ............................................................................... 136
8.5.3 Mail Surveys ......................................................................................... 137
8.5.4 Computer Direct Interviews .................................................................138
8.5.5 Email Surveys ....................................................................................... 138
8.5.6 Internet/Intranet (Web Page) Surveys .................................................. 139
8.5.7 Scanning Questionnaires ......................................................................140
8.5.8 Summary of Survey Methods ............................................................... 141
8.6 Household data ........................................................................................... 141
8.6.1 Questionnaire design ...........................................................................141
8.6.2 Survey administration ..........................................................................142
8.7 Data preparation ......................................................................................... 142
8.7.1 Data correction .................................................................................... 143
8.7.2 Sample expansion ................................................................................ 143
8.7.3 Validation of results ............................................................................. 143
8.8 Other surveys .............................................................................................. 144
8.8.1 O-D survey............................................................................................ 144
8.8.2 Road side interviews ............................................................................ 144
8.8.3 Cordon and screen-line survey ............................................................. 144
8.9 Summary .....................................................................................................144

1. INTRODUCTION
1.1 Transport planning and modeling
1.1.1 Overview
Many treatises and learned essays have been written about the subject of planning
and the role of the professional planner in various societal functions.
One hears of urban, economic, financial, corporate, industrial, water resource,
environmental, and many other kinds of planning.
In the field of transportation, professional designations such as highway planner,
airport planner, and urban transportation planner, are common.
It is thus clear that planning is considered to be an important function in modern
society and that, whatever this function is, it has a specific focus-that is, it is
concentrated in particular areas, subjects, or systems.
Planning may be defined as the activity or process that examines the potential of
future actions to guide a situation or a system toward a desired direction, for
example, toward the attainment of positive goals, the avoidance of problems, or
both.
Being the conceptual, the process that precedes a decision to act in a certain way,
planning is a fundamental characteristic of all human beings.
The most important aspect of planning is the fact that it is oriented toward the
future: A planning activity occurs during one time period but is concerned with
actions to be taken at various times in the future.
However, although planning may increase the likelihood that a recommended action
will in fact take place, it does not guarantee that the planned action will inevitably be
implemented exactly as conceived and on schedule.
It is often said that everything is related to everything else.
Therefore, any event or human action affects everything else, ultimately in ways that
are beyond the limits of human comprehension.
As a matter of practicality, planning is not a search for ultimate answers but only a
means to specific ends that is based on the proposition that better conditions would
result from premeditative as opposed to impulsive actions.
How much premeditation is necessary (i.e., how much planning is good planning) in a
particular situation is always an open question: Too little planning is almost like no
planning, and too much planning is self-defeating, as it leads to inaction.
By necessity, any particular planning effort has a limited scope and is oriented toward
bringing about specific desirable ends.

Since desirability cannot be divorced from the value system of human beings,
planning is necessarily directed toward the satisfaction of the goals and objectives of
particular groups of people.
However, within its social context, planning cannot afford to ignore the reactions of
other groups; it must, in fact, anticipate these responses as well.
In addition, when the group on behalf of which planning is undertaken is
heterogeneous, the planning effort must deal with the presence of internal conflicts
relating to specific objectives and aspirations.
This is especially critical when the government participates in or regulates the
planning effort.
The fundamental purpose of transportation is to provide efficient access to various
activities that satisfy human needs.
Therefore, the general goal of transportation planning is to accommodate this need
for mobility.
Within specific contexts, however, whose mobility, for what purpose, by what means,
at what cost and to whom, and who should do the planning and how are questions
that are not amenable to easy-answers.
Contemporary responses to these questions are largely rooted in history and have
been influenced by a confluence of many factors, including technological innovations,
private interests, and governmental policies.
Transportation planning is the process of making decisions related to the future of
the transportation system.
It focuses on issues such as the future demand for transportation; interaction among
different transportation systems and facilities; the relationships among land use,
economic activity, and transportation; alternative ways of operating transportation
systems; the social, economic, and environmental impacts of proposed
transportation systems; and the financial and institutional arrangements needed to
implement transportation proposals.
1.1.2 Models and their Role
A model can be defined as a simplified representation of a part of the real world - a
system of interest - which concentrates on certain elements considered important for
its analysis from a particular point of view.
Incorporate both physical and abstract models.
In this course, we are concerned mainly with an important class of abstract models -
mathematical models.
A model is only realistic from a particular perspective or point of view.

1.2 Characteristics of Transportation Problems
1.2.1 Characteristics of Transportation Demand

The demand for transportation is derived, it is not an end in itself. People travel in
order to satisfy a need (work, leisure, health) undertaking an activity at particular
locations.
The demand for transportation services is highly qualitative and differentiated. There
is a whole range of specific demands for transportation which are differentiated by
time of day, day of week, journey purpose, type of cargo, importance of speed and
frequency.
Transportation demand takes over space. It is the distribution of activities over space
which makes for transportation demand. The most common approach to treat space
is to divide study areas into zones and to code them, together with transportation
networks.
Transportation demand has very strong dynamic elements. The time-variable
character of transportation system could be coped well with the average demand of
travel in an area but it breaks down during peak periods.
1.2.2 Characteristics of Transportation Supply

The first distinctive characteristic of transportation supply is that it is a service and
not a good. A transportation service must be consumed when and where it is
produced, otherwise its benefit is lost.
Transportation supply requires a number of fixed assets, the infrastructure, and a
number of mobile units, the vehicles. It is the combination of these, together with a
set of rules for their operation that makes possible the movement of people and
goods.
Infrastructure and vehicles are not own nor operated by the same group or company.
Transportation infrastructure is "lumpy". One cannot provide half a runway or one-
third of a railway station. There may be scope for providing a gradual build-up of
infrastructure to match growing demand. In this way, the provision of infrastructure
can be adjusted to demand and avoid unnecessary early investment in expensive
facilities.
Investments in transportation infrastructure are not only lumpy but also take long
time to be carried out. There are usually large projects.
Transportation investment has an important political role.
The separation of providers of infrastructure and suppliers of services introduces
economic complexities. The use of taxes on vehicles and fuels is only a rough
approximation to charging for provision of infrastructure.

One of the most important features of transportation supply is congestion. Note that
the contribution an additional car makes to the delay of all users is greater at high
flows than at low flow levels.
Figure 1-1 Congestion and its external effects
1.2.3 Equilibration of Supply and Demand

Consider a set of volumes on a network, V, a corresponding set of speed, S, operating
capacity, Q, under a management system M:
S = f(Q, V, M)
The capacity Q would depend on the management system M and on the levels of
investment, I, over the years:
Q = f(I, M)
The level of demand, D, is dependent on the level of service provided by the
transportation system and also on the allocation of activities A over space:
D = f(S, A)
Combine S and D for a fixed activity system one would find the set of equilibrium
points between supply and demand for transportation. However, activity system
itself would probably change s levels of service change over time and space.
Therefore, one would have two different sets of equilibrium points: short-term and
long-term ones.
A typical example is the car and public transportation vicious circle in the Figure
below:
Figure 1-2 Car and public transportation vicious circle
Figure 1-3 Breaking the car and public transportation vicious circle
1.3 Modeling and Decision Making
1.3.1 Decision making Styles

1.3.1.1 Decisions Based on Master Plans
Decisions are based on interpretations of the master plan which provides the rules
governing contingencies, performance expectations and what can or cannot be done.
Master plans may be reasonable when the environment is stable.
1.3.1.2 Normative Decision Theory or Substantive Rationality

It is sometimes referred to as the "system approach".
The decision problem is seen as one of choosing options from a complete set of
alternatives and scenarios, with estimates on their probability of occurrence; the
utility of each alternative is quantified in term of benefits and costs and other criteria
like environmental protection, safety, etc.
Some problems of applying normative decision theory are:
Some elements of the objective function or constrains may be difficult to
quantify.
The accusation of insensitivity to the aspiration of the public;
Its high costs
The alienation of decision makers who may not understand nor accept the
analytical treatment of the problem.
1.3.1.3 Behavioral Decision Theory
Decision makers are not utility maximizes but simply satisfiers. The search for better
solution is often stopped once an acceptable one is found. This approach combines
searching, learning, and decision making.
1.3.1.4 Group Decision Making

Decision making becomes a learning process inside a group with decision authority
and a specific remit.
Individuals contribute their expertise and knowledge and the group tries to apply
these to the decision problem.
1.3.1.5 Adaptive Decision Making
It recognizes the interaction between pressure groups, none holding complete
decision-making power. Each group sees the problem in a different way and
therefore negotiation and compromise are required to reach a decision.
The approach is common in legislative decision making and diplomacy.
1.3.1.6 Mixed-mode Decision-Making Strategies
It combines many of above approaches into a flexible strategy. This is quite common
in transport studies.
1.3.2 Choosing Modeling Approaches

The decision -making context
Strategic issues or schemes, tactical schemes, or operational problems;
Help define requirements on the model to be used, variables to be included or
considered given or exogenous.
Accuracy required
The availability of suitable data
The state of the art in modeling
Behavioral richness;

Mathematical and computer tractability;
Available of good solution algorithms.
Resources availability for the study
Data processing requirements
Level of training and skills of the analysts
1.4 Issue in Transportation Modeling
1.4.1 General Modeling Issues

1.4.1.1 Model Specification
Model structure: Is it possible to replicate the system to be modeled with a simple
structure which assumes that all alternative are independent?
Functional Form: Is it possible to use linear forms or does the problem require
postulating more complex non-linear functions?
Variable specification: Which variables to use and how (which form) they should
enter a given model.
1.4.1.2 Model calibration, validation, and use

Calibrating a model requires choosing its parameter, assumed to have a non-null
value, in order to optimize one or more goodness-of-fit measures which are a
function of the observed data.
Estimation involves finding the values of the parameters which make the observed
data more likely under the model specification. In this case, one or more parameters
can be judged non-significant and left out of the model.
Because the large majority of transportation models have been built on the basis of
cross-sectional data, there has been a tendency to interpret model validation
exclusively in terms of the goodness-of-fit achieved between observed behavior and
based-year predictions. Validation requires comparing the model predictions with
information not used during the process of model estimation.
A model is normally used to test a range of alternative plans for a range of possible
assumptions about the future value of the other variables.

Figure 1-4 Modeling and sampling
1.4.1.3 Modeling, forecasting and judgment

Modeling focuses on building and applying appropriate tools that are sensitive to the
choices of interest and respond logically to changes in key policy instruments. The
successful modeler will provide useful and timely advice to the decision-making
process, even if the data and timescales are limited.
Forecasting is an attempt to envision and quantify future conditions. It normally
involves estimating future travel demand and the resulting multimodal flows and
costs over time.
Given nature of analytical models, interpretation of their output is essential.
Interpretation requires good judgment and this is only acquired with experience and
thorough understanding of the theories underpinning and their limitations.
1.4.2 Aggregate and Disaggregate Modeling
When the model at base aims at representing the behavior of more than one
individual, such as aggregate or first-generation models, a certain degree of
aggregation of the exogenous data is inevitable. When the model at base attempts to
represent the behavior of individuals, such as disaggregate or second-generation
models, it is conceivable that exogenous information can be obtained and used
separately for each traveler.
In most case, the forecasts and sensitivity estimates must be provided at the
aggregate level. They must represent the behavior of the entire population of
interest. The analyst using disaggregate models must find a sound method for
aggregating model results to provide these indicators.
1.4.3 Cross Section and Time Series
A fundamental assumption of the cross-sectional approach is that a measure of the
response to the incremental change may simply be found by computing the
derivatives of a demand function with respect to the policy variables in question.
It has two potentially serious drawbacks. A given cross-sectional data set may
correspond to a particular "history" of changes in the values of certain key variables
influencing choice. Second, data collect only one point in time will usually fail to
discriminate between alternative model formulations.
Longitudinal or time series data should be use to construct more dependable
forecasting models.
Longitudinal data can take the form of panels or more simply before-and-after
information.
1.4.4 Reveal and Stated Preferences
Reveal Preference data is based on information about observed choices and
decisions. Within this approach, project evaluation requires expressing policies in
terms of changes in attributes which "map onto" those considered to influence
current behavior. This has practical limitations basically associated with survey costs
and the difficulty of distinguishing the effects of attributes which are not easy to
observe. Another practical embarrassment has been "new option" problem, it is
required to forecast the likely usage of a facility not available at present.
Stated Preference techniques base demand estimates on an analysis of the response
to hypothetical choice: these can cover a wider range of attributes and conditions
than the real system.
1.5 The structure of the Classic Transportation Model
1.5.1 The Explicit Demand Models

A related travel-demand theory states that an individual makes travel choices
simultaneously rather than in a sequence of discrete steps and that a demand model
should be calibrated to reflect this behavior.
An often-cited example of simultaneous models is the Quandt and Baumol
formulation of intercity travel demand, takes the general form
CIJK a 4 H D
QIJK a0 ( PI )a1 ( PJ ) a 2 (CIJ * ) a3 ( ) ( H IJ * )a5 ( IJK ) a 6 ( IJK )a 7 (YIJ ) a8
CIJ * H IJ * DIJ *
where:
QIJK : travel flow between cities I and J via mode K
PJ, PJ : populations of /and J
C IJ* : least cost of travel between /and J

C IJK : cost via mode K
HIJ : shortest travel time between I and J
HIJK : travel time via mode K
DIJ* : departure frequency of the most frequent
DIJK : departure frequency of mode K
YIJ : weighted average incomes of I and J
a0...a8 : calibration parameters
This model is a simultaneous trip-generation/trip-distribution modal choice equation
employing land-use variables (populations), socio-economic characteristics (income
levels), and interzonal impedances by mode (costs, travel times, and frequency of
service) to estimate the interzonal demands by mode (QIJK).
In keeping with the purpose of the demand-estimating methodology, these interzonal
flows would presumably be assigned to the networks of the modes K serving the
region to find the equilibrium link flows.
The earlier-raised question of consistency between the assumed levels of some of the
explanatory variables (travel times, for example) and the levels implied in the results
of the assignment phase resurfaces.
In urban situations, the calibration and application of such large models is, to say the
least, cumbersome.
However, they may be useful for rather coarse estimates at the regional level if the
number of zones and the degree of detail in specifying the transportation network
are kept to a minimum.
1.5.2 The Implicit Demand Models

In this implicit demand models, the outputs of each step-become inputs to the
following step, which also takes relevant inputs from the specification of the
alternative plan under study and from the land-use and socio-economic projection
phase, and this widely known as the four-step models.
Some comments on the four-step models:
Several criticisms have been levied against the traditional four-step travel-
demand estimation process.
Among them are that it is cumbersome, expensive, and requires a large
amount of data.
It is based on cross-sectional data obtained at one point in time, and the
results are reached through iterative processes rather than through direct
optimal solutions.
The errors at one step are likely to magnify at other steps, since the inputs to
one model are the outputs of another.

The generation of trips is Independent of the transportation supply
characteristics and possible technological improvements, and the models are
generally site-specific-that is, they are not transferable from one urban area to
another.
Despite all those criticisms, the process is still the most used, primarily
because it has been well jested and is completely operational.
Several modifications and refinements have been carried out on this
procedure, such as the inclusion of transit assignment and disaggregate
behavioral modal-split models.
Quick-response travel-demand estimation is gaining popularity.
However, it is not designed or meant to replace the four-step process.
Simplified procedures, mainly based on the four-step sequence, have been
developed the used in small urban areas to avoid the costs and the time
needed to run the large models.
A fair amount of research work has been conducted in the travel forecasting
field, but so far has failed to come up with any radical departure from the set
of four models that have been described in this lecture.
In contrast to the four-step process and to disaggregate models are those that
are highly aggregate in nature.
However these are relatively new and not widely understood or utilized vet.
The four-step process generally has remained the key approach to date.

Figure 1-5 Four Basic Models Used in Transportation Planning
1.6 The Urban Transportation Planning Process
UTP = UT (studies) + UT (models)

UT (studies) measure existing conditions at t = 0
UT (models) determining conditions at future time t
Many possible ways to describe the UTP
Example 1.1
Figure 1-6 Urban Transportation Planning Process
1.6.1 Formulation of goals and objectives

State the purpose of the planning exercise
Identify the issues
Recognize the problems and constraints
Goal is an ideal, and as such is expressed in abstract terms - (goal is
abstract and unattainable)
Objective is a specific statement denoting a measurable end to be
reached or achieved.
Example
Goal Objective
Improve public safety Reduce occurrence of traffic accidents
Reduce injuries and deaths resulting from traffic accidents
Improve personal security of urban travelers
1.6.2 Inventory
Data collection (collect relevant data)
population pattern and quantity
land use activity and intensity
socio-economic income, labor forces, car ownership
travel pattern O-D studies
Survey of existing facilities and current behavior
street classification
volume studies
travel time studies
accident studies
parking studies
traffic control devices studies
physical street system ... evaluation for present and future use ... widths,
geometries, etc.
1.6.3 Analytical methods

analyze the data
develop appropriate quantitative methods to understand the system
use information gathered in the inventory step to develop and calibrate
(mathematical) models
the models are then used in forecasting the travel demand for the future
1.6.4 Forecasting
derive forecasts of future situation
future requirements
predict future conditions .in the urban area population, socio-economic
activity,
land use activity, etc.
general target year or "horizon year" or "design year" about 5-15-20 years.
1.6.5 Formulating of alternative plans
several alternative plans should be developed
1.6.6 Evaluation
choose most appropriate plan
establish performance of each plan
evaluation methods
examining and testing all of the alternatives the plan that "best" satisfies
the community goals and objectives and is technically and economically
feasible
economic factor: cost-benefit analysis
environmental and social factors
1.6.7 Implementation
recommend appropriate course of action
capital work programs
detailed design
land acquisition, bid-letting

funding
Example 1.2
UTP involves a sequence of information-gathering and decision-making steps.
These include:
Goal definition, that is, a determination and statement of the goals of the
transportation system. These are often based on the community values, as identified
by the planner. Goals are often expressed in very broad terms, with more specific
objectives and (possibly) measures of effectiveness then being deduced from the
goals.
Identification of needs. This involves comparison of the actual performance of the
transportation system with its goals, objectives, and measures of effectiveness.
Where performance falls short of the standard, some sort of improvement is
"needed."
Development of alternative solutions to address each need identified.
Evaluation of alternative solutions, in terms of physical, economic, and financial
feasibility; cost; demand; environmental impact; etc.
A decision process, in which particular alternatives are selected for implementation.
1.6.8 Information needed
Typical information required can be grouped into four categories, enumerated as
below.
Socio-economic data: Information regarding the socio-economic characteristics of
the study area. Important ones include income, vehicle ownership, family size, etc.
This information is essential in building trip generation and modal split models.
Travel surveys: Origin-destination travel survey at households and traffic data from
cordon lines and screen lines. Former data include the number of trips made by each
member of the household, the direction of travel, destination, the cost of the travel,
etc. The latter include the traffic flow, speed, and travel time measurements. These
data will be used primarily for the calibration of the models, especially the trip
distribution models.
Land use inventory: This includes data on the housing density at residential zones,
establishments at commercial and industrial zones. This data is especially useful for
trip generation models.
Network data: This includes data on the transport network and existing inventories.
Transport network data includes road network, traffic signals, junctions etc. The
service inventories include data on public and private transport networks. These
particulars are useful for the model calibration, especially for the assignment models.

1.6.9 Study area
Once the nature of the study is identified, the study area can be defined to
encompass the area of expected policy impact. The study area need not be confirmed
by political boundaries, but bounded by the area influenced by the transportation
systems. The boundary of the study area is defined by what is called as external
cordon or simply the cordon line. A sample of the zoning of a study area is shown in
Figure 1-7
Zone Study area
Figure 1-7 Zoning of a study area
Interactions with the area outside the cordon are defined via external stations which
effectively serve as doorways to trips, into, out of, and through the study area.
In choosing the study area one must consider the decision-making context, the
schemes to be modeled, and the nature of the trips of interest: mandatory,
optional, long or short distance, and so on.
For strategic studies one would like to define the study area so that the
majority of the trips have their origin and destination inside it; however, this
may not be possible for the analysis of transport problems in smaller urban
areas where the majority of the trips of interest are through-trips and a bypass
is to be considered.
Study area should be defined such that majority of trips have their origin and
destination in the study area and should be bigger than the area-of-interest
covering the transportation project.
1.6.10 Zoning
Once the study area is defined, it is then divided into a number of small units called
traffic analysis zones (TAZ) or simply zones. The zones with in the study area are
called internal zones.
Zones are modeled as if all their attributes and properties were concentrated in a
single point called the zone-centroid. The centroids are connected to the nearest
road junction or rail station by centroid connectors. Both centroid and centroid
connectors are notional and it is assumed that all people have same travel cost from
the centroid to the nearest transport facility which is the average for a zone. The
intersection from outside world is normally represented through external zones. The
external zones are defined by the catchment area of the major transport links feeding
to the study area. Although the list is not complete, few guidelines are given below
for selecting zones.
zones should match other administrative divisions, particularly census zones.
zones should have homogeneous characteristics, especially in land use,
population etc.
zone boundaries should match cordon and screen lines, but should not match
major roads.
zones should be as smaller in size as possible so that the error in aggregation
caused by the assumption that all activities are concentrated at the zone
centroids is minimum.
zones do not have to be of equal size. It should be generated smaller zones in
congested than in uncongested areas.
Figure 1-8 A road network coded as nodes and links

1.6.11 Network
Transport network consists of roads, junctions, bus stops, rails, railway station etc.
Normally road network and rail network are represented separately. Road network is
considered as directed graph of nodes and links. Each node and links have their own
properties. Road link is normally represented with attributes like starting node,
ending node, road length, free flow speed, capacity, number of lanes or road width,
type of road like divided or undivided etc. Road junctions or nodes are represented
with attributes like node number, starting nodes of all links joining the current node,
type of intersection (uncontrolled, roundabout, signalized, etc.). Similarly public
transport network like bus transit network and rail network are represented, but with
attributes relevant to them. These may include frequency of service, fare of travel,
line capacity, station capacity etc. This completes the inventory of base-year
transportation facility.

2. DATA AND SPACE
2.1 Basic Sampling Theory
2.1.1 Statistical Considerations

Statistics may be defined as the science concerned with gathering, analyzing and
interpreting data in order to obtain the maximum quantity of useful information.
Data usually consist of a sample of observations taken from a certain population of
interest which is not economically feasible to observe in its entirety.
Sample design aims at ensuring that the data to be examined provide the greatest
amount of useful information about the population of interest at the lowest possible
cost. Two difficulties exist:
How to ensure a representative sample;
How to extract valid conclusions from a sample satisfying the above condition.
2.1.1.1 Basic Definition
Sample is defined as a collection of units which has been especially selected to
represent a larger population with certain attributes of interest.
Population of interest is the complete group about which information is sought.
Sampling method:
Simple random sampling is not only the simplest method but constitutes the
basis of all the rest.
Stratified random sampling: A priori information is first used to subdivide the
population into homogeneous strata and then simple random sampling is
conducted inside each stratum using the same sampling rate.
Choice based sampling consists in stratifying the population based on the
results of the choice process under consideration. A major advantage is that
data may be produced at a much lower cost than with the other sampling
methods. However, it may not be random and therefore the risk of bias in the
expanded values is greater.
Sampling error is simply due to the fact that we are dealing with a sample and not
with the total population. This type of error does not affect the expected values of
the means of the estimated parameters. It only affects the variability around them.
Sampling bias is caused by mistakes made either when defining the population of
interest, or when selecting the sampling method, the data collection technique and
any other part of the process.
It differs from the sampling error:
Transportation Systems Planning (ECIV 540) | DATA AND SPACE 24

It can affect not only the variability around the mean of the estimated
parameters but the values themselves.
While the sampling error may not be avoided, the sampling bias may be
eliminated by taking extra care during the various stages of sampling design
and data collection.
Sample size: Determining sample size is a problem of trade-offs, as:
A much too large sample is too expensive;
A too small sample may lead to unacceptable results.
2.1.1.2 Sample size to estimate population parameters
This depends on three main factors: variability of the parameters in the population,
degree of accuracy and population size.
Consider a population of size N and a specific property which is distributed with mean
and variance 2. The Central Limit Theorem states that the distribution of the mean
( ) of successive samples is distributed Normal with mean and standard deviation
se( ) , known as the standard error of the mean:
se( ) = ( ) /[ ( 1)]
If only one sample is considered, the best estimate of is , and the best estimate of
2 is S2 (the sample variance). Thus:
se( ) = ( ) /
For large populations and small sample sizes:
se( ) = /
Sample size may be estimated by two stages. First, calculate n from the equation
above:
=
( )
Then, correcting for finite population size, if necessary, by
S can only be calculated once the sample has been taken, so it has to be estimated
from other sources. The standard error of the mean is related with the desired
degree of confidence to be associated with the use of the sample mean as an
estimate of the population mean. Therefore:
A confidence level for the interval must be chosen (E.g. the typical 95% level
implies an acceptance to err in 5% of cases);
It is necessary to specify the limits of the confidence interval around the mean.

Example 2.1
If a Normal distribution is assumed and a 95% confidence level is specified, this

means that a maximum value of 1.96se( ) would be accepted for the confidence
interval. If a 10% error is specified we would get the interval ( 0.1) and it may be
seen that:
se( ) = 0.1/1.96 = 0.051
n' = (s/0.051)2 = 384CV2
2.1.1.3 Obtaining the Sample
Example 2.2
Consider a certain area the population of which may be classified in groups according
to: automobile ownership and household. Let assume that m observations are
required by cell in order to guarantee a 95% confidence level in the estimation of trip
rates.
Car ownership Household size % of population
With car Four or less 9

More than four 16
Without car Four or less 25

More than four 50
There are ways to proceed:

1. Achieve a sample with m observations by cell by means of a random sample. The
sample size would be:
n = (1 + + + )m = 11.1m
2. One can undertake first a preliminary random survey of 11.1m households. This
low-cost survey can be used to obtain m observations even in the smallest cell.
Subsequently, it would be sufficient to randomly select a (stratified) sample of 3m
observations from the other cells to be interview in detail.
2.1.2 Conceptualization of the Sampling Problem

We assume that each sampled observation may be described on the basis of the
following two variables:
i = Observed choice of the sample individual;
X = Vector of characteristics (attributes) of the individual (age, sex, income, car
ownership) and of the alternatives in his choice set (waiting time, travel cost, ..).

= Parameters of a model;
The joint distribution of i and X is : P(i, X/)
The probability of choosing alternative i among a set of options with attributes X is:
P(i/X, )
2.1.2.1 Random sample
The distribution of i and X in the sample and population is:
f(i, X/) = P(i/X, )
2.1.2.2 Stratified Sample

The sampling process is defined by a function f(X), giving the probability of finding an
observation with characteristics X.
f(i, X/) = P(X) . P(i/X, )
2.1.2.3 Choice based Sampling

The sampling procedure is defined by a function f(i), giving the probability of finding
an observation that chooses i.
f(i, X/) = f(i) . P(X/i, )
Example 2.3
Assume that for the purposes of a transport study the population of a certain area
has been classified according to two income categories, and that there are only two
modes of transport available (car and bus) for the journey to work. Let assume that
the population distribution is given by:
Low income High income Total
Bus user 0.45 0.15 0.60

Car user 0.20 0.20 0.40
Total 0.65 0.35 1.00
1. Random sample. If a random sample is taken, it is clear that the same population
distribution would be obtained.
2. Stratified sample. Consider a sample with 75% low income (LI) and 25% high
income (HI) travelers. From the previous table it is possible to calculate the
probability of a low-income traveler using bus, as:
P(Bus/LI) = P(LI and Bus) / [P(LI and Bus) + P(LI and Car)] = 0.45/ (0.45 + 0.20) = 0.692
The probability of finding a bus user with low income in the sample is: 0.75 0.692 =
0.519. Doing this for the rest of the cells:

Bus user 0.519 0.107 0.626

Car user 0.231 0.143 0.374
Total 0.750 0.250 1.000
Choice-based sample. Let us assume now that we take a sample of 75% bus users
and 25% car users. The probability of a bus user having low income may be calculated
as:
P(LI/Bus) = P(LI and Bus) / [P(LI and Bus) + P(HI and Bus)] = 0.45 / (0.45 + 0.15) = 0.75
Therefore, the probability of finding a low-income traveler choosing bus in the
sample is 0.75 0.75 = 0.563.
Bus user 0.563 0.187 0.750

Car user 0.125 0.125 0.250
Total 0.688 0.312 1.000
Each sampling method produces in general a different distribution in the sample. The
importance of the above example will increase when we consider what is involved in
the estimation of models using the various samples.
2.1.3 Practical consideration in sampling
2.1.3.1 The implementation problem
Stratified (and choice based) sampling requires random sampling inside each stratum.
It is necessary to isolate the relevant group and this may be difficult in some cases.
Another problem is that in certain cases even if it is possible to isolate all
subpopulations and conforming data, it may be difficult to ensure a random sample
inside each stratum.
2.1.3.2 Finding the size of each subpopulation
Given certain stratification, there are several methods available to find out the size of
each subpopulation.
Direct measurement;
Estimation from a random sample. If a random sample is taken, the proportion
of observations corresponding to each stratum is a consistent estimator of the
fraction of the total corresponding to each subpopulation;
Solution of a system of simultaneous equations. Assume we are interested in
stratifying by chosen mode and that we have data about certain population

characteristics. Taking a small on-mode sample we can obtain modal average
values of these variables and postulate a system equations which as the
subpopulation fractions as unknowns.
Example 2.4
Assume the following information is available:

Average income of population (I): 33 600 $/year
Average car ownership (CO): 0.44 cars/household
Assume also that small on-mode surveys yield the following:
Mode I ($ / year) CO (cars/household)
Car 78 000 1.15

Bus 14 400 0.05
Metro 38 400 0.85
If Fi denotes the subpopulation fraction of the total, the following system of

simultaneous equations holds:
33 600 = 78 000F1 + 14 400F2 + 38 400F3
0.44 = 1.15F1 + 0.05F2 + 0.85F3
1 = F1 + F2 + F3
the solution of which is:
F1 = 0.2451
F2 = 0.6044
F3 = 0.1505
If the total population was 180 000 inhabitants, there would be 44 100 car users 108
800 bus users and 27 100 metro users.
2.2 Error in Modeling and Forecasting
2.2.1 Different Types of Error

Measurement Errors: These occur due to the inaccuracies inherent in the process of
actually measuring the data in the base year, such as: questions badly registered by
the interviewee, answers badly interpreted by the interviewer, network
measurement errors.
Sampling Errors: These arise because the models must be estimated using finite data
sets.

Computational Errors: These arise because models are generally based on iterative
procedures for which the exact solution, if it exists, has not been found for reasons of
computational costs.
Specific Errors: These arise either because the phenomenon being modeled is not
well understood or because it needs to be simplified. The types of error are following:
Inclusion of an irrelevant variable
Omission of a relevant variable
Exclusion of taste variations on the part of the individuals
The use of model forms is not appropriate, etc.
Transfer Errors: These occur when a model developed in one context (time and/or
place) is applied in a different one.
Aggregation Errors: These arise basically out of the need to make forecasts for groups
of people while modeling often needs to be done at the level of individual in order to
capture behavior better.
Data aggregation: When network models are used there is aggregation over
routes, departure times and zones, this means that the values thus obtained
for the explanatory variables are averages for groups of travelers rather than
exact values for any particular individual.
Aggregation of alternatives: It may not be feasible to attempt to consider the
whole range of options available to each traveler (minibuses, bus, ..).
Model aggregation

Figure 2-1 Attribute measurement and choice
2.3 Basic Data Collection Methods
2.3.1 Practical Considerations

Length of study: It determines indirectly how much time and effort it is
possible to devote to the data collection stage.
Study Horizon: If the design year is too close, there will not be much time to
conduct the study. On the other hand, this allow time to employ almost any
time of analytical tool, it also means that error in forecasting will only be
known in very long time later.
Limits of the Study Area: It is important to ignore formal political boundaries
and concentrate on the whole area of interest. The definition of the area of
interest depends on the type of policies examined and decisions to be made.
Study Resources: How many personnel and of what level, computing facilities
will be available for the study.
2.3.2 Type of Survey
Figure 2-2 Scope of data collection needed for a metropolitan OD survey
Household survey: trips made by all household members by all modes of transport
both within the study area and leaving/arriving to the area during the survey period;
this survey should include socio-economic information (income, car ownership, family
size and structure, etc.).
Intercept survey, external cordon: data on people crossing the study area border,
particularly non-residents of the study area. This data can also be used to check and
amplify the household data on study area crossings, since there is usually only a small
amount of data collected, even in a very large survey.
Intercept surveys, internal cordons and screen lines: these are required to measure
trips by non-residents, and again to verify household data to some extent.
Traffic and person counts: they are low cost and are required for calibration,
validation and for further checks to other surveys.
Travel time surveys: these are required to calibrate and validate most models and
may be needed for both car and public transportation travel.
Other related data: to create robust forecasting models as needed in large
metropolitan area, including:
Land use inventory; residential zones, commercial and industrial zones,
parking spaces, ..
Infrastructure and existing services inventories (public and private
transportation network, fares, frequency, signal and timing, etc.
Information from special surveys on attitudes and elasticity of demand.
2.4 Stated Preference Surveys
2.4.1 Introduction
Reveal Preference (RP) information means data about actual or observed choices
made by individuals. RP data have limitations:
Observations of actual choices may not provide sufficient variability for
constructing good models for evaluation and forecasting.
Observed behavior may be dominated by a few factors making it difficult to
detect the relative importance of other variables.
The difficulties in collecting responses for policies which are entirely new.
Stated preference (SP) methods cover a range of techniques, which have in common
the collection of data about respondent's intentions in hypothetical setting as
opposed to their actual actions as observed in real markets.
The three most common methods are Contingent Valuation (CV), Conjoin Analysis
(CA) and State Choice (SC).
2.4.2 Contingent Valuation (CV)

CV primarily deals with eliciting willingness to pay (WTP) information for various
policy or product options. In CV studies, the policy or product is kept static and the
outcome is for the entire product or policy. CV questions cannot be used to
disentangle the WTP for individual characteristics or attributes of the product or
policy under study.
Example 2.5
If no action is taken, this highway's quality is expected to deteriorate in the next few
years (show picture). To get the highway back to its current state (show picture) the
government will have to spend money and this will mean raising taxes.
1. Open-ended elicitation: What is the maximum amount that you would be
prepared to pay every year [e.g. through a tax surcharge]?
2. Bidding game elicitation:
If Yes: Interviewer keeps increasing the bid until the respondent answers No. Then
maximum WTP is elicited.
If No: Interviewer keeps decreasing the bid until respondent answers Yes. Then
maximum WTP is elicited.
3. Payment card elicitation: Which of the amounts listed below best describes your
maximum willingness to pay every year, through a tax surcharge?
$0, $1, $2, $3, $4, $5, $6, ..

2.4.3 Conjoint Analysis (CA)
CA analysis allows the researcher to examine the preferences, and even WTP if a
price or cost attribute is included, not only entire product but also individual
characteristics of the object under study. In CA, respondents are presented with a
number of alternative policies or products and are asked to either rate or rank them.
Traditional CA has had limited acceptance in transportation studies since:
The dependent variable of a linear regression model should be, at the
minimum, interval scaled.
Respondents in real life do not rate or rank alternatives and even if they did
different people would approach such scales in psychologically different
manners.
Figure 2-3 Example of stated-preference ranking exercise
2.4.4 State Choice Methods

Whereas CA asks respondents to rank or rate the alternatives, respondents
undertaking a SC survey are asked to choose their preferred alternative from unit a
subset of the total number of hypothetical alternatives constructed by the analyst.
A further distinction between the two methods is that CA tasks typically present
respondents with a relatively large number of alternatives, simultaneously, to rate or
rank, whereas SC methods typically present only a few alternatives at a time,
changing them and having respondents repeat the choice task.
On the other hand, the primary distinction between RP and SC surveys is that in the
latter case individuals are asked about what they would choose to do (or how would
they rank/rate certain options) in one or more hypothetical situations.
A crucial problem with stated preference data collection in general, is how much faith
we can put on individuals actually doing what they stated they would do when the
case arises (for example, introducing a new option). E.g. only half the people doing
what they said they would.
The main features of an SC survey may be summarized as follows:
It is based on the elicitation of respondents statements of how they would
respond to different hypothetical alternatives:
Each option is represented as a package of different attributes like travel
time, price, headway, reliability, etc.
The analyst constructs these hypothetical alternatives so that the individual
effect of each attribute can be estimated.
The researcher has to make sure that respondents are given hypothetical
alternatives they can understand, appear plausible and realistic, and relate to
their current level of experience.
The responses given by individuals are analyzed to provide quantitative
measures of the relative importance of each attribute.
Figure 2-4 Example of SC tasks

3. TRIP GENERATION
Trip generation is the first stage of the classical first generation aggregate demand
models.
The objective of a trip-generation model is to forecast the number of person-trips
that will begin from or end in each travel analysis zone within the region for a typical
day of the target year. In other words this stage answers the questions to "how many
trips" originate at each zone, from the data on household and socioeconomic
attributes.
The total number of person-trips generated constitutes the dependent variable of the
model.
The independent or explanatory variables include land use and socio-economic
factors that have been shown to bear a relationship with trip making.
In other words, trip generation models accept land use and socio-economic
characteristics as input to produce zonal trip ends, the outputs.
Mathematically
Oi = f(Li)
Dj = f(Lj)
Where:
Oi = no. of trips originating in zone i
Dj = no. of trips attracted in zone j
Li, Lj = measures of land use intensity in zones i and j
The main factors affecting personal trip production include income, vehicle
ownership, household structure and family size. In addition factors like value of land,
residential density and accessibility are also considered for modeling at zonal levels.
The personal trip attraction, on the other hand, is influenced by factors such as
roofed space available for industrial, commercial and other services. At the zonal
level zonal employment and accessibility are also used. In trip generation modeling in
addition to personal trips, freight trips are also of interest. Although the latter
comprises about 20 percent of trips, their contribution to the congestion is
significant. Freight trips are influenced by number of employees, number of sales and
area of commercial firms.
Transportation Systems Planning (ECIV 540) | TRIP GENERATION 36

Land use classification Measure of Land use activity No. of trip ends
Residential Li, no. of persons living in zone Oi

Lj, no. of workers living in zone (Km2) Oi
Lj, no. of jobs in zone Dj
Industrial Lj, industrial area in sq., km. in zone Dj
Lj, no. .of parking space in zone Dj
Commercial Lj, no. of sq. m. of office floor area Dj
Lj, no. of hotel rooms in zone Dj
Recreational Lj, seating capacity of generator in zone Dj
3.1 Types of trips
Some basic definitions are appropriate before we address the classification of trips in
detail. We will attempt to clarify the meaning of journey, home-based trip, non-
home-based trip, trip production, trip attraction and trip generation. Trips can be
classified by trip purpose, trip time of the day, and by person type.
3.1.1 By trip purpose
Trip generation models are found to be accurate if separate models are used based
on trip purpose. The trips can be classified based on the purpose of the journey as
trips for work, trips for education, trips for shopping, trips for recreation and other
trips. Among these the work and education trips are often referred as mandatory
trips and the rest as discretionary trips. All the above trips are normally home based
trips and constitute about 80 to 85 percent of trips. The rest of the trips namely non
home based trips, being a small proportion are not normally treated separately.
Journey is an out way movement from a point of origin to a point of destination,
where as the word "trip" denotes an outward and return journey. If either origin or
destination of a trip is the home of the trip maker then such trips are called home
based trips and the rest of the trips are called non home based trips. Trip production
is defined as all the trips of home based or as the origin of the non home based trips.
See Figure 3-1

Production Attraction
Home Work
Home-based trip
Work Shop
Attraction Production
Non-home-based trip
Figure 3-1 Trip Production and Attraction
3.1.2 By time of day
The second way of classification is based on the time of the day when the trips are
made. The broad classification is into peak trips and off peak trips.
Table 3-1 Example of trip classification
3.1.3 By person type

The third way of classification is based on the type of the individual who makes the
trips. This is important since the travel behavior is highly influenced by the socio
economic attribute of the traveler and are normally categorized based on the income
level, vehicle ownership and house hold size.
3.2 Trip generation by expansion factors
Early transportation studies, such as the Detroit Metropolitan Area Traffic Study,
used simple expansion (or growth) factors to estimate future trip ends for traffic
zones or districts.
In its simplest form, the method relates data collected in the movement studies to
data collected in the land-use survey, to develop a trip generation rate for major land
uses. Its basic equation is:
Ti = Fi ti
where Ti and ti are respectively future and current trips in zone i, and Fi is a growth
factor. Normally the factor is related to variables such as population (P), income (I)
and car ownership (C), in a function such as:
( , , )
=
( , , )
where f is a function with no parameters, and d and c are the design and current
years respectively.
Example 3.1 The total population of Gi-Gi City in 2000 was 1,000,000 people and
there were 250,000 total trips made in 2000. If the total population in the horizon
year is expected to be 1,250,000 persons, then what would be the total trips in
horizon year?
Answer:
Expansion factor = 1,250,000/1,000,000
= 1.25
Future trips = 1.25 x 250,000 = 312,500 trips
Example 3.2 The following information were collected for RCA City in 2005. The
forecasted information of 2015 are also provided for RCA City.
Year Population No. of Workers Car Ownerships

2005 800,000 300,000 100,000
2015 900,000 320,000 120,000
If the total auto trips/day in 2005 were 125,000 trips, then what would be the total
auto trips in 2015?
Assuming that the auto trips/day = f(population, no. of workers, car ownerships)
Solution:
This method is called "Simple Growth Factors"
Growth Factors = (900,000 /800,000); (375,000 /300,000); (120,000 /100,000)
= 1.125 ; 1.25 ; 1.2
Overall growth factor = 1.125 x 1.25 x 1.2 = 1.69
Total auto trips in 2015 = 1.69 x 125,000 = 211,250 trips

Example 3.3
Consider a zone with 250 households with car and 250 households without car.
Assuming the average trip generation rates of each group:
car-owning households produce: 6.0 trips/day
non-car-owning households produce: 2.5 trips/day
The current number of trips per day is:
ti = 250 2.5 + 250 6.0 = 2125 trips/day
Assuming that in the future all households will have a car and income and population
remain constant. A growth factor is:
Fi = Cdi /Cci = 100% / 50% = 2
We could estimate the number of future trips as:
Ti = 2 2125 = 4250 trips/day
However, if we use our information about average trip rates, we could estimate the
future number of trips as:
Ti = 500 6 = 3000
which means that the growth factor method would overestimate the total number of
trips by approximately 42%.
In general growth factor methods are mostly used in practice to predict the future
number of external trips to an area; this is because they are not too many in the first
place (so errors cannot be too large) and also because there are no simple ways to
predict them.
3.3 Cross-Classification Models
Cross-classification is a technique for trip generation that uses categories or cells in

which the changes in one dependent variable (trips) can be measured when the
changes in other independent variables (land use, socio-economic status) are
analyzed.
Example 3.4
An urban zone contains 200 acres of residential land, 50 acres devoted to commercial
uses, and 10 acres of park land. The following table presents the zone's expected
household composition at some future (target) year.

Vehicles per Persons per household
household 1 2,3 4 5
0 100 200 150 20
1 300 500 210 50
2+ 150 100 60 0
Using the calibrated cross-classification table above, estimate the total non-work
home-based trips that the zone will produce during a typical target-year day. The
rates are given as trips per household per day.
Table 3-2 Example of Total Home-based, Non-work Trip Rates
Vehicles available per Persons per

Cross class household
household area type
1 2,3 4 5+
1. Urban-high density 0 0.57 2.07 4.57 6.95
1 1.45 3.02 5.52 7.90
2+ 1.82 3.39 5.89 8.27
2. Suburban-medium density 0 0.97 2.54 5.04 7.42
1 1.92 3.49 5.99 8.37
2+ 2.29 3.86 6.36 8.74
3. Rural-low density 0 0.54 1.94 4.44 6.82
1 1.32 2.89 5.39 7.77
2+ 1.69 3.26 5.76 8.14
Solution:
The total trip productions are estimated by summing the contribution of each
household type:
where Pi = Total trips produced in zone i

Nh = No. of households of type h
Rh= Trip rates that are corresponding to households of type h
Pi = 5760 trips/day
Note: 5760 = 100 x 0.57 + 300 x 1.45 + ...

3.4 The FHWA-Simplified Trip-Production Procedure
Example 3.5 A number of suburban zones have a total of 1000 dwelling units (DU).
The average income per DU is $12,000. Using the curves a, b, and c provided,
estimate the number of trips produced by the zones.
Solution:
Refer to Figure below.

1. Enter curve a with zonal income per dwelling unit to determine car ownership level
by household:
2% 0 auto households = 20 dwelling units
2. Enter curve b with income, to determine the total production (person-trips) from
each household:
Trips from 0 auto household = 5.5 trips/DU x 20 DU

= 110 trips
= 3840 trips
= 8060 trips
= 2408 trips
Total trips- = 14,418
Average trips/dwelling unit = 14.4
3. Enter curve with income to determine the trips produced by purpose:
Home-to-work trips = 19% x 14,418 = 2739 trips
Home-to-shop trips = 11% x 14,418 = 1586 trips
Home-to-school trips = 14% x 14,418 = 2018 trips
Home-to-other trips = 34% x 14,418 = 4903 trips
Non-home-based = 22% x 14,418 = 3172 trips
Total = 14,418 trips

Figure 3-2 Example of Urban Trip Production Procedure
3.5 Regression Analysis
Regression is a mathematically based procedure that has been programmed for most
electronic computers.
The technique is, therefore, readily available to the analyst, and many statistical
software packages exist that will perform regression analyses easily and quickly.
In trip generation step, no. of trip ends are treated as dependent variable, which is a
function of one or more independent variables.
Mathematically,
Pi = f(x 1 ,x 2 ,x 3 , ... )
Aj = f (k1, k2, k3, ... )
where
Pi : trip productions from zone i
Aj : trip attractions to zone j
Xi : predictive factors for productions (independent variables)
Ki : predictive factors for attractions (independent variables)
Example:
Pi = a + bx1 + cx2 + dx3
Pi = trip productions
X1 = no. of car ownerships
X2 = family income
X3 = no. of population
a, b, c, d = parameters determined through a calibration process
Model parameters and variables vary from one study area to another and are
established by using base-year information.
Once the equations are calibrated, they are used to estimate future travel for a target
year.
3.5.1 Forms of Regression
Linear Regression
Simple linear: Y = ax + b
Multiple linear: Y = a + bx1 + cx2 + dx3 + ...
Non-linear Regression
Y= eax + b

Y = ax2 + b
Y = ax + bx2 + c ,
Multiple Linear Regression
Four guidelines are helpful in deciding which independent variables to include in a
linear regression model. The selected explanatory variables
Must be linearly related to the dependent variable
Must be highly correlated with the dependent variable
Must not be highly correlated between themselves
Must lend themselves to relatively easy projection.
3.5.2 Correlation
Coefficient of determination, R2
Coefficient of correlation, R
The goodness of fit a regression line increases with the proportion of the total
variation that is explained by the line. The coefficient of determination, R2
quantifies this fact.
It ranges from zero when none of the total variation is explained by the
regression line to unity when all the variation is explained by the line.
It is denoted as a squared quantity to capture the fact that it is always non
negative.
The square root of the coefficient of determination is called the coefficient of
correlation.
Its value can range from -1 to +1.
In the case of linear regression, the sign of R is the same as the sign of the
slope b of the regression line.
Figure below illustrates that R is near +1, there exists a high positive
correlation;
If it is near -1, there exists a high negative correlation;
And if it is around zero, there exists no correlation between X and Y
The following formula gives the proper magnitude and sign for R2.
R 2

n x y x y
i i i i
2
n x x n y y
2
i
2
i
2
i i
2
Example 3.6
If the coefficient of correlation, R of the two regression models are R1 = - 0.95 and
R2= -0.39, which of the two models are superior ?
Solution:

Since both correlations are negative, so the first model represents a better fit than
the second.
Note: R2 = 0.49 Meaning = ???
R2 = 0.49 means that 49 % of the variance of the dependent variable can be
explained by the model.
R2 scores ranging between 0.30 and 0.60 generally denote a good model fit.
R2 scores less than 0.20 represent poor model fit, and such models should be
used with particular caution for real-world applications.
R2 scores greater than 0.70 when derived from large samples (i.e. a few
hundred observation or more) may be suspicious.
Co linearity effects and other problems may be resulting in exaggerated
goodness of fit.
3.5.3 Hypothesis tests for a
A typical null hypothesis is H0: a = 0 is: t = a/Sa
This value needs to be compared with the critical value of the Student statistics for a
given significance level and the appropriate number of degrees of freedom. One
problem is that the alternative hypothesis H1 may imply unilateral (a > 0) or bilateral
(a not equal 0) tests; this can only be determined by examining the phenomenon
under study.
Example 3.7
Assume we are interested in studying the effect of income (I) in the number of trips
by non-car-owning households (T), and that we can use the following relation:
T = aI + b
As in theory we can conclude that any influence must be positive (i.e. higher income
always means more trips) in this case we should test H0 against the unilateral
alternative hypothesis H1: a > 0. If H0 is true, the t-value is compared with the value
t;d, where d are the appropriate number of degrees of freedom, and the null
hypothesis is rejected if t > t;d.
Figure 3-3 Rejection region for = 5%

On the other hand, if we were considering incorporating a variable the effect of
which in either direction was not evident (for example, number of female workers, as
these may or may not produce more trips than their male counterparts), the null
hypothesis should be the bilateral H1: a = 0, and H0 would be rejected if 0 is not
included in the appropriate confidence interval for a.
Example 3.8
Consider the variables trips per household (Y), number of workers (X1) and number of
cars (X2). The Table below presents the results of successive steps of a stepwise
model estimation; the last row also shows (in parenthesis) values for the t-ratio.
Assuming large sample size, the appropriate number of degrees of freedom (n 2) is
also a large number so the t-values may be compared with the critical value 1.645 for
a 95% significance level on a one-tailed test.
Table 3-3 Example of stepwise regression
Model Equation R2
1 Y = 2.36 X1 0.203
2 Y = 1.80 X1 + 1.31 X2 0.325
3 Y = 0.91 + 1.44X1 + 1.07X2 0.384
(3.7) (8.2) (4.2)
Example 3.9
Find the coefficients a and b in linear regression: y = ax + b

Solution:
We need to find the best fit straight line - the one that passes as closely as possible to
as many points as possible.
If yi is the measured value of the dependent variable of data i;
is the calculated value of the dependent variable of data i;
i = yi -
Objective of the regression: To find out the relationship that minimizes the sum of
squares of errors. It means that must be minimized.
i = yi -
--> (i) = ( yi - )2
2
= yi2 - 2yi +
But = ax + b:
(i)2 = yi2- 2yi (axi + b) + (axi + b)2

= yi2- 2yi axi- 2yib + a2xi2 + 2axib + b2
--> = yi2- 2ayi xi- 2byi + a2xi2 + 2abxi + nb2 needs to minimize.
A = minimizes --> = 0, = 0:
= -2yi xi + 2axi2 +2bxi = 0 (1)
= -2yi + + 2nb = 0 (2)
From (1): axi2 + bxi = yi xi (3)

From (2): axi + nb = yi (4)
Let xav = xi /n, yav = yi /n,
From (4): b = yav - axav
Put it into (3):

=

( )
=

Example 3.10
Let the trip rate of a zone is explained by the household size done from the field
survey. It was found that the household size are 1, 2, 3 and 4. The trip rates of the
corresponding household are as shown in the table below. Fit a linear equation
relating trip rate and household size.
Household size
Zone
1 2 3 4
1 1 2 4 6
2 2 4 5 7
3 2 3 3 4
Total 5 9 12 17
The linear equation will have the form y = ax + b where y is the trip rate, and x is the
household size, a and b are the coefficients. For a best fit, a is given by

=
( )
=
= 31 + 32 + 33 + 34 = 30
= 90

= 5 + 9 + 12 + 17 = 43
= 127
= 43/12 = 3.58
= 30/12 = 2.5

= = 1.3
( )
= = 0.33
y = 1.3x + 0.33

4. TRIP DISTRIBUTION
After trip generation step, the analyst knows the numbers of trip productions and
trip attractions that each zone will have (we know Oi and Dj).
But where do the attractions in zone come from and where do the productions
go?
What are the zone-to-zone travel volumes?
Trip distribution procedures determine where the trips produced in each zone
will go ... how the trips produced in a zone are distributed among all of the other
zones.
Methods
1. Growth factor methods
2. Fratar method
3. Gravity Models
4.1 Definitions and notations
Trip matrix
The trip pattern in a study area can be represented by means of a trip matrix or
origin-destination (O-D) matrix. This is a two dimensional array of cells where rows
and columns represent each of the zones in the study area. The notation of the trip
matrix is given in Table 4-1.
Table 4-1 OD matrix
1 2 .. j .. n Oi
1 V11 V12 V1j V1n O1
2 V21 V2n O2
..
i Vij Oi
..
m Vm1 Vmn On
Dj D1 D2 Dj Dn
The cells of each row i contain the trips originating in that zone which have
destinations in the corresponding columns. Vij is the number of trips between origin i
and destination j. Oi is the total number of trips originating in zone i and Dj is the total
number of trips attracted to zone j. The sum of the trips in a row should be equal to
Transportation Systems Planning (ECIV 540) | TRIP DISTRIBUTION 50

the total number of trips emanating from that zone. The sum of the trips in a column
is the number of trips attracted to that zone. These two constraints can be
represented as: Vij = Oi, Vij = Dj .If reliable information is available to estimate
j i
both Oi and Dj, the model is said to be doubly constrained. In some cases, there will
be information about only one of these constraints, the model is called singly
constrained.
Generalized cost
One of the factors that influences trip distribution is the relative travel cost between
two zones. This cost element may be considered in terms of distance, time or money
units. It is often convenient to use a measure combining all the main attributes
related to the disutility of a journey and this is normally referred to as the generalized
cost of travel. This can be represented as
Cij =a1tvij + a2twij + a3ttij + a4 Fij + a5j +
where tvij is the in-vehicle travel time between i and j,
twij is the walking time to and from stops,
ttij is the waiting time at stops,
Fij is the fare charged to travel between i and j,
j is the parking cost at the destination,
is a parameter representing comfort and convenience,
a1, a2, a3, a4, a5 are the weights attached to each element of the cost
function.
4.2 Growth factor methods
Three different growth factor methods of trip distribution have been developed, each
based on the assumption that present travel patterns can be projected into the
future, using expected differential zonal rates of growth.
The three growth factor methods in chronological order of their development are
1. Uniform growth factor
2. Singly constrained growth factor
3. Doubly constrained growth factor
4. Average factor
5. Detroit
4.2.1 Uniform growth factor

The uniform factor is the oldest and simplest method of projecting future trip
distribution.

A single growth factor is calculated for the entire area under study, and this is used to
multiply all existing interzonal movements to produce estimates of future interzonal
movements.
Mathematically this can be expressed as
Vtij = V0ij F
Vtij Future no. of trips from zone i to zone j
V0ij Present no. of trips from zone i to zone j
F Average area wide growth factor
Example 4.1
Consider the simple four-by-four base-year trip matrix of Table 5.2. If the growth in
traffic in the study area is expected to be of 20% in the next three years, it is a simple
matter to multiply all cell values by 1.2 to obtain a new matrix.
1 2 3 4
1 5 50 100 200 355
2 50 5 100 300 455
3 50 100 5 100 255
4 100 200 250 20 570
205 355 455 620 1635
The assumption of uniform growth is generally unrealistic except perhaps for very
short time spans of, say, one or two years. In most other cases one would expect
differential growth for different parts of the study area.
1 2 3 4
1 6 60 120 240 426
2 60 6 120 360 546
3 60 120 6 120 306
4 120 240 300 24 684
246 426 546 744 1962
4.2.2 Singly constrained growth factor

Consider the situation where information is available on the expected growth in trips
originating in each zone, for example shopping trips. In this case it would be possible
to apply this origin-specific growth factor (i) to the corresponding rows in the trip
matrix. The same approach can be followed when the destination-specific growth
factors (j) would be applied to the corresponding columns.
Tij = i tij for origin-specific factors
Tij = j tij for destination-specific factors
Example 4.2
Consider the Table below with growth predicted for origins:

1 2 3 4 Target Oi
1 5 50 100 200 355 400
2 50 5 100 300 455 460
3 50 100 5 100 255 400
4 100 200 250 20 570 702
205 355 455 620 1635 1962
This problem can be solved immediately by multiplying each row by the ratio of
target Oi over the base year total (j ).
1 2 3 4 Target Oi
1 5.6 56.3 112.7 225.4 400 400
2 50.5 5.1 101.1 303.3 460 460
3 78.4 156.9 7.8 156.9 400 400
4 123.2 246.3 307.9 24.6 702 702
257.7 464.6 529.5 701.2 1962 1962
4.2.3 Doubly constrained growth factor

When information is available on the growth in the number of trips originating and
terminating in each zone, we know that there will be different growth rates for trips
in and out of each zone and consequently having two sets of growth factors for each
zone. This implies that there are two constraints for that model and such a model is
called doubly constrained growth factor model. One of the methods of solving such a
model is given by Furness who introduced balancing factors ai and bj as follows:
Tij = tij ai bj
In such cases, a set of intermediate correction coefficients are calculated which are
then appropriately applied to cell entries in each row or column. After applying these
corrections to say each row, totals for each column are calculated and compared with
the target values. If the differences are significant, correction coefficients are
calculated and applied as necessary. The procedure is given below:
1. Set bj = 1
2. With bj solve ai for to satisfy trip generation constraint.
3. With ai solve bj for to satisfy trip attraction constraint.
4. Update matrix and check for errors.
5. Repeat steps 2 and 3 till convergence.
Here the error is calculated as: E = |Oi - Oi1| + |Dj - Dj1| where Oi corresponds to
the actual productions from zone i and Oi1 is the calculated productions from that
zone. Similarly Dj are the actual attractions from the zone j and Dj1 are the calculated
attractions from that zone.

Example 4.3
1 2 3 4 Target Oi
1 5 50 100 200 355 400
2 50 5 100 300 455 460
3 50 100 5 100 255 400
4 100 200 250 20 570 702
i 205 355 455 620 1635
Target Dj 260 400 500 802 1962
The solution to this problem, after three iterations on rows and columns can be
shown as:
1 2 3 4 Target Oi
1 5.25 44.12 98.24 254.25 401.85 400
2 45.30 3.81 84.78 329.11 462.99 460
3 77.04 129.50 7.21 186.58 400.34 400
4 132.41 222.57 309.77 32.07 696.82 702
i 260 400 500 802 1962
Target Dj 260 400 500 802 1962
4.2.4 Average factor
The average factor method was an early attempt to take some account of the
differential rates of growth of movement which occur in urban areas.
It utilizes a growth factor for each zone within the study area which, like the
uniform factor method, is derived from land-use and trip generation
predictions.
Mathematically it can be expressed:
Vtij = V0ij (Fi + Fj)/2
Vtij Future no. of trips from zone i to zone j

V0ij Present no. of trips from zone i to zone j
Fi Growth factor of zone i
Fj Growth factor of zone j
4.2.5 Detroit
The Detroit method of trip distribution was developed in connection with the Detroit
Metropolitan Area Traffic Study, in an attempt to overcome the shortcomings of the
simpler growth factors.

This method introduces the assumption that although the no. of trips generated in
zone i will increase as predicted the appropriate growth factor Fi, these will be
distributed to zone j in proportion to the appropriate growth factor Fj divided by the
growth factor for the area as a whole.
Mathematically
Vtij = V0ij (Fi Fj)/FTB
Fi Growth factor of zone i
Fj Growth factor of zone j
FTB Average area wide growth factor. FTB= Fi/n
Example 4.4 Trip interchange between zone 1 and zone 2 in the base year is 10,000
trips. In the next 10 years forecast period, zones 1 and 2 are expected to grow at
rates of 1.75 and 2.05, respectively. The average rate of growth area wide is expected
to be 2.2. Compute the future trip interchange between zones 1 and 2.
Solution
Vtij= 10,000(1.75 2.05)2.2 = 16,307 trips
4.2.6 Advantages and limitations of growth factor model

The advantages of this method are:
Simple to understand.
Preserve observed trip pattern.
Useful in short term-planning.
The limitations are:
Depends heavily on the observed trip pattern.
It cannot explain unobserved trips.
Do not consider changes in travel cost.
Not suitable for policy studies like introduction of a mode.
4.3 The Fratar Method
A technique for trip distribution utilizing growth factors was introduced by

Thomas J. Fratar.
While the technique is seldom used now as a study wide distribution
model, the technique is considered by many to be a particularly useful way
of dealing with external to external trips, that is, between external stations
of the study area.
The method is applied iteratively with the interchanges being computed
according to the relative attractiveness of each interzonal movement from
the point under consideration.
The method involves the following steps:

1. Future traffic growth is estimated for each traffic zone and expressed as a
"growth factor." (The growth factor is simply the ratio of expected future
traffic to the existing traffic.)
2. Future traffic originating in (or destined to) a given zone is estimated by
multiplying growth factors and existing traffic.
3. This traffic is distributed to other zones in proportion to existing interzonal
travel and growth factors, for example,
(Trips)ij = (estimated future total) x
(present travel)ij (growth factor)j
all destination zones (growth factor) (present travel)
4. This distribution will yield two values (different) for each movement, for
example, Vij and V ji. Average these two values.
5. The sum of these average values for a particular zone probably will be
different from the existing traffic to (or from) that zone multiplied by its growth
factor (desired volume). Obtain new growth factors:
desired volume
New growth factors =
volume obtained from sum of movements
6. Make a second approximation using these growth factors.

7. Repeat the process until there is reasonable harmony between interzonal
traffic sums and desired volume.
Example 4.5
For zones 1 to 4 , the present traffic volumes and patterns and growth factors are
indicated below. Determine future traffic volumes and patterns by the Fratar
method.
To zone
1 2 3 4
1 - 10 12 18
2 10 - 14 14
From 3 12 14 - 6
zone 4 18 14 6 -
Zone
1 2 3 4
Present totals 40 38 32 38
Growth factors 2 3 1.5 1
Estimated future totals 80 114 48 38

Solution:
Find Vij
V12 F2
V121 80
V12 F2 V13 F3 V14 F4
10 3
V121 80 36.4
10 3 12 1,5 18 1
V21F1
V211 114
V21F1 V23 F3 V24 F4
10 2
V211 114 41.5
10 2 14 1,5 14 1
12 1,5
V131 80 21.8
10 3 12 1,5 18 1
12 2
V311 48 16
12 2 14 3 6 1
18 1
V141 80 21.8
10 3 12 1,5 18 1
18 2
V411 38 15.7
18 2 14 3 6 1,5
36.4 41.5
V12' 39
2
21 .8 16
V13' 18 .9
2
21 .8 15 .7
V14' 18 .8
2
Likewise: V23 = 35.7
V24 = 23.6
V34 = 4.0
F:nd New Growth Factors , F'i
Check Fi Vik= Vik

2 (10 + 12 + 18) = 39 + 18,9 + 18,8
1 2 3 4 Vik Fi Vik
1 - 39 18.9 18.8 76.7 80
2 39 - 35.7 23.6 98.3 114
3 18.9 35.7 - 4.0 28.6 48
4 18.8 23.6 4.0 - 46.4 38
Calculate again
Fi Vik
Fi '
Vik
2 40
F1' 1.04
76,7
114
F2' 1.16
98,3
48
F3' 0.82
58,6
38
F4' 0.82
46,4
It is iteration process, so repeat all calculations again (with new Fi') until Fin = Fin-1
4.4 The Gravity Model
The gravity model is one of the most widely used trip distribution techniques in
transportation planning.
The gravity model gets its name from the fact that it is conceptually based on
Newton's law of gravitation, which states that the force of attraction between two
bodies is directly proportional to the product of the masses of the two bodies and
inversely proportional to the square of the distance between them, or
m1 m 2
F k
d2
Early studies measured trip generation and attraction components in terms of zonal
populations, and the resistance function was assumed to be related to an inverse
function of distance.

This relationship was similar to Sir Isaac Newton's theory of gravity, expressed in
mathematical terms as
Pi Pj
Vij K
Dn
where
Vij = interaction between i and j
Pi = population at i
PJ = population at j
D = distance between i and j
K = some constant
n = some exponent
Gravity model formulations in current use are based on the hypothesis that the trips
produced at an origin and attracted to a destination are directly proportional to:
Total trip productions at the origin
Total trip attractions at the destination
A calibrating term
A socio-economic adjustment factor
The form of this relationship may be written as:
Vij C Pi A j Fij K ij
where
Vij = trips produced at i and attracted at j
C = a constant
Pi = total trip production at i
Aj = total trip attraction at j
Fij = a calibration term for interchange ij (friction factor)
Kij = a socio-economic adjustment factor for interchange j
i = an origin zone number
j = a destination zone number
n = number of zones
A value for C for any origin i (Ci) can be established when it is specified that the sum
of all Tij's for origin i must be equal to Pi. Therefore

n n
Pi Vij (C i Pi A j Fij K ij )
j 1 j 1
n
= C i Pi ( A j Fij K ij )
j 1
1
Ci n
(A F K
j 1
j ij ij )
and it becomes
Pi A j Fij K ij
Vij n
A F K
j 1
j ij ij
This equation is the standard form of the gravity model.

The term Fij is the calibrating term and generally is found to be an inverse
exponential function of impedance.
In developing the model, the output from this equation normally will show
production (row) totals to be correct but attraction (column) totals will not
necessarily match their desired values.
In order to match the desired values, an iterative procedure is employed.
Example 4.6
Consider the results of the trip generation step as shown in the Figure below, find the
trip interchanges between zone 3 and all other zones. Assuming that
F31 = 6, F32 = 29 , F33 = 45 , F34 = 18 and F35 = 4. All Kij = 1.
Solution:
Oi D j Fij K ij
Vij n
D F K
j 1
j ij ij
602 1080 6 1.0

V31 147 trips
1080 6 1 531 29 1 76 45 47 18 82 4
Likewise,
T32 = 350
T33 = 78
T34 = 19
T35 = 8
4.4.1 Gravity Model by Drew's Technique

In the gravity model, the no. of trips between zones is directly related to activities in
the area represented by trip generation and inversely related to the separation
between areas represented as a function of travel time.
The idea is: given Oi, Dj , tij , x find the elements of trip distribution matrix, Vij for the
horizon year subject to the following constraints:
1. V ij
= Oi
j
2. V
i
ij = Dj
3. O = D
i
i j
j
OD matrix
1 2 .. j .. n Oi
1 V11 V12 V1j V1n O1
2 V21 V2n O2
..
i Vij Oi
..
m Vm1 Vmn On
Dj D1 D2 Dj Dn
General form of the gravity model:
Vij = Kij Oi Dj Z(tij)

where
Z(tij) is a factor impending travel, often represented by
Z(tij) = (tij)-x
and x is a measure of the importance of proximity to the traveler
1/2 < x < 1 for work trips
2 < x < 3 for non-work trips
For this lecture, will be represented by ai and bj
where
ai , bj = set of parameters needed to fulfill the constraints
Vij = ai bj Oi Dj / tijx
where
Vij = no. of trips from zone i to zone j
tij = travel time from zone i to zone j
From the above equation => Need to find ai & bj
ai = 1/ b j D j / t ijx
j
x
bj = 1/ aiOi / tij
i
Computational Procedure
1) Assume all bj = 1
2) Calculate all ai(1) using
1
ai(1)
(bj Dj / tijx )
j
3) Find all Vij(1)

Vij(1) = ai(1)bj Oi Dj /tijx
4) Calculate bj(2) using
1
b (j 2)
(a Oi / tijx )
i
(1)
i
(2)
5) Calculate ai using
Vij(2) = ai(2)bj(2) Oi Dj /tijx
6) Find Vi(2)
7) Continue until

ain = ain-1
bin = bin-1
Vijn = Vijn-1
Example 4.7 Find the trip interchanges for the following trip distribution matrix using
the gravity model with x =1
1 2 Oi
2 18 13 35000
3 19 14 10000
4 30 25 35000
Di 40000 40000
Solution
Step 1:
1
= = 1.887 10
40000 40000
+
18 13
1
= = 2.015 10
40000 40000
+
19 14
1
= = 3.409 10
40000 40000
+
30 25
1
=
35000 10000 35000
1.887 10 + 2.015 10 + 3.409 10
18 13 30
= 1.148
1
=
35000 10000 35000
1.887 10 + 2.015 10 + 3.409 10
13 14 25
= 0.885
Step 2:
1
= = 1.895 10
40000 40000
1.148 + 0.885
18 13
1
= = 2.021 10
40000 40000
1.148 + 0.885
19 14
..
V21 = 1.89510-4 1.1483500040000/18 = 16931

V22 = 1.89510-4 0.8853500040000/13 = 18069
..
Zone 1 2
2 16931 18069
3 4887 5113
4 18182 16818

5. MODAL SPLIT
5.1 Overview
Modal split techniques are used by the planner to supply answers to the
following questions: "By what mode of transport will trips from this zone be made?"
In many transportation planning studies, a particular subgroup (referred to as
the transit- captive subgroup) has been singled out for special treatment.
As this group's name implies, it consists of people who for various reasons do
not have ready access to private transportation and, hence, whose mobility is almost
exclusively dependent on the public-transit system.
Included in this group are many of the elderly, the poor, the very young, and
even the second primary individual of one-car households.
Since transit captives do not have access to private transportation, they may
be identified on a zonal basis as a percentage of the trip generation.
The remaining trip-makers who do have a choice between competing private
and public modes are treated by the modal split model, which estimates the
percentage of trip-makers who choose to use the public-transportation system (i.e.,
the choice riders).
The total ridership of the public transportation system can be obtained by
summing the estimated captive and choice riders.
5.1.1 Factors Affecting Modal Split

Before we can predict how travel will be split among the modes available to
the travelers, we must analyze the factors that affect the choices that people make.
Three broad categories of factors are considered in mode usage:
1. The characteristics of the trip maker (e.g., family income, number of autos
available, family size, residential density, household structure,.. )
2. The characteristics of the trip (e.g., trip distance, time of day, trip purpose, ..)
3. The characteristics of the transportation system:
Quantitative factors are:
Relative travel time: in-vehicle, waiting and walking times by each
mode; time spent outside the vehicle, e.g., walking, parking,
transferring, etc.
Relative monetary costs (fares, fuel and direct costs);
Availability and cost of parking
Qualitative factors which are less easy to measure are:
Transportation Systems Planning (ECIV 540) | MODAL SPLIT 65

Comfort and convenience
Reliability and regularity
Protection, security
These are the categories of independent variables that would be included in
mathematical models of modal choice, the dependent variable being the market
share or the percent of travelers that are expected to use each of the available
modes.
5.1.2 Methods
1) Diversion Curve
2) Stratified Diversion-Curve Model
3) Logit Model
5.2 Types of modal split models
5.2.1 Trip-end modal split models

Traditionally, the objective of transportation planning was to forecast the growth in
demand for car trips so that investment could be planned to meet the demand.
When personal characteristics were thought to be the most important determinants
of mode choice, attempts were made to apply modal-split models immediately after
trip generation. Such a model is called trip-end modal split model. In this way
different characteristics of the person could be preserved and used to estimate
modal split. The modal split models of this time related the choice of mode only to
features like income, residential density and car ownership.
The advantage is that these models could be very accurate in the short run, if public
transport is available and there is little congestion. Limitation is that they are
insensitive to policy decisions example: Improving public transport, restricting parking
etc. would have no effect on modal split according to these trip-end models.
5.2.2 Trip-interchange modal split models

This is the post-distribution model; that is modal split is applied after the distribution
stage. This has the advantage that it is possible to include the characteristics of the
journey and that of the alternative modes available to undertake them. It is also
possible to include policy decisions. This is beneficial for long term modeling.
5.2.3 Aggregate and disaggregate models

Mode choice could be aggregate if they are based on zonal and inter-zonal
information. They can be called disaggregate if they are based on household or
individual data.

5.3 Diversion Curve
One of the simplest modal split models employs simple diversion curves as shown in
the Figure below:
Figure 5-1 Diversion Curve for Modal Split
To find the modal split between auto and transit, one needs to find the travel time
ratio between these two modes.
5.4 The Stratified Diversion-Curve Model
Figure A illustrates a typical stratified diversion-curve model of the modal choice,

which superposes the calibration results obtained in Washington, D.C., Philadelphia,
and Toronto for work trips and two competing modes (i.e., auto and transit).
The independent variables employed by this model include three characteristics of
the competing modes - the ratio of transit to auto travel time (TTR), the ratio of
transit to auto costs (CR), and the ratio of transit to auto service'tL), which is defined
as the time spent outside the vehicle - and one characteristic of the trip makers,
income category (EC).
Each cell of the table corresponds to an income range and a cost-ratio level and
contains a family of regression curves relating the travel-time ratio to the percentage
of trips choosing transit for each of the four levels of relative service.
Relative travel time, TTR
TTR represents the ratio of the door-to-door travel time by transit to that by
automobile:
Transit travel time
TTR =
Auto travel time
if
x1 = time spent in the transit vehicle
x2 = transfer time between transit vehicle
x3 = time spent waiting for a transit vehicle
x4 = walking time to transit vehicle
x5 = walking time from transit vehicle
X6 = auto driving time
x7 = parking delay at destination
x8 = walking time from parking place to destination,
x1 x 2 x 3 x 4 x 5
then TTR
x 6 x 7 x8
Relative travel cost, CR
Transit fare
CR =
Out of pocket Auto travel cost
x9
CR
( x10 x11 0,5 x12 ) / x13
if
x9 = transit fare
x10 = cost of gasoline
x11 = cost of oil change and lubrication
x12 = parking cost at destination
x13 = average car occupancy
Economic status of the trip maker, EC
Income, for example:
EC1: 0$ - 3,100$

EC2: 3,100$ - 4,700$
EC3: 4,700$ - 6,200$
EC4: 6,200$ - 7,500$
EC5: 7,500$ ++
Relative travel service, L
"L" is difficult to quantify because it pertains to such intangible factors as ... the
atmosphere within the vehicle, comfort, appearance, ride smoothness, availability of
seats, and convenience of transfer.
L is designated somewhat arbitrarily as
time spent outside transit vehicle

L =
auto parking + walking time
x 2 x3 x 4 x5
L
x7 x8
Example 5.1
The peak-hour target-year work-trip Interchange volume between two zones has
been estimated to be 2000 person-trips as follows: 200 of these trips are made by
captive riders, 600 are made by trip-makers from households of income category EC3,
and the remaining 1200 from economic category EC5. Apply the stratified model of
Fig. A to find the market share of the transit system assuming that this model has
been calibrated as a disaggregate model based on household-income categories. The
following data apply to the interchange under consideration:
TTR = 2.0 CR = 0.8 L = L4
Solution:
For a disaggregate model, the contribution of each economic subgroup to the transit
patronage must be computed separately and then summed to arrive at an estimate
of the total transit share:
Captives: 200 x 1.0 = 200 person-trips
EC3 group: 600 x 0.25 = 150
EC5 group: 1200 x 0.2 = 240

5.5 Probabilistic Models.... Logit Model
Although simple in concept, the stratified diversion model described previously is

awkward to calibrate and use, especially if more than two competing travel modes
are to be included.
More computationally efficient probability-based models of modal choice have been
developed including probit analysis models, and the most popular logit analysis
models.
These models of human choice have been applied to many situations to explain how
people select between competing alternatives.
Each alternative is described by a utility (or disutility) function, and the probability
associated with an individual's choosing of each of the competing alternatives is
expressed mathematically in terms of these utilities.
Extended to groups of individuals via the theory of probability, these models estimate
the proportion of the group that is likely to choose each of the competing
alternatives.
The development of each model involves two steps: the selection of its mathematical
form and the calibration of appropriate utility functions that render the selected
model capable of reproducing the available base-year data.
5.5.1 Utility and Disutility Functions

A utility function measures the degree of satisfaction that people derive from their
choices.
A disutility function represents the generalized cost that is associated with each
choice.
The utility (or disutility) function is typically expressed as the linear weighted sum of
the independent variables or their transformation, that is,
U = a0 + a1X1 + a2X2 + . + a1X1
where U is the utility derived from a choice defined by the magnitudes of the
attributes X that are present in that choice and that are weighted by the model
parameters a.
5.5.2 Binary logit model

Binary logit model is the simplest form of mode choice, where the travel choice
between two modes is made. The traveler will associate some value for the utility of
each mode. If the utility of one mode is higher than the other, then that mode is
chosen. But in transportation, we have disutility also. The disutility here is the travel
cost. This can be represented as
cijm = a1 tijv,m + a2 tijw,m + a3 tijt,m + a4 Fijm+ a5 jm +

where tijv,m is the in-vehicle travel time between i and j of mode m, tijw,m is the
walking time to and from stops, tijt,m is the waiting time at stops, Fijm is the fare
charged to travel between i and j, jm is the parking cost, and is a parameter
representing comfort and convenience. If the travel cost is low, then that mode has
more probability of being chosen. Let there be two modes (m=1,2) then the
proportion of trips by mode 1 from zone to zone is (Pij1). Let cij1 be the cost of
traveling from zone i to zone j using the mode 1, and cij2 be the cost of traveling from
zone i to zone j by mode 2, there are three cases:
if cij2 - cij1 is positive, then mode 1 is chosen.
if cij2 - cij1 is negative, then mode 2 is chosen.
if cij2 - cij1 = 0 , then both modes have equal probability.
This relationship is normally expressed by a logit curve as shown in Figure 5-2.
Therefore the proportion of trips by mode 1 is given by:
This functional form is called logit, where cij is called the generalized cost and is the
parameter for calibration. The graph in Figure 5-2 shows the proportion of trips by
mode 1 (Tij1/ Tij ) against cost difference.
Figure 5-2 Logit function
Example 5.2
Let the number of trips from zone i to zone j is 5000, and two modes are available
which has the characteristics given in Table 5-1. Compute the trips made by mode
bus, and the fare that is collected from the mode bus. If the fare of the bus is reduced
to 6, then find the fare collected.
Table 5-1 Trip characteristics
tijv,m tijw,m tijt,m Fijm jm

Car 20 - 18 4
Bus 30 5 3 9
ai 0.03 0.04 0.06 0.1 0.1
Table 5-2 Binary logit model example: Solution
tijv,m tijw,m tijt,m Fijm jm cijm Pijm Tijm

Car 20 - 18 4 2.08 .52 2600
Bus 30 5 3 9 2.18 .475 2400
ai 0.03 0.04 0.06 0.1 0.1
Solution
The base case is given below.
Cost of travel by car (Equation)= cijcar = 0.03 20 + 18 0.06 + 4 0.1 = 2.08
Cost of travel by bus (Equation)= cijbus = 0.03 30 + 5 0.04 + 3 0.06 + 9 0.1= 2.18
Probability of choosing mode car (Equation)= 0.52
Probability of choosing mode bus (Equation)= 0.475
Proportion of trips by car = 50000.52 = 2600
Proportion of trips by bus = 50000.475 = 2400
Fare collected from bus = Tijbus Fij = 24009 = 21600
When the fare of bus gets reduced to 6,
Cost function for bus = 0.03 30 + 5 0.04 + 3 0.06 + 6 0.1= 1.88
Probability of choosing mode bus (Equation)= 0.55
Proportion of trips by bus = 5000 0.55 = 2750
Fare collected from the bus = 2750 6 = 16500
5.5.3 Multinomial Logit Model

The multinomial logit model calculates the proportion of travelers that will select a
specific mode K according to the following relationship:
eUk
Pk
eUk
Uk = utility of mode k

Example 5.3
A calibrated utility function for travel in a medium-sized city by automobile, bus, and
light rail is
U = a - 0.002X1 - 0.05X2
where Xi is the cost of travel (cents), and X2 is the travel time (min). Calculate the
modal split for the given values.
Mode a x1 x2
Automobile - 0.30 130 25
Bus. - 0.35 75 35
Light rail - 0.40 90 40
Solution:
Automobile: Ua = - 0.30 - 0.002(130) - 0.05(25) = - 1.81
Bus: Ub = - 0.35 - 0.002 (75) - 0.05(35) = - 2.25
Light rail: Ur: =- 0.40-0.002(90) -0.05(40) =- 2.58
P(a) = e-1.81/ (e-1.81+ e-2.25+ e-2.58) = 0,475
P(b) = 0,304
P(r) = 0,221
Modal Split = Auto of 47.5 %
Bus of 30.4 %
Rail of 22.1 %
5.6 Behavioral Model
A behavioral model is based on a representation of individual choice when faced with

alternatives.
If Pijm = fraction of travelers with choice going from i to j using mode m.
then Pijm = f(Zijm)
Zijm = generalized cost of going from i to j using mode m
= Cijm + v(tijm )
Cijm = money spent from i to j using m
m
tij = time spent from i to j using m
v = value of trip maker's time

Choose a mathematical form of Pijm = f(Zijm) that is realistic
m
exp ( zijm )
P
ij
exp ( zijm )
Computational procedure illustrated (use v = $ 6/hr)
m Cijm Tijm Zijm exp (-Zijm) Pijm VijChoice Vijm
B 0.25 20 2.25 0.1054 0.279 3347
C 0.5 8 1.30 0.2725 0.5 12,000 8653
0.3779 12,000
Determination of value of trip maker's time

We know that the major determinants of modal choice are
1) Relative travel cost, CijC - CijB
2) Relative travel time, tijB - tijC
In order to place these two determinants in common terms, we multiply
(tijB - tijC) by "v", the value of time
Assumptions are :
if CijC - CijB < v(tijB - tijC), rational driver will take car
if CijC - CijB > v(tijB - tijC), rational driver will take bus
Value of Trip Makers' Time
Example 5.4
Find the value of traveler's time from the following commuter

Car Travel time (Min.) Travel cost ($) Mode
Owner Chosen
Car Bus Car Bus
1 6 18 0.45 0.1 Car
2 15 25 0.80 0.1 Bus
3 40 55 1.0 0.2 Bus
4 11 21 0.6 0.1 Bus
5 21 36 0.6 0.2 Car
6 25 37 0.85 0.2 Bus
7 18 33 0.80 0.2 Car
8 60 80 1.30 0.1 Bus
9 8 18 0.55 0.1 Car
10 15 35 0.95 0.2 Car
Solution:
Use Car Use Bus
CijC - CijB tijB - tijC CijC - CijB tijB - tijC
0.35 12 0.7 10
0.4 15 0.8 15
0.6 15 0.5 10
0.45 10 0.65 12
0.75 20 1.2 20
$ 2.55 < 72v $3.85 >72v
$ 2.13/hr < v < $ 3.44/hr

6. TRAFFIC ASSIGNMENT
6.1 Overview
The final step of the sequential transportation-forecasting process is the assignment

of zonal trip interchanges to the individual transportation facilities.
The basic resulting output is in the form of traffic volumes on each portion of the
transportation system.
As a result, the planner can determine for the base year how well the assignment
process has simulated observed traffic volumes.
When the traffic assignment is used with future trip interchanges, the model
indicates how well the proposed facilities will serve the anticipated travel demand.
6.2 Link cost function
As the flow increases towards the capacity of the stream, the average stream speed
reduces from the free flow speed to the speed corresponding to the maximum flow.
This can be seen in the graph shown below (Figure 6-1).
That means traffic conditions worsen and congestion starts developing. The inter-
zonal flows are assigned to the minimum paths computed on the basis of free-flow
link impedances (usually travel time). But if the link flows were at the levels dictated
by the assignment, the link speeds would be lower and the link travel time would be
higher than those corresponding to the free flow conditions. So the minimum path
computed prior to the trip assignment will not be the minimum after the trips are
assigned.
Travel time
Volume (x)
Figure 6-1 Two Link Problem with constant travel time function
A number of iterative procedures are done to converge this difference. The relation
between the link flow and link impedance is called the link cost function and is given
by the equation as shown below:
Transportation Systems Planning (ECIV 540) | TRAFFIC ASSIGNMENT 80

= 1+
where t and x is the travel time and flow, respectively on the link, to is the free flow
travel time, and k is the practical capacity. The parameters and are specific the
type of link and is to be calibrated from the field data. In the absence of any field
data, following values could be assumed: = 0.15, and = 4.0.
The types of traffic assignment models are all-or-nothing assignment (AON),
incremental assignment, capacity restraint assignment, user equilibrium assignment
(UE), stochastic user equilibrium assignment (SUE), system optimum assignment (SO),
etc. The frequently used models all-or-nothing, user equilibrium, and system
optimum will be discussed in detail here.
6.3 All-or-nothing traffic assignment
Assignment using the all-or-nothing method is a simple process.

The procedure is assign all trips (or volumes) to the Shortest Time route and nothing
to other routes
Mathematically 1 to the shortest route
Pijmr =
0 to all other routes
mr
where Pij = fraction of trips (or volumes) from i to j using m taking route r.
Example 6.1 Assign the trip interchanges shown in the Table below to the network of
the four zones area as shown in the Figure below. The time shown in the Figure is the
travel time along the links, expressed in minutes.
To Zone
1 2 3 4
1 0 500 750 350
From 2 275 0 1050 475
Zone 3 650 1870 0 950
4 1250 350 2050 0
100
1 2
110
170
70 80 70 50
165
60
4 3
100

Solution:
110
1 2
70
From zone 1
170
4 3
100
1 2
70
130
From zone 2 4 3
1 2
From zone 3 140 50

60
4 3
1 2
From zone 4 80 150

100
4 3
Assign the trips:
500
1 2
750 + 350
From zone 1
750
4 3

275
1 2
From zone 2
1050 + 475
475
4 3
1 2
From zone 3 650 1870
4 3
950 + 650
1 2
1250 350
From zone 4
4 3
2050 + 350
275
1 2
500
0
1100 1900 1525 2220
0
2075
4 3
3150
6.4 Diversion
Diversion refers to the allocation of a trip interchange to two possible routes in a

designated proportion that depends on some specified criterion.
One of the more widely used diversion techniques in early studies was that
developed by Campbell.
This curve (Figure 6-2) is based upon relative travel times between the new facility
and the quickest alternative route.

Time via arterial highway/Time via quicker alternate route
Figure 6-2 Time ratio diversion curves
The state of California method developed later incorporated both time and distance
savings in the diversion curves, shown in Figure 6-3.
t = Time saved via freeway route, (min)
Figure 6-3 California time and distance savings diversion curves
The California curves consist of a family of hyperbolas.

It can be seen that with equal time and distance on the freeway and on the best
alternative arterial route, 50 percent of the trips are assigned to the free way using
these curves.
This relationship can also be presented mathematically as
P = 50 + 50 (d + 1/2t) [(d - 1/2t)2 + 4.5] -1/2
where: P = percent of trips via freeway
d = distance saved in miles via freeway
t = time saved in minutes via freeway
The most widely used diversion procedure in the 1960s was that available in the
series of traffic planning computer programs of the Bureau of Public Roads (now
called the Federal Highway Administration).
This form of diversion is dependent on one parameter only, the ratio of travel times
by the quickest combined arterial-freeway route to the quickest arterial-only route.
With a one-parameter relationship, one single diversion curve (see Figure 6-4) defines
the relationship.
Travel time ratio
Figure 6-4 FHWA diversion curve
6.5 User equilibrium assignment (UE)
The user equilibrium assignment is based on Wardrop's first principle, which states
that no driver can unilaterally reduce his/her travel costs by shifting to another route.
User Equilibrium (UE) conditions can be written for a given O-D pair as:
fk (ck u) = 0, k
ck u 0, k
where fk is the flow on path k, ck is the travel cost on path k, and u is the minimum
cost. Last equation can have two states.
If ck u = 0, so that fk > 0. This means that all used paths will have same travel time.
If ck u > 0, then from equation 2: fk = 0.
This means that all unused paths will have travel time greater than the minimum cost
path.
Assumptions in User Equilibrium Assignment
The user has perfect knowledge of the path cost.
Travel time on a given link is a function of the flow on that link only.
Travel time functions are positive and increasing.
The solution to the above equilibrium conditions given by the solution of an
equivalent nonlinear mathematical optimization program,
Minimize: = ( ) (4)
Subject to 0
Where, xa equilibrium flows in link a, ta travel time on link a.
The equations above are simply flow conservation equations and non negativity
constraints, respectively. These constraints naturally hold the point that minimizes
the objective function. These equations state user equilibrium principle. The path
connecting O-D pair can be divided into two categories: those carrying the flow and
those not carrying the flow on which the travel time is greater than (or equal to) the
minimum O-D travel time. If the flow pattern satisfies these equations no motorist
can better off by unilaterally changing routes. All other routes have either equal or
heavy travel times. The user equilibrium criteria are thus met for every O-D pair. The
UE problem is convex because the link travel time functions are monotonically
increasing function, and the link travel time a particular link is independent of the
flow and other links of the networks. To solve such convex problem Frank Wolfe
algorithm is useful.
Example 6.2
Let us suppose a case where travel time is a function of flow as shown below.
Figure 6-5 Two link problem with travel time functions
Solution
Substituting the travel time in equation yield to
subject to x1 + x2 = 12.
Substituting x2 = 12 - x1 in the above formulation will yield the unconstrained
formulation as below :
Differentiate the above equation and equate to zero, and solving for x1 and then x2
leads to the solution x1= 5.8, x2= 6.2.
Example 6.3
Two routes connect a city and a suburb. During the peak hour commute, a total of
4500 vehicles travel from the suburb to the city. Route 1 has a 60 mi/h speed limit
and is six miles in length; route 2 is three miles in length with a 45 mi/h speed limit.
Studies show that the total travel time on route 1 increases two minutes for every
additional 500 vehicles added. Minutes of travel time on route 2 increase with the
square of the number of vehicles, expressed in thousands of vehicles per hour.
Determine user-equilibrium travel times.
Solution:
Determine free-flow travel times, in minutes, gives
Route 1: 6 mi/ (60 mi/h) x 60 min/h = 6 min;
Route 2: 3 mi/ (45 mi/h) x 60 min/h = 4 min;
With these data, functions can be written as:
t1 = 6 + 4x1
t2 = 6 + x 2 2
where, t1, t2 : Average travel times on routes 1 and 2 in minutes;
x1, x2 : Traffic flow on routes 1 and 2 in thousands of vehicles per hour.
Also, q = x1 + x2 = 4.5
where, q = total traffic flow between the origin and destination.
Because t1(4.5) > t2(0) and t2(4.5) > t1(0) so that both routes are used. If t1(0) > t2(4.5),
4500 vehicles use only route 2.
With both routes are used, Wardrop's user equilibrium definition gives:
t1 = t 2
or 6 + 4x1 = 6 + x22
--> x2 = 2899 veh/h; x1 = 1601 veh/h.
Another way,
.
Min S(x) = (6 + 4 ) + (4 + )
= 27 - 6x2 + 40.5 - 18x2 + 2x22 + 4x2 + x23/3

Differentiate the above equation and equate to zero, and solving for and then leads
to the solution x2= 2899, x1= 1601.
6.6 System Optimum Assignment (SO)
The system optimum assignment is based on Wardrop's second principle, which

states that drivers cooperate with one another in order to minimize total system
travel time. This assignment can be thought of as a model in which congestion is
minimized when drivers are told which routes to use. Obviously, this is not a
behaviorally realistic model, but it can be useful to transport planners and engineers,
trying to manage the traffic to minimize travel costs and therefore achieve an
optimum social equilibrium.
subject to 0; q =
Example 6.4
To demonstrate how the assignment works, an example network is considered. This

network has two nodes having two paths as links. Suppose that travel time is a
function of flow as shown in the figure below.
Figure 6-6 Two Link Problem with time functions
Solution
Substituting the travel time in equation , we get the following:
min Z(x) = x1 * (10 + 3x2) + x2 * (15+2x2)
= 10x1 + 3x12 + 15x2 +2xx2
Substituting x2 = x1 -12,
min Z(x) = 10x1 + 3x12 + 15(x1 -12) +2(x1 -12)2
Differentiate the above equation to zero, and solving for x1 and then x2 leads to the
solution x1= 5.3, x2 = 6.7
which gives Z(x) = 327.55
6.7 Incremental assignment
Incremental assignment is a process in which fractions of traffic volumes are assigned

in steps. In each step, a fixed proportion of total demand is assigned, based on all-or-
nothing assignment. After each step, link travel times are recalculated based on link
volumes. When there are many increments used, the flows may resemble an
equilibrium assignment; however, this method does not yield an equilibrium solution.
Consequently, there will be inconsistencies between link volumes and travel times
that can lead to errors in evaluation measures. Also, incremental assignment is
influenced by the order in which volumes for O-D pairs are assigned, raising the
possibility of additional bias in results.
6.8 Stochastic user equilibrium assignment
User equilibrium assignment procedures based on Wardrop's principle assume that

all drivers perceive costs in an identical manner. A solution to assignment problem on
this basis is an assignment such that no driver can reduce his journey cost by
unilaterally changing route. Van Vilet considered as stochastic assignment models, all
those models which explicitly allows non minimum cost routes to be selected.
Virtually all such models assume that drivers perception of costs on any given route
are not identical and that the trips between each O-D pair are divided among the
routes with the most cheapest route attracting most trips. They have important
advantage over other models because they load many routes between individual
pairs of network nodes in a single pass through the tree building process, the
assignments are more stable and less sensitive to slight variations in network
definitions or link costs to be independent of flows and are thus most appropriate for
use in uncongested traffic conditions such as in off peak periods or lightly trafficked
rural areas.
6.9 Dynamic Assignment
Dynamic user equilibrium, expressed as an extension of Wardrop's user equilibrium

principle, may be defined as the state of equilibrium which arises when no driver can
reduce his disutility of travel by choosing a new route or departure time, where
disutility includes, schedule delay in addition in to costs generally considered.
Dynamic stochastic equilibrium may be similarly defined in terms of perceived utility
of travel. The existence of such equilibrium in complex networks has not been proven
theoretical and even if they exist the question of uniqueness remains open.

6.10 Capacity-Restraint Techniques
Many different capacity restraint equations have been developed and tested and are
available for use, which use the volume-capacity ratio.
The underlying premise of a capacity restraint model is that the travel time on any
link is related to traffic volume on that link.
The Bureau of Public Roads (BPR) Method
T = T0 [1 + 0,15 (V/C)4]
where T = travel time at traffic flow V
T0 = free flow travel time ( "zero-flow" travel Time )
text books recommended = travel time at practical capacity x 0.87
V = assigned volume
C = practical capacity
Drew's technique
An iterative procedure that uses a relationship between travel time ratio and volume
to capacity ratio as:
t ijmr
mr
f (Vijmr / Qijmr )
t 0
where V ijmr = volume of trips from i to j using m over r

Q ijmr = capacity, maximum volume, veh/hr.
tijmr = travel time from i to j using m over r for a volume Vijmr
t0mr = Free flow travel time from i to j using m over r
Form of the capacity restraint:
t mr
t mr
1 (V mr
ij
/ Qijmr ) (1 k r )
ij 0
1 Vijmr / Qijmr
where kr = level of service factor which varies for different types of highway facility
0 < kr < 1/2 where f = freeway or expressway
1/2 < kr < 1 where a = arterial routes
(kr = 0 for a "perfect" highway facility)
Example 6.5
For the design year, zones 1 and 2 are to be joined by an expressway with 3 lanes in
each direction with a capacity of 2000 vph per lane, an existing 2 lanes/direction
arterial street with a capacity of 1200 vph/lane. The free flow travel time over the
expressway and arterial street are 10 min and 20 min, respectively. Their level of

service factors are 1/4 and 3/4, respectively. Find the traffic volumes on the
expressway and arterial street if analysis showed that 6,600 vph can be expected to
travel between zones 1 and 2 in the design year.
Solution:
Assume V 12cf = 5400 vph
V 12ca = 1200 vph
5400 1
1 (1 )
= 10 3 2000 4 = 32 min.
5400
1
6000
1200 3
1 (1 )
= 20 2 1200 4 = 35 min.
1200
1
2400
Since t 12cf t 12ca : Must try again
6.11 Summary
Traffic assignment is the last stage of traffic demand modeling. There are different
types of traffic assignment models. All-or-nothing, User-equilibrium, and System-
optimum assignment models are the commonly used models. All-or-nothing model is
an unrealistic model since only one path between every O-D pair is utilised and they
can give satisfactory results only when the network is least congested. User-
equilibrium assignment is based on Wardrop's first principle and it's conditions are
based on certain assumptions. Wardrop's second principle is utilized by System-
optimum method and it tries to minimise the congestion by giving prior information
to drivers regarding the respective routes to be chosen. Other assignment models are
also briefly explained.

7. URBAN PUBLIC TRANSPORTATION
7.1 Transit planning
Transit planning takes many forms. It includes strategic planning, which takes a broad
global look at how an agency might function in its surrounding environment; long-
range planning, which generally relates to major facility development; short-range
planning, which traditionally produces a transit development plan; and service or
operations planning, which looks at service changes on a continuing basis.
All types of planning are closely linked to system finance and administration.
The transit planning effort should reflect specific local circumstances and needs.
It should assess existing problems and how they are likely to change, identify
improvement options, and suggest directions.
It should provide essential information to decision makers relative to ridership, costs,
performance, and environmental and economic aspects.
It should produce transit plans that are compatible with an area's needs, goals, and
resources.
Short and long-range transit planning studies traditionally include
setting goals and objectives;
conducting system inventories;
surveying travel patterns, especially those of transit patrons;
identifying existing needs and problems;
forecasting future ridership;
identifying and assessing improvement alternatives;
developing improvement programs; and
preparing financial analyses.
It may lead to environmental impact assessments of proposed system developments.
The system planning process involves planners, transportation engineers, citizens,
and public officials.
Collectively, these groups should identify the most promising candidate plans for
decision makers to examine and from which to select the most appropriate
improvements.
7.2 Definitions of transit service
Definitions of urban public transportation can be defined according to general

function, technology, and type of service as follows:
Transportation Systems Planning (ECIV 540) | URBAN PUBLIC TRANSPORTATION 92

7.2.1 General function definitions
Urban public transportation systems for intra-urban or intraregional travel
include transit and paratransit.
Urban transit (public transit or mass transit) is a common carrier passenger
transportation service that is available to any person who pays a prescribed
fare. It operates on established schedules along designated routes with
specific stops. Examples include bus, light rail, and rapid transit.
Paratransit service may not follow fixed routes or schedules, may not be
available to the general public, or both. Examples include taxis, vanpools, club
buses, and demand- responsive services.
Charter service is based on contracts between a transit agency and individuals
for occasional or regular hire of vehicles and drivers, is restricted to the use of
the contracting party, and follows whatever route and schedule this party
desires.
7.2.2 Technological definitions
Right-of-way refers to the land used by transit systems. It may be shared
(street transit), semi-exclusive (light rail transit or buses and high-occupancy
vehicles on reserved lanes), or exclusive (rapid transit).
Guideway or infrastructure refers to the right-of-way plus the special
improvements required for operations (e.g. tracks, power distribution, control
system).
Revenue vehicles are the rolling stock operating on highways or guideways
that furnishes the actual passenger transportation. The term transit unit
designates an individually scheduled unit, which may be a single vehicle or a
train of vehicles.
The following vehicles are commonly used:
7.2.2.1 Road-based vehicles

a. Articulated bus: A transit bus (diesel or electric) with a permanently
attached semitrailer, with full interior passenger circulation.
b. Double-decker bus: A bus with two levels.
c. Dual-powered bus: A transit bus that can be propelled by electricity or
by a diesel engine.
d. Transit bus: A vehicle usually propelled by a diesel engine that is
generally designed for frequent-stop service with front and center
doors, and without luggage storage compartments or restroom
facilities. A transit bus less than 25 feet long is called a minibus.
e. Trolleybus: A transit bus propelled by electricity obtained from
overhead wires.

f. Van: A passenger vehicle or an automobile or light truck chassis,
propelled by an internal combustion engine, with a capacity of 8 to 15
persons.
Figure 7-1 Articulated bus
Figure 7-2 Double-decker bus

Figure 7-3 Trolleybus
7.2.2.2 Rail and fixed guideway vehicles

a. Automated guideway vehicle (AGV): Small and medium-sized vehicles
that operate fully automatically on guideways with exclusive rights-of-
way on a loop or as a shuttle within central business districts, airports,
or other high activity centers.
b. Commuter railroad car: A standard railroad passenger car with high-
density seating. It may be self-propelled (by electricity or diesel engines)
or designated for haulage by locomotive.
c. Monorail: An electrically powered transit vehicle suspended from or
supported by a guideway formed by a single beam or rail.
d. Rail transit car (rapid transit car, heavy rail transit car): An electrically
propelled vehicle usually operated in trains on exclusive right-of-way.
e. Street car or light rail vehicle (LRV): An electrically propelled rail vehicle
operated singly or in trains on shared, semi-exclusive, or exclusive right-
of-way.
7.2.2.3 Other vehicles

Ferry boat: A passenger-carrying marine vessel that provides service over a fixed
route and on a published schedule between two or more points.
Figure 7-4 AGV, Powered by electric motors, the computer-driven cars
Figure 7-5 Monorail in Disneyland
7.2.3 Service definitions

Local transit service operates on streets or other rights-of-way with frequent
stops and at relatively low speeds; it serves adjacent land uses within
acceptable walking distances.
Express service does not attempt to serve all land areas through which it
passes, but offers faster speeds to a selected number of stops spaced more
widely apart. It includes limited stop services.
Basic service comprises routes that operate all day (although the length of the
"day" may vary from about 14 to 24 hours) and at least 5 days per week.
Peak service comprises routes that operate during peak demand periods only.
Special service comprises irregular routes operated for special events or for
seasonal traffic generators.
Short-haul transit refers to service within major activity centers over short
routes. Special technology is sometimes used.
7.3 Bus Transit Systems
Most bus service operates in mixed traffic over streets and highways.
These include motor buses and trolley buses operating on public streets.
The local transit authority usually prescribes the routes, frequencies, fares, and stops.
A uniform flat fare or one based on zones or distances is charged.
The vehicles can carry anywhere from 12 to 240 passengers and a mixture of standing
and seated passengers can be accommodated.
Local services may entail frequent stops, whereas express service may require only a
few stops.
Service standards are perceived in terms of reliability, frequency, journey time, and
quality of ride.
Bus transit enjoys the advantage of being quite flexible in meeting changes in
demand, with virtually no cost.
Extensions, expansions, and new routes can be introduced with little effort and low
cost.
Trolley bus systems, however, do not have this same flexibility because they are
constrained by overhead electric transmission lines.
Transit systems using buses with a seating capacity of 40 to 120 are capable of
carrying from 2400 to 15,000 passengers in mixed traffic.
Journey speed (including stops) in mixed traffic is likely to be in the range 12 to 25
km/hr; where several lanes are available in the same street, one can expect volumes
of between 25,000 and 30,000 bus passengers per hour in one direction.
Journey speed and capacity can be enhanced by utilizing reserved bus lanes.
Maximum bus transit performance can be provided by exclusive busways in which
buses are physically separated from other traffic by medians or barriers, with grade
separation or priority at intersections.
Volumes in excess of 30,000 passengers per hour per lane with speeds up to 30
km/hr can be achieved with off-line stations and multiple boarding platforms.

7.3.1 Priority treatment of buses
Bus priority lanes are used on downtown streets in many large cities (i.e., New York,
Paris, Singapore), and bus (or HOV) lanes are provided along radial freeways in many
cities (i.e., Washington, D.C., Los Angeles, San Diego, Seattle, Houston).
Busways (sometimes called transitways) are found in Adelaide (Australia), Pittsburgh,
Ottawa, Houston, and Runcorn (England).
The speed, capacity, and reliability of buses can be enhanced by a variety of
techniques, such as
1. Exclusive or preferential transit lanes on sections of streets and freeways
2. Exclusive transit turns at intersections
3. Metered freeway entry with bus preference
4. Passive traffic signal priority measures, such as cycle-length adjustments, split
phases, and timing plans selectively favoring buses
5. Active traffic signal priority measures, such as unconditional or conditional
signal preemption by buses
6. Exclusive transit streets, malls, and ramps
7. Exclusive busways
Frequently, the application of the first three techniques is limited to peak periods
and/or directions where transit patronage is relatively high.
Other high-occupancy vehicles (HOVs) are sometimes permitted to share use of
preferential transit lanes, ramps, and bypasses on freeways.
If street and highway space is considered for the exclusive use of buses, one test of
reasonableness is that the combined auto and transit passenger volumes in the
direction(s) and for the period(s) involved will not be reduced substantially below
existing levels, and preferably will be increased.
Another criterion sometimes considered is that the total travel or delay times for all
persons involved are reduced.
Bus priority treatments should result in a net reduction in total person delay.
Criteria for bus lanes: most of the warrants are qualitative and subjective
Quantitative
where qB = hourly volumes of buses

qA = hourly volumes of autos (including trucks as passenger car equivalents)
N = total no. of lanes per direction

X = ratio of average auto to bus occupancies
7.4 Para Transit
The term applies to small passenger vehicles operating informally on a fare-paying

basis, and serving in some places as an alternative to regular bus transit services.
Of the many classifications possible, the one that helps to capture the essence of
paratransit alternatives is shown in Figure 7-6 differentiated by time and space
characteristics.
TIME
Fix Variable
Rail Rapid Transit

Fixed Light rail Jitney
Fixed-route bus
SPACE
Road Deviation Shared ride

Point Deviation Auto transit
Variable Vanpool Taxi

Subscription bus Dial a ride
Carpool Auto rental
Figure 7-6 Temporal Spatial Characteristics of Urban Transport Alternatives
Although conventional transit operates on a fixed-schedule (time) and fixed-route

(space) basis (quadrant 1), paratransit operates in the environment of fixed space-
variable time (jitney), or fixed time-variable space (vanpool/carpool), or variable
time-variable space (taxi, dial-a-ride).
Paratransit systems are capable of offering
1. personalized door-to-door service,
2. shared services with routes determined by individual passengers, or
3. regular services along fairly well-defined routes, similar in some
respects to bus transit.

Figure 7-7 Jitney in Atlantic city
7.5 Light Rail Transit Systems
A light rail transit (LRT) is a generic term embracing a wide range of electrically
powered vehicles running on steel rails.
Light rail transit systems are found throughout the world, especially in Western
Europe and North America.
At one extreme are streetcars operating on tracks sharing the roadway with cars and
buses: the other could include LRT metros operating on exclusive rights-of-way.
Passengers usually board from the road surface or from low platforms.
Streetcars (SCRs) or trams usually run on fixed rails flush with the roadway streets in
mixed traffic.
The vehicles carry about 100 to 300 sitting and standing passengers and are
comparatively simple to operate and maintain.
Routing is constrained by the alignment, and rerouting is expensive.
SCRs operating in mixed traffic can carry from 4000 to 15,000 passengers per track
per hour at journey speeds of about 12 to 20 km/hr.
LRT systems operate along streets, but may be provided with exclusive rights of way
over all or part of their routes.
Grade-separated and priority signalized routes at intersections are common.
The system operates with trains of one to four cars.
A typical LRT with two cars has a capacity of 500 passengers.
LRT is intended to provide high-capacity service, fast and frequent.
Technologically, LRTs are superior to street-cars where signaling and control is
concerned.
Where LRT system operate on exclusive rights-of-way, capacity can be as high as
20,000 passengers per hour per track at speeds of between 15 and 40 km/hr.
Because of these facts, LRT has been adopted by a number of metropolitan areas
across the world.
Figure 7-8 Light rail transit in Toronto
7.6 Rapid Rail Transit Systems
Rapid rail transit (RRT) systems, called MRT, MTR, metros, the underground, or the
tube, operate on exclusive rights-of-way and at relatively high speeds and thus
provide the highest line capacity available.
Rapid transit is the main peak-hour travel mode for CBD travelers in most large cities.
Its high capacity and ability to operate below ground makes it possible to support a
large CBD employment; conversely ridership correlates closely with office
employment in the city center.
Speed, service reliability, and the ability to alleviate street congestion and impact
land development are among its desirable attributes.
Elevated and underground rights-of-way are common.
Flat fares, zone fares, or distance-based fares are collected through automatic or
other ticketing systems.
Four to 10 cars per train is not uncommon.
A typical train with, say, six cars may have a capacity of 1500 passengers, seated and
standing.
Operating at headways of 2 minutes and with a speed of 100 km/hr, the line capacity
can be 70,000 passengers per hour per line.
RRTs usually require sophisticated signaling and control devices to maintain high
speeds and frequencies with very high safety standards.
Rapid loading and unloading of passengers is achieved by providing high-level
platforms.
Ventilation systems and escalators are needed for underground operations.
The costs of construction, maintenance, and operation are enormous and changes in
routing are almost impossible.
To maximize patronage, it is generally necessary to supplement RRTs with feeder
systems such as buses and LRTs and other flexible modes.
The reliability of RRTs is very high, particularly when it is underground.
Figure 7-9 Metro in San Francisco
7.7 Automated Guideway Transit (AGT)
Automated Guideway Transit systems are characterized by the use of exclusive

guideways and vehicles operated without a driver on board.
Automated systems provide urban transit service in several cities and also operate in
controlled environments such as airports and amusement parks (e.g. Disneyworld).
Figure 7-10 Automated Guideway Transit in Kobe
7.8 Transit Costs
Figure 7-11 provides an overview of the composition of operating expenses for the
U.S. transit industry.
These major categories include transportation expense, which is basically the cost of
providing the service in the form of drivers, supervisory personnel, and fuel and
constitutes over one-third of the total costs (37.1%), and maintenance and garage
expenses, which involve primarily repairs to rolling stock, including the labor
associated with that function- this category constitutes 26.5% on the average.
The next largest category involves administrative and general expenses, including
personnel costs, insurance, and safety, which constitute 18.8% of expenses.
The balance of all other categories is 17.6%.
In total, labor-related expenses in the form of wages, salaries and fringe benefits vary
by system, but generally constitute 60 to 75% of total cost.
Operating expenses vary significantly by mode of operation and within modes for
different operating systems.
Figure 7-11 Distribution of Nationwide Transit of Operating Expenses
For comparative purposes, it is useful to examine these statistics on unit bases in

terms of the cost of providing the service per unit of service provided (number of
vehicle-mi operated annually) and per passenger carried.
These statistics are presented for overview purposes only; caution is advised in using
this kind of simplified unit pricing for cost estimation.
Two reasons for caution were previously pointed out; an additional one is the fact
that many operating expenses do not vary by veh-mi.
For instance, wage-related expenses conform more closely to vehicle-hours of
operation, with speed being an obviously important variable to examine.
Furthermore, costs related to maintenance of fixed facilities (for example, garages)
are a function of the size of the fleet required to operate peak service.
For these reasons, methods of calculating operating expense for these modes have
been developed utilizing multivariable cost allocation models that are calibrated for
the expense conditions of the existing or proposed transit system being analyzed.
A second cost allocation technique is utilization of models that have been developed
to estimate the incremental cost (or savings) of changes in existing service by
isolating fixed and variable costs.
Another operating expense estimation technique (more appropriate to situations
where no existing system is in place) is to "buildup" expenses by estimating numbers
of personnel and materials for each functional department.
These techniques have been developed in response to the common questions of
managers and planners of transit systems:
1. What is the relative financial performance of each of the routes in my system
(that is, route revenue versus route cost)? This can be answered with fully allocated
cost models.
2. What would the cost be if I modified a route (that is, more or less than current
cost)? - This can be answered with incremental cost models.
3. What would it cost to institute a brand new service (for example, light rail)?
This can be answered with a cost buildup model.
7.9 Cost Allocation Model
The preferred technique in developing multivariable cost allocation models is

illustrated and derived here.
The technique is valid for and can be applied to any urban area with proper unit data.
Transit operations can be viewed as an economic input-output model.
Money is supplied to the system and transit resources (for example, vehicle-hr,
vehicle-mi, and peak vehicles) are output by the transit system.
Mathematically, this economic relationship can be stated as shown below
Ct = f(R1, R2, R3, .., Rn)
where Ct = total cost of transit services (input)
R = transit resources provided (output)
n = number of resources provided
and the Unit Cost Model can be presented by
C = aR1 + bR2 + cR3 + dR4 + ...
7.10 Bus Service Cost Models
In a similar fashion to that used for the region rail carriers, expense accounts for the
10 major bus operators were allocated to one of four resources or variables: vehicles,
vehicle- mi, peak vehicle needs, and system revenue.
Vehicle-hours
Operating employees' wages represent by far the largest single element of cost in
most bus transit properties.
Employees engaged in operating vehicles are paid on an hourly basis; hence, the
allocation of wage expense is most properly made on the basis of hours of service on
the system.

Similarly, supervision of transportation operations is directly related to the number of
hours of service provided, and this item is also properly allocated to the vehicle-hr
category.
Vehicle-miles
Many costs are related directly to the miles a bus system operates.
Expenses such as fuel, tires, and equipment maintenance are a direct function of
miles operated.
Material expenses for vehicle bodies, brakes, engines, chassis, and transmissions are
also a function of exposure in terms of miles of service.
Consequently, these costs, together with the cost of motor fuel, taxes, and certain
other miscellaneous expenses, are assigned to the category of vehicle-mi.
Peak vehicle needs
Many individual expense items do not vary as functions for either of the foregoing
parameters (vehicle-hr or vehicle-mi).
For example, the cost of providing operating and maintenance facilities for vehicles is
determined by the number of vehicles required rather than the number of hours or
miles of service provided.
Various material expenses are also related to peak vehicle needs, including the
maintenance of building, fixtures, shop and garage, service car equipment, and other
miscellaneous shop items.
A number of broad overhead expenses will vary with the number of vehicles required
to operate the system, including depredation of equipment, general office costs, and
the salaries of general office clerks and officials.
System revenue
Operating costs resulting from injuries and damages are logically assigned to the
system revenue category.
Traffic promotion, station expenses, and federal income and other taxes are also
assigned to this category because they relate primarily to system revenue.
Example 7.1
The development of the cost allocation model for the Chicago Transit Authority bus
operations resulted in the apportionment of 14.69% of aggregate costs on the basis
of vehicle-mi, 54.62% on the basis of vehicle-hr, 25.60% allocated on peak vehicle
needs basis, and the remaining 5.09% as a function of system revenue (Table 7-1).

Table 7-1 Operating-cost-allocation-model Development - Chicago Transit Authority
- Bus
Basis of Total units Total cost % of total cost Unit cost

allocation allocated
Vehicle-miles 90,701,804 $25,431,448 14.69 $0.28/veh-mi

Vehicle-hours 8,500.071 94,572,897 54.62 $11.13/veh-mi
Peak vehicles 2210 44,330,511 25.60 $20,059.22/peak veh
System revenue $138,832,579 8,806,063 05.09 $0.06/$1 of system
Total $173,140,919 100.00 revenue
For the Chicago Transit Authority, the resultant bus operations cost allocation
formula follows:
C = 11.13H + 0.28M + 20,059.22V+ 0.06R
where C = annual cost of system operation
H = annual vehicle-hours of service
M = annual vehicle-miles of service
V = peak vehicle needs
R = annual system revenue
7.11 Costing Process
The cost characteristics of transit systems vary considerably.

For example, the operating costs of bus systems are high in comparison to capital
costs, in the ratio 5:1, whereas the opposite is true for, say, underground rail systems
(1:3).
Operating costs, in turn, are greatly influenced by labor, energy, and material costs.
Capital costs are related to the useful lives of vehicles and the infrastructure: 8 to 15
years for buses, 30 to 40 years for rail cars, and 100 years for tunnels.
In calculating comparative costs, the operating costs of each system and the capital
cost in terms of annual depreciation and interest changes are examined.
The cost effectiveness of the options under consideration can then be compared by
expressing total costs in terms of passenger-miles or passenger-kilometers.
To sum up, the following procedure provides a quick way of determining transit costs
in just enough detail to permit broad comparison of the options.

1. The characteristics of each system are examined: type, demand, capacity,
performance.
2. Approximation of operating costs can be calculated by applying unit rates for the
following:
a) Distance-related costs (energy, maintenance, servicing of vehicles, and so on,
in terms of vehicle-miles)
b) Time-related costs (operating staff wages, and other costs in terms of total
number of hours run by the fleet of vehicles in vehicle-hours)
c) Route-related costs (maintenance of roadway, track, signals, stations, in terms
of cost per mile per day)
Distance-and time-related costs are called variable costs.
Approximate operating unit costs are given in Table 7-2.
Table 7-2 Approximate Operating Unit Costs (Dollars)
Cost Bus LRT RRT

Distance cost 0.70 2.00 2.00
(car-mile)
Time cost 16 10 11
(car-hour)
Route cost 25 320 1200
(per mile of route/day)
3. Approximation of capital costs are annualized and represent depreciation and

interest changes, and are calculated on the following basis:
a) Each category of capital costs is assumed to be financed by a loan for a
term equal to its useful life.
b) An interest rate is assumed (say, 6%).
c) Constant annual payments are made on the loan and are calculated by
annualizing the cost of each element, using conventional tables.
4. The cost effectiveness of various systems is compared by expressing total costs
in terms of cost per passenger-mile.
Example 7.2
A light-rail system is proposed for a linear city about 1 million population. The track
will run 75% grade-separated and 25% through regular street intersections. The
following details apply:

Route length = 15 miles
Spacing of stop/location = 0.3 mile
Operating hours/day = 18
Operating days/years = 365
Average trip length = 3.75 miles
Journey speed = 12.5 mph
Peak-period operation = 3 hours
Passengers/day = 500,000
Average hourly boarding (peak = 12%) = 60,000
Heaviest flow in one direction (peak) = 24,000 on busiest section
Headway = 120 sec (peak); 240 sec (off-peak)
Capacity/train = 900 passengers
Capacity/car = 225 passengers
Load factor = 90%
Capital costs:
Segregated ROW: 9.16 million/mi; life 40 years
Track (double): 3.33 million/mi; life 30 years
Signals: 1.67 million/mile; life 30 years
Power: 5.00 million/mile; life 30 years
Stations (stops): 0.15 million each; life 40 years ,
Yards (2): 12.5 million each; life 40 years
Workshops (1): 25.0 million each; life 40 years
Rolling stock: 0.80 million each; life 25 years
Interest rate = 6% per annum
Operating costs:
Distance cost per car-mile = $2
Time cost per car-hour = $9
Route costs per mile per day = $375
What is the cost per passenger-mile?
Solution:
1. Daily passenger boardings:
Peak: 3 hours x 60,000 = 180,000 in 3 hours

= 60,000 passengers/hr
Off-peak: 50,000 - 180,000 = 320,000 in 15 hours
= 21,333 passengers/hr
2. Heaviest flow
Peak: 24,000; off-peak: 12,000 (assuming 50% peak-hour)
Vehicle requirements:
3. Hourly capacity (90% loading) = line 90%
Peak : 24,000/0.90 = 27000
Off - peak :12,000/0.90 = 13,000
4. Headways: peak = 120 sec; off-peak = 240 sec.
5. Frequency (trains/hr): peak:30; off-peak: 15
6. Capacity/train: 900
7. Capacity/car: 225
8. Cars/train: 4
9. Cars/hr: peak = 120; off-peak = 60
10. Round-trip time (including stopover of 21 min.):
2 hr, 24 min + 21 min = 2 hr, 45 min = 2.75 hours
11. Fleet size: 120 cars/hr x 2.75 hours / 90% = 367 cars (90% availability)
12. Car-miles/day:
Peak: 120 cars/hr x 3 hours/day x 30 mi = 10,800
Off-peak: 60 cars/hr x 15 hours/day x 30 mi = 27,000
Total = 37,800 car-miles/day
13. Train operating hours/day:
Peak: 30 trains/hr x 3 hours/day = 90 trains/day
Off-peak: 15 trains/hr x 15 hours/day = 225 trains/day
Total = 315 trains/day x 2.75 hours = 866 train-hours/day
14. Car operating hours/day:
866 train-hours/day x 4 cars/train = 3464 car-hours/day
Cost:
15. Total capital cost (in millions) = 637.3 principal
%
Annual Cost = where i = 6%
( %)

16. Annualized capital cost = $46.8 million
17. Annual operating costs:
Daily distance cost: 37,800 car-miles x $2.00 = $75,600
Daily time cost: 3464 car-hours x $9.00 = $31,176
Daily route cost: 15 route-miles x $375.00 = $5,625
Total = $112,401
Annual operating cost = $112,400 x 365 = $41.03 million
Element Unit Cost Cost Life Annual cost

(millions of (millions of
dollars) dollars)
Segregated ROW (15 mi) $9.16/mi 137.5 40 9.14
Track (15 mi) 3.33/mi 50.0 30 3.63
Signals (15 mi) 1.67/mi 25.0 30 1.82
Power (15 mi) 5.00/mi 75.0 30 5.45
Stations/stops (45 + 5) @0.15 7.5 40 0.50
Yards (2) @12.5 25.0 40 1.66
Workshop (1) @25.0 25.0 40 1.66
Rolling stock (367) @0.80 293.6 25 22.96
Total 638.6 46.8
18. Total annual cost = 46.8 + 41.03 = $87.83 million

19. Annual passenger-miles = 500,000 x 3.75 x 365
= 685 million
20. Cost/passenger-mile = 87.83/685 = $0.128

7.12 Transit System Design and Operation
Definition
L = one way route length, km
= cycle time or round trip time ... the time required for a transit unit to
complete one round trip including terminal time or layover time
V = average vehicular speed over the entire route, kph
Vs, VP = also known as "schedule speed" or" platform speed"
f = frequency of service in transit units per hour
h = headway . . . time interval between two successive transit units
N = no. of transit units
n = no. of vehicles (cars) per transit unit (n = 1 for bus system)
P = no. of passengers per vehicle, persons
Q = (passengers) flow rate, passengers per hour
Cc = car capacity, passengers/car
CT = train capacity, passengers/train
CT = n Cc
nmax =?
nmax = S/lc
S = length of the platform
lc = length of the car
Basic Relationships
Flow rate, Q
Q = fnp
f = 60 /h ( if h in min. )
= 3600/h ( if h in sec. )
Q = 3600np/h
Round trip time,
= 2 x L /V (km/km/hr)
= 2 x 60 x L/V min.
= 2 x 3600 x L/V sec.
The faster the transit travel, the shorter the travel time
No. of transit unit required, N

N = /h
= 120L/Vh ( h in min. )
The shorter the headway, the more the number of transit units required.
Speed
- average speed, (schedule speed), (platform speed)
Distance
S1
Deceleration
X Cruising
S2
Acceleration
S3
t1 t2 t3 Time
Figure 7-12 Distance Time Profile
Figure 7-13 Speed Time Profile
X = station spacing
(distance between stations)
S1, S2, S3 = travel distances corresponding to t1, t2 and t3, respectively
t1 = time to accelerate to cruise speed
t2 = time spent at cruise speed
t3 = time to decelerate to a stop
T = station dwell time
The total station-to-station travel time = travel time + dwell time
= t1 + t 2 + t3 + T

t1 = Vc/a
t3 = Vc/d
where a = acceleration rate
d = deceleration rate
Vc = cruise speed
How do we find t2 ?
S1 = at2/2 = a(V/a)2/2 = v2/2a;
S3 = dt2/2 = d(V/d)2/2 = v2/2d;
X = S1 +S2 + S3 --> S2 = X - S1 S3
t2 = S2/Vc = (X - S1 S3 )/Vc = X /Vc - (V2c/2a)/Vc - (Vc2/2d)/Vc
= X /Vc Vc/2a Vc/2d
t1+ t2+ t3 = X/ Vc + Vc/2a + Vc/2d
Station spacing
Schedule speed =
Station-to-station travel time
X
=
t1 +t2 +t3 + T
X
V =
X V V
+ + +T
V 2a 2d
What is the equation for Vs if "X" is not uniform?

then
L
V =
X V V
+ + +T
V 2a 2d
(k = no. of stops/station)
and if uniform stop dwell times are assumed for each stop/station
X
V =
V V L
k + +T +
2a 2d V
but all these equations did not include the layover time, so

X
V =
V V L
k + +T + +r
2a 2d V
7.13 Rail Capacity
The capacity of a rail line is determined by station capacity or way capacity,

whichever is smaller; usually, station capacity governs.
Capacity depends on
1) car size and train-station length,
2) allowable standees as determined by scheduling policy, and
3) minimum headway (spacing) between trains.
The minimum headway is a function not only of dwell time at major stations, but also
train length, acceleration and deceleration ratio, and train control systems.
Passenger capacity in the peak direction during peak hours can be estimated from
the following equations:
passengers/hr = (trains/hr).(cars/train).(seats/car).(passengers/seat) (1)
passengers/hr = (cars/train).(seats/car).(passengers/seat) (2)
or basing the equations on allowable levels of passenger space:
passengers/hr = (trains/hr).(cars/train).(ft2/car).( ft2/passengers) (3)
Equation 3 derives passenger capacity that is independent of the seating
configuration and is related directly to the area of the car.
Example 7.3
A rail rapid transit operates 10 four-car trains per track per hour. Schedule loads
average 1.90 passengers per seat. How many people can the line carry? Cars are 80 ft
long and can seat 80 people.
Solution
passengers/hr = (trains/hr).(cars/train).(seats/car).(passengers/seat)
= 10 x 4 x 80 x 1.9 = 6080 persons/hr
7.14 Operational Design
7.14.1 Rail Operation Design

The purpose of this section is to give an elementary idea of how the concepts of
capacity, volume, headway, and safety considerations are used in rail and bus
operation design.
We have seen that volume may be defined as the number of vehicles passing a fixed
point on the guideway in a unit of time.

Volume is related to headway given by: V = 3600/h (4)
Similarly, capacity is related to headway as given by the equation
Cv = 3600/hm (5)
where Cv , is the theoretical vehicular capacity or maximum volume (veh/hr), and h m
is the minimum headway (sec).
Theoretical passenger capacity is given by
CP = n Cc Cv = 3600nCc/hm (6)
where
Cp = theoretical passenger line capacity (number of passengers)
n = vehicles/train
Cc = maximum passenger per vehicle
If practical vehicular capacities are to be considered, the ratio of practical to
theoretical vehicular line capacities is introduced.
This ratio, called the guideway utilization factor, is denoted by .
Therefore, actual vehicle capacity is given by:
Ca = 3600/hm (7)
A load factor is usually used to express the percentage of vehicle occupancy; hence,
Actual passenger capacity = 3600 n Cc /hm (8)
where is the load factor.
Load Factor (LF) is a ratio of the average vehicle occupancy to the occupancy level for
which the vehicle was designed.
When the load factor = 1, it denotes that the vehicle is fully occupied.
The maximum number of passengers who can theoretically be squeezed into a
vehicle is called its crush load, and therefore the load factor can exceed 1.00 during,
say rush hours.
In rail design particularly, safe stopping distance is a major concern.
A safety factor is used for safe design on the brick-wall-stop (BWS) concept.
Say, for example, that when the lead vehicle on a track stops instantaneously, the
following vehicle must be able to stop safely, with a factor of safety of K.
K can be assumed to be 1.5.
The following vehicle in such cases is considered to stop with constant deceleration.
Based on this principle,
Kv o nl c
Minimum headway, hm = (9)
2d vo

where
vo = cruise speed (ft/sec)
lc = vehicle length (ft)
d = deceleration rate (ft/sec2)
Therefore, we can now write the theoretical capacity equation as
3600 nC c
Cp = (10)
Kv o / 2 d nl c / v o
For maximum capacity, we differentiate Cp with respect to v0 and obtain

2 nl c
vo = for maximum capacity (11)
K
and if we substitute v0 in Eq. 9, we get
2 nl c K
hm =
d
nd
and the maximum theoretical capacity = 2546 C c
lc K
The basic line-capacity equation tells us that if we wish to increase the line capacity,
we could achieve it in one of five ways:
1. Increase the number of passengers carried by each vehicle.
2. Increase the length of the trains.
3. Decrease the minimum allowable headway.
4. Improve the load factor.
5. Improve the guideway utilization.
Example 7.4 Rail Operation Design
A transit authority needs to design a rapid rail line to meet peak-hour demand of
10,000 passengers per hour, with a required speed of 35 to 40 ft/sec (24 to 27 mph).
The following assumptions are made: deceleration 2 ft/sec2, safety factor K = 1.35;
minimum headway = 120 sec; maximum headway = 240 sec; load factor = 0.9;
guideway utilization factor = 0.6, station platform limit = 10 vehicles (maximum); car
length = 70 ft; car capacity = 130 passengers. How many cars should a train consist of
to provide adequate passenger volume capacity? What will be the corresponding
headway?
Solution
1. Determine headway.
Cx = 3600. . . n. Cc/hx

10000 = 3600. 0.6. 0.9. 130. n/hx
Therefore, nx = 0.03937hx
nx (veh/train) hx (headways; sec)

1 25.27
2 50.54
3 75.82
'4 101.09
5 126.36 Possible range
6 151.63
7 176.91 (min. hx = 120 sec)
8 202.18
9 227.45 (max. hx = 240 sec)
10 252.72
2. 2. Examine computed headways and train size. From the brick-wall-stop (BWS)
concept:
2 nl c 2 nl c K
vo = , hm =
K d
Number of cars per Speed, vo Time headways, h 0 (sec)

train, n
Computed (hx) BWS (hx)
1 14.4 25.27 9.72

2 20.3 50.54 13.75
3 24.8 75.82 16.84
4 28.8 101.09 19.44
5 32.2 126.36 21.73
a
6 35.2 151.63 23.81
7 38.1a 176.91 25.71
8 40.7 202.18 27.49
9 43.1 227.65 29.20
a
Acceptable
3. Evaluate. Examination of the preceding table should be based on three criteria: (a)
computed speed should be in the range 35 to 40 ft/sec: (b) minimum headway = 120
sec; (c) BWS h0 should be less than 120 sec.

4. Conclusion: Six-or-seven-car trains are all right.
Six-car train: speed 35.28 ft/sec, hx = 151.63 sec
Seven-car train: speed 38.10 ft/sec, hx = 176.91 sec
Discussion:
These results are meant for peak-hour service. Naturally, for off-peak hours, the train
lengths will be different, depending on what policy headways are needed.
7.14.2 Bus Operation Design

The design of a bus route is somewhat different from rail design operation and the
differences will be evident from the description that follows.
Ultimately, an operation plan would contain information regarding the adopted
headway, cycle time, terminal time, fleet size, and the average vehicle speed.
Here dwell time is the total time spent by a bus at a stop.
Dwell time is calculated from one of the following formulas:
aA + bB + C (for two-way flow through busiest door)
td = aA + C (for one-way flow, alighting)
bB + C (for one-way flow, boarding)
where :
a, b = average alighting and boarding time per passenger in seconds,
respectively;
a = 1.5 to 2.0 sec, b = 2.5 to 3.5 sec
A, B = number of alighting and boarding passengers, respectively
C = clearance time = lost time in opening and closing doors, or to traffic delays
when bus is ready to leave; C is usually 15 sec
The frequency of service is given by
f = Q/Cc
where
f = frequency (buses/hr) required
Q = demand for service (passengers/hr)
Cc = maximum number of passengers per bus
The capacity of a bus route is governed by four factors: the street capacity, the bus
station platform capacity, the vehicle capacity, and the headway.
Each of the first three factors is independent of one another, and the headway is
influenced by all three.
Vehicle capacity depends on two factors: seating capacity and standing capacity.

The passenger capacity of a bus is given by
Ct = Ca + Cb
where
Ct = total passenger capacity per vehicle
Ca = vehicle seating capacity
Cb = vehicle standing capacity
= fraction of Cb allowed
Hence, capacity Rc of a bus routing during any time period is
60C t 60(C a C b )
Rc = =
hm hm
The fleet size, or the number of vehicles needed to serve a particular route, can be
determined, based on the time it takes a bus to complete a round trip.
d
Thus, =
vc
where:
= round-trip travel (hr)
d = distance of a round trip (miles or km)
vc= average vehicles speed (mph or km/hr)
A minimum layover and recovery time (say, 10 minutes) is provided at the end of
each round trip.
The number of vehicles needed (fleet size) can be determined from

Nf =
h
where Nf is the fleet size.
Example 7.5
A bus system needs to be set up between the Washington State University Campus
and the University of Idaho, a distance of 8.5 miles. The operating time is 30 minutes.
It has been estimated that the peak-hour demand is 400 passengers/hr and 45-seater
buses are available, which can safely accommodate 20 standees. Design the basic
system and determine the fleet size, assuming that the policy headway is 30 minutes
and that the minimum terminal time is 7.5 minutes, which may be revised if
necessary.
Solution
Operating speed,
v0 = 60L/to = 60 x 8.5/30 = 17mph; to = operating time

Policy headway = 30 min (which is arbitrary)
Terminal time = 7.5 min
60 ( 45 20 )
Headway, hmin = 60 Q/Rc = = 9.75 min (adopt 10 min)
400
Cycle time, = 2(t0 + tt) = 2(30 + 7.5) = 75 min;
Fleet size, Nf = /h = 75/10 = 7.5 = 8 vehicles;
Revised cycle time, ' = Nf h = 8 x 10 = 80 min
Revised terminal time, t't = (' - 2to)/2
= [80 - (2 x 30)]/2 = 10 min
Average speed, vc = d/ = 120L/ = 120 (8.5)80 = 12.75 mph.
In summary,
Headway, h = 10 min
Cycle time, = 80 min
Terminal time, tt = 10 min
Fleet size, Nf = 8 vehicles
Average speeds, vc = 12.75 mph
7.15 Capacity of Urban Transit
The passenger-carrying capacity of an urban transit route is the maximum number of

people that can be carried past a given location during a given period of time, under
specified operating conditions, without unreasonable delay, hazard, or restriction,
and with reasonable certainty.
Passenger-carrying capacity is naturally a less precise measure than vehicle-carrying
capacity.
The latter is a measure of the maximum number of vehicles capable of passing a
point during a given period of time multiplied by the maximum number of passengers
that can be transported in each vehicle.
For example, the person capacity of a freeway lane with bus and car traffic under
prevailing conditions of flow can be estimated as
Cp = fO1 + [(1800 1.5f)O2]
where f' = number of buses per hour
O1 = bus occupancy
O2 = car occupancy
CP = person capacity (people/hr)

A freeway lane where f = 28, O1 = 45, and O2 = 1.25 would have a capacity of
Cp = 28 x 45 + [1800 - (28 x 1.5)] x 1.25 = 3457 people/ hr
The same freeway approaching the downtown area, where the level of service has
deteriorated, would naturally have a capacity quite different from the example
because the freeway capacity of 1800 veh/hr/ln (without buses) may not prevail.
Example 7.6
An urban freeway during the peak hour carries 5200 cars with an average vehicle
occupancy of 1.25 persons and 35 buses with 45 passengers each. In addition, there
are 90 vanpools carrying 10 passengers each (including the driver). Calculate the
person flow. What percentage of passenger flow is represented by cars, vans, and
buses?
Solution
Veh/hr Persons/veh Persons/hr Percent of Percent of
vehicles persons
Cars 5200 1.25 6500 97.65 72.42
Vanpools 90 10.00 900 1.69 10.02
Buses 35 45.00 1575 0.66 17.56
Total 5325 8975 100.0 100.0
The total person flow is 8975 in the peak hour.

Cars, which represent 97.65% of the total vehicles, carry only 72.42% of the
passengers.
Vanpools and buses, which represent just 1.69% and 0.66% of the vehicles, carry
10.02% and 17.56% of the passengers, respectively.
The passenger capacity of a transit line depends on four major factors:
1. The maximum-number of vehicles per transit unit (bus, car, train)
2. The passenger capacity of the individual transit vehicles
3. The maximum possible headway, or time spacing between individual vehicles
or trains
4. The number of movement channels, loading positions, or station platform
capacity
Factors 1, 2, and 4 are independent of one another.
Factor 3 is influenced by the other three.
7.16 Route Schedules
Route schedules, or timetables, give the times that successive buses pass points
called time points.

Route schedules are distributed to the public; and are also an important starting
point in calculating requirements for vehicles and drivers.
Route schedules are most commonly constructed by maintaining approximately
uniform headways among successive vehicles during periods of time known as
schedule blocks.
The main motive for using uniform headways is that they are easier for passengers to
remember than are irregular ones.
Schedule blocks commonly Include morning and evening peaks, base (or midday)
service, night sen/ice, and (where appropriate) late night or owl service.
In addition, separate schedules may be operated on weekends and holidays.
7.16.1 Headways
o In general, headways are influenced by the demand for service, with the shortest
headways being maintained during the busiest periods.
For high-demand routes, operators may provide capacity headways - that is,
headways such that the vehicles are just filled at the maximum load point on the
route.
It should be noted, however, that there may be significant random variations in loads
from trip to trip, and that the capacity of a vehicle is at best a rough approximation,
since passengers are allowed to stand.
Consequently, "capacity headways" are likely to actually be based on some arbitrary
value of the average load factor at the maximum load point (that is, the ratio of
passengers to seats) that is intended to represent the limit of acceptable crowding.
At the other extreme, on lightly traveled routes, and at off-peak times, bus operators
often operate policy headways, that is, arbitrarily determined maximum headways
that are intended to represent the minimum acceptable standard for frequency of
service.
A third possibility is to provide headways that minimize the sum of the operating
costs and the value of time spent by passengers waiting for service.
For a given maximum load factor and spatial peaking pattern on a route, capacity
headways vary directly with the reciprocal of the patronage, measured in passengers
per hour.
The capacity headway is given by
Ca L f
h=
rQ
where
h = headway
Ca = bus seating capacity

Lf = maximum acceptable average load factor
Q = total patronage for the route in passengers per hour
r = ratio of maximum load to passengers per trip
For a given maximum load factor and spatial peaking pattern on a route, capacity
headways vary directly with the reciprocal of the patronage, measured in passengers
per hour.
Headways based on minimization of operating costs and time spent by passengers
waiting for service vary with the reciprocal of the square root of the total patronage.
In this case, the operating cost varies directly with the size of the fleet needed to
operate the route; fleet size, in turn, is a function of the total cycle time for the route
(round trip travel time, including layovers - that is, time spent stopped at the ends of
the route) and the headway.
The fleet size to operate the route is given by: N = /h
where N = number of vehicles needed to operate the route and 6 = cycle time.
The operating cost, in turn, is given by
Co = o N = o/h
where o = operating cost per vehicle-hour of operation.
The average number of passengers boarding during any given trip is Qh.
If it is assumed that, on the average, passengers wait half a headway for service, the
total cost for the passenger's waiting time is
Cw = wQh/2
where w =value of passenger's waiting time in dollars/hr.
The total cost of operating the route then, is
C = Co + C w = o/h + wQh/2
Taking the derivative of c with respect to h, setting this equal to zero, and solving for
h results in an optimum headway of
2 o
h=
wQ
It is common practice to round headways off to "convenient" intervals such as 10, 15,
or 30 min. especially if they are to be 15 min or more.
This is done because it is believed that most passengers will try to time their arrivals
at the stop to coincide with the arrival of the bus if headways are more than about 10
or 15 min, and that headways based on multiples of 10 or 15 min are easier for
passengers to remember.

Example 7.7
An urban bus route has a patronage of 500 passengers per hour and a cycle time of
2.5 hr. It is operated with buses having a seating capacity of 50 passengers. The
operating cost is $60/bus-hr and the transit operator believes that passengers value
waiting time at $10/hr. The ratio of maximum load to the total number of passengers
boarding is 0.60, and the operator's maximum load factors standard is 1.20.
Determine the capacity headway, the headway that minimizes the sum of operating
costs and passengers' time costs, and the actual headway to be operated. All
headways are to be rounded off to the nearest minute.
Capacity headway:
Ca L f 60 min (50)(1.2)(60)
h= = = 12 min
rQ hr (0.6)(500)
Total cost minimum headway:
2 o 2(60)(2.5)
h= = 60 = 14.7 min or 15 min
wQ (10)(500)
Since the capacity headway is less than the headway minimizing the sum of the
operating costs and the passengers' time costs, the buses will be overloaded if the
latter is used. Use the capacity headway.
h = 12 min.
7.16.2 Constructing Timetables
Once headways are determined for each schedule block, the actual route timetable
can be constructed.
The simplest case is that in which there are constant average travel times on each
route segment, uniform headways, and no interchange of vehicles among routes.
Under these conditions, there is a constant cycle time, the number of vehicles serving
the route must be an integer, and the cycle time must be an integral multiple of the
headway.
Note that if segment travel times vary in a nonrandom fashion, as they might as
traffic congestion increases or decreases on the route, or if vehicles are interchanged
among routes, these constraints do not apply.
In the case in which travel times are changing, it may not be possible to maintain
uniform headways, except at the dispatch point.
Where vehicles are interchanged among routes, fleet sizes for individual routes do
not have to be integers, and total cycles do not have to be integral multiples of the
headways.
Where the conditions outlined above do apply, schedules may be determined by the
following process.

First, time points are established and the average travel time is determined for all
route segments between time points.
Note that even though vehicles traveling in opposite directions most often use the
same streets, travel times in opposing directions may be different, so that separate
travel times should be determined from A to B and B to A.
In practice, travel times for routes that are not yet in existence must be established
roughly from driving times, with allowances made for time spent picking up and
discharging passengers.
Where schedules for existing routes are being adjusted, travel times may be observed
directly.
Given a headway and travel times, the cycle time and number of vehicles needed to
operate the route may be determined.
The cycle time will consist of the sum of the travel times on the individual route
segments, minimum layovers required to dampen variations in running time and
provide work breaks for drivers, and excess layover needed to make the cycle an
integral multiple of the headway.
That is,
= T + tL
where
T= t
i
i tl
tl = max(t, tw)
ti : travel time on segment i;
tL : excess layover time;
tl : minimum layover required;
tw : layover required to provide work breaks for drivers; this will often be
established as part of a labor agreement, and may be stated as either a fixed time or
a fraction of the cycle or travel time;
t : layover required to dampen variations in travel times so as to ensure that
vehicles returning from the route are dispatched on time for their next trip.
Then, from Equation N = /h and the requirement that the number or vehicles be an
integer,
N = int(T/h)
where int ( ) indicates that the quantity in parenthesis is to be rounded up to the next
highest integer, and
tL = - T

Then, given the dispatch time (or time passing any point) for any vehicle in the
schedule, the entire schedule can be determined.
Example 7.8
Construct a bus route schedule based on the running times given below. Headways
are 30 min. and minimum layovers are 7 min. at each end or 10 percent of running
time (total both layover), whichever is greater. The first bus from A to D should leave
at 8:00 a.m. The schedule should cover the time block from 8:00 a.m. to 12:00 noon.
Segment travel times:
Segment Travel time, min.

A-B 20
B-C 15
C-D 40
D-C 38
C-B 18
B-A 22
t
i
i = 20 + 15 + 40 + 38 + 22 = 153 min
tw = 2. 7 min = 14 min
t = 0.1 . 153 min = 15.3 min
T = 153 + 15.3 = 168.3 min
N = int(168.3/30) = int(5.6) = 6 vehicles
= 6. 30 = 180 min
tL = 180 168.3 = 11.7 min
tl + tL = 11.7 + 15.3 = 27 min Total layover, both terminals
Use 14-min layover at A and 13 min at fe (Note: any combination between 7 min. at A
and 20 min at & and 20 min, at A and 7 min. at is acceptable.)

Schedule
Bus A B C D Bus D C B A
1 8:00 8:20 8:35 9:15 4 7:58 8:36 8:54 9:16
2 8:30 8:50 9:05 9:45 5 8:28 9:06 9:24 9:46
3 9:00 9:20 9:35 10:15 6 8:58 9:36 9:54 10:16
4 9:30 9:50 10:05 10:45 1 9:28 10:06 10:24 10:46
5 10:00 10:20 10:35 11:15 2 9:58 10:36 10:54 11:16
6 10:30 10:50 11:05 11:45 3 10:28 11:06 11:24 11:46
1 11:00 11:20 11:35 12:15 4 10:58 11:36 11:54 12:16
2 11:30 11:50 12:05 12:45 5 11:2 8 12:06 12:24 12:46
3 12:00 12:20 12:35 1:15 6 11:5 8 12:36 12:54 1:16
7.16.3 The Scheduling Algorithm

The logic of the algorithm is to determine the minimum number of transit vehicles of
a given size required on the route during each time period, consistent with demand
and service standards.
Procedure:
1) Determine if the minimum even headway, HE , supplies enough capacity. If it
does not, calculate the required lower headway, HR , the no. of vehicles, N , and go to
step 7.
2) If HE does supply enough capacity, calculate the no. of vehicles, N , required to
meet the design volume, Q.
3) Calculate the vehicle headways, H , implied by N.
4) If the headway, H , is greater than the maximum allowable headway, HMAX ,
set N = N + 1 and return to step 3.
5) Select the next larger (than H) even headway, HR , from the set of allowable
even headways.
* If the obtained value is longer than 6 min., the only numbers that should be
used are 7.5, 10, 12, 15, 20, 30, 40, 45, and 60 min. Thus vehicle departure times
repeat themselves every hour except for a headway of 45 min.
6) If HR does not supply enough capacity, set N = N + 1 and return to step 3.
7) Record the required no. of vehicles, N , and the vehicle headways, HR.

Figure 7-14 Calculation of Vehicle Headways
7.17 Design the transit line
7.17.1 Transit vehicle travel analysis

One of the most important attributes of a transit vehicle is the time it takes to carry
passengers from their origins to their destination. The ideal situation might be for the
vehicle to carry its passengers from a single origin to a single destination, without any
intermediate stops to slow it down and increase the trip time. This is not ideal,
however, if there are other persons desiring service along the non-stop route, The
trade-off becomes:
A. Increase the number of stops along a route, to improve access to the transit
services and perhaps increase ridership.
B. Reduce the number of stops, to increase average operating speed and
reduce travel time along the route.
In this section, strategy B can be revised and refined as follows:
B1. Determine the best distance between transit stops on a route to make the
best use of the performance characteristics of the transit vehicles assigned to
that route.
Determine the best performance characteristics for transit vehicles assigned to
a particular route, given a specified spacing between transit stops on that
route.
7.17.2 Transit Vehicle Travel Regimes

In either strategy B.1 or B.2. the objective is to maximize the average operating speed
along the route. A transit vehicle with a high maximum speed seems desirable but (a)
such a vehicle will cost more to purchase, (b) it will probably consume more energy,
and (c) it may spend little or no time at its maximum speed, depending on the
distance between stations. To analyze this situation, five regimes of motion must be
defined and examined.
Figure 7-15 Diagram of five travel regimes

1. Acceleration regime. The rail vehicle leaves the station (or transit stop) and begins
to increase speed, until the desired top speed Vtop is reached. If an average
acceleration rate a is used, the time ta needed to accelerate to Vtop is ta = Vtop/a. The
distance sa needed to reach Vtop is found by substituting ta = Vtop/a into sa = 1/2 (ata2)
to get sa = 1/2 (Vtop 2/a)
2. Constant speed regime. The speed Vtop may be the maximum possible speed for a
particular vehicle, the maximum speed allowed by company policy for safety or other
reasons, or the speed reached before deceleration must begin.
3. Coasting regime. An advantage to steel wheels on steel rail is the very low rolling
resistance. A rail vehicles motors could be shut off, with little loss in speed but with
significant energy savings. Coasting is simply deceleration at a constant rate c,
primarily because of friction and air resistance. During the coasting regime, the
transit vehicles speed is reduced from Vtop to Vec. The coasting regime lasts tc time
units. Therefore, Vec = Vtop - ctc and tc = (Vtop - Vec )/c.
4. Braking regime. The brakes are applied with an average deceleration rate b, and
the vehicle is stopped from an initial speed Vi where Vi = Vtop or V i = Vec. The time
needed to stop is tb = Vi/b. The corresponding braking distance is sb = 1/2 (Vi2/b).
5. Station standing time, or dwell time. Time is needed to allow passengers to board
and leave the transit vehicle. The greater the number of passengers using a given
stop, the longer must be the time the vehicle remains motionless at that stop.
During the coasting regime, the transit vehicles speed is reduced from Vtop to Vec. The
coasting regime lasts tc time units with an average speed
Vc = 1/2 x (Vtop + Vec)
The total distance between two consecutive stations, S, is the sum of the
acceleration, coasting, and deceleration regime distances:
S = sa + sc + sbc
The distance sa needed to reach Vtop is
sa = 1/2 x (Vtop2/a)
The distance spent coasting uses the relationship s = vt:
sC = 1/2 x (Vtop + Vec)tc
The distance sbc needed to brake to a stop from the end-of-coasting speed Vec

sbc= 1/2 x (Vtop2/b)
The resulting solution for V ec is:
( )
=
( )
Example 7.9
The longest distance between stations in the proposed LRT corridor is expected to be
2.35 miles. The LRT vehicle would have an average acceleration rate of 3.0
mph/second and an average deceleration rate of 2.85 mph/second, the coasting
deceleration is c = 0.2 mph/sec. The vehicle would be able to achieve the desired top
operating speed of 44 mph. After the acceleration to Vtop, sustain that constant
speed until coasting should begin. How long (in time and distance) should the
constant speed regime last?
Solution:
For Vtop = 44mph; sa = 474.3 ft; for brake from 30mph, sc = 3807.3 ft, sb = 232.1 ft, tb =
10.5 sec. Distance at Vtop= 12,408 - 474.3 - 3807.3 - 232.1 = 7894.3 ft. tv = 125.6 sec.
T = 15.4 + 125.6 + 70 + 10.5 = 221.5
Example 7.10
Repeat Example 7.9, but maintain the top speed until the braking regime needs to
begin. Draw the resulting time-speed diagram and label the key points with time and
distance values.
Solution:
For Vtop = 44mph; sa = 474.3 ft; ta = 14.7 sec. for brake from 44 mph, sb = 499.2 ft, tb =
15.4 sec. Distance at Vtop= 12,408 - 474.3 - 499.2 = 11,434.5 ft., tv = 176.8 sec. T =
206.9 sec.
Example 7.11
Compare the total travel time and average speed for the operation described In
Example 10.9 and Example 10.10. For each of the two cases, include an average
dwell time of 30 sec. at the destination station in your calculation of total travel time
and average speed. Use an interstation spacing of 2.35 miles.
Solution:

Case Distance Time Avg. speed Avg. speed
ft/sec. mph
Top speed 12,408 206.9 60.1 40.8
Coasting 12,408 221.5 56.0 38.1

8. DATA COLLECTION
8.1 Overview
The four-stage modeling, an important tool for forecasting future demand and
performance of a transportation system, was developed for evaluating large-scale
infrastructure projects. Therefore, the four-stage modeling is less suitable for the
management and control of existing software. Since these models are applied to
large systems, they require information about travelers of the area influenced by the
system. Here the data requirement is very high, and may take years for the data
collection, data analysis, and model development. In addition, meticulous planning
and systematic approach are needed for accurate data collection and processing. This
chapter covers three important aspects of data collection, namely, survey design,
household data collection, and data analysis. Finally, a brief discussion of other
important surveys is also presented.
8.2 Survey design
Designing the data collection survey for the transportation projects is not easy. It
requires considerable experience, skill, and a sound understanding of the study area.
It is also important to know the purpose of the study and details of the modeling
approaches, since data requirement is influenced by these. Further, many practical
considerations like availability of time and money also has a strong bearing on the
survey design. In this section, we will discuss the basic information required from a
data collection, defining the study area, dividing the area into zones, and transport
network characteristics.
8.3 Information needed
Typical information required from the data collection can be grouped into four
categories, enumerated as below.
Socio-economic data: Information regarding the socio-economic characteristics of

the study area. Important ones include income, vehicle ownership, family size, etc.
This information is essential in building trip generation and modal split models.
Travel surveys: Origin-destination travel survey at households and traffic data from
cordon lines and screen lines (defined later). Former data include the number of trips
made by each member of the household, the direction of travel, destination, the cost
Transportation Systems Planning (ECIV 540) | DATA COLLECTION 134

of the travel, etc. The latter include the traffic flow, speed, and travel time
measurements. These data will be used primarily for the calibration of the models,
especially the trip distribution models.
Land use inventory: This includes data on the housing density at residential zones,
establishments at commercial and industrial zones. This data is especially useful for
trip generation models.
Network data: This includes data on the transport network and existing inventories.
Transport network data includes road network, traffic signals, junctions etc. The
service inventories include data on public and private transport networks. These
particulars are useful for the model calibration, especially for the assignment models.
8.4 Selecting Your Sample
There are two main components in determining whom you will interview. The first is
deciding what kind of people to interview. Researchers often call this group the
target population. If you do not interview the right kinds of people, you will not
successfully meet your goals.
The next thing to decide is how many people you need to interview. Statisticians
know that a small, representative sample will reflect the group from which it is
drawn. The larger the sample, the more precisely it reflects the target group.
However, the rate of improvement in the precision decreases as your sample size
increases. For example, to increase a sample from 250 to 1,000 only doubles the
precision. You must make a decision about your sample size based on factors such as:
time available, budget and necessary degree of precision.
Sample Size
Z p (1 p)
ss =
c
Where:
Z = Z value (e.g. 1.96 for 95% confidence level)
p = percentage picking a choice, expressed as decimal (.5 used for sample size
needed)
c = confidence interval, expressed as decimal (e.g., .04 = 4)
Correction for Finite Population
ss
new ss =
ss 1
1+
pop
Where: pop = population

8.5 Survey Methods
Once you have decided on your sample you must decide on your method of data
collection. Each method has advantages and disadvantages.
8.5.1 Personal Interviews

An interview is called personal when the Interviewer asks the questions face-to-face
with the Interviewee. Personal interviews can take place in the home, at a shopping
mall, on the street, outside a movie theater or polling place, and so on.
Advantages
The ability to let the Interviewee see, feel and/or taste a product.
The ability to find the target population. For example, you can find people who
have seen a film much more easily outside a theater in which it is playing than by
calling phone numbers at random.
Longer interviews are sometimes tolerated. Particularly with in-home
interviews that have been arranged in advance. People may be willing to talk longer
face-to-face than to someone on the phone.
Disadvantages
Personal interviews usually cost more per interview than other methods. This
is particularly true of in-home interviews, where travel time is a major factor.
Each mall has its own characteristics. It draws its clientele from a specific
geographic area surrounding it, and its shop profile also influences the type of client.
These characteristics may differ from the target population and create a non-
representative sample.
8.5.2 Telephone Surveys

Surveying by telephone is the most popular interviewing method in the USA. This is
made possible by nearly universal coverage (96% of homes have a telephone).
Advantages
People can usually be contacted faster over the telephone than with other
methods. If the Interviewers are using CATI (computer-assisted telephone
interviewing), the results can be available minutes after completing the last
interview.
You can dial random telephone numbers when you do not have the actual
telephone numbers of potential respondents.
CATI software, such as The Survey System, makes complex questionnaires
practical by offering many logic options. It can automatically skip questions, perform
calculations and modify questions based on the answers to earlier questions. It can
check the logical consistency of answers and can present questions or answers

choices in a random order (the last two are sometimes important for reasons
described later).
Skilled interviewers can often elicit longer or more complete answers than
people will give on their own to mail, email surveys (though some people will give
longer answers to Web page surveys). Interviewers can also ask for clarification of
unclear responses.
Some software, such as The Survey System, can combine survey answers with
pre-existing information you have about the people being interviewed.
Disadvantages
Many telemarketers have given legitimate research a bad name by claiming to
be doing research when they start a sales call. Consequently, many people are
reluctant to answer phone interviews and use their answering machines to screen
calls. Since over half of the homes in the USA have answering machines, this problem
is getting worse.
The growing number of working women often means that no one is home
during the day. This limits calling time to a "window" of about 6-9 p.m. (when you can
be sure to interrupt dinner or a favorite TV program).
You cannot show or sample products by phone.
8.5.3 Mail Surveys

Advantages
Mail surveys are among the least expensive.
This is the only kind of survey you can do if you have the names and addresses
of the target population, but not their telephone numbers.
The questionnaire can include pictures - something that is not possible over
the phone.
Mail surveys allow the respondent to answer at their leisure, rather than at
the often inconvenient moment they are contacted for a phone or personal
interview. For this reason, they are not considered as intrusive as other kinds of
interviews.
Disadvantages
Time! Mail surveys take longer than other kinds. You will need to wait several
weeks after mailing out questionnaires before you can be sure that you have gotten
most of the responses.
In populations of lower educational and literacy levels, response rates to mail
surveys are often too small to be useful. This, in effect, eliminates many immigrant
populations that form substantial markets in many areas.

8.5.4 Computer Direct Interviews
These are interviews in which the Interviewees enter their own answers directly into
a computer. They can be used at malls, trade shows, offices, and so on. The Survey
System's optional Interviewing Module and Interview Stations can easily create
computer-direct interviews. Some researchers set up a Web page survey for this
purpose.
Advantages
The virtual elimination of data entry and editing costs.
You will get more accurate answers to sensitive questions. The National
Institute of Justice has also found that computer-aided surveys among drug users get
better results than personal interviews. Employees are also more often willing to give
more honest answers to a computer than to a person or paper questionnaire.
The elimination of interviewer bias. Different interviewers can ask questions in
different ways, leading to different results. The computer asks the questions the
same way every time.
Ensuring skip patterns are accurately followed. The Survey System can ensure
people are not asked questions they should skip based on their earlier answers.
These automatic skips are more accurate than relying on an Interviewer reading a
paper questionnaire.
Response rates are usually higher. Computer-aided interviewing is still novel
enough that some people will answer a computer interview when they would not
have completed another kind of interview.
Disadvantages
The Interviewees must have access to a computer or one must be provided for
them.
As with mail surveys, computer direct interviews may have serious response
rate problems in populations of lower educational and literacy levels. This method
may grow in importance as computer use increases.
8.5.5 Email Surveys

Email surveys are both very economical and very fast. More people have email than
have full Internet access. This makes email a better choice than a Web page survey
for some populations. On the other hand, email surveys are limited to simple
questionnaires, whereas Web page surveys can include complex logic.
Advantages
Speed. An email questionnaire can gather several thousand responses within a
day or two.
There is practically no cost involved once the set up has been completed.

You can attach pictures and sound files.
The novelty element of an email survey often stimulates higher response
levels than ordinary snail mail surveys.
Disadvantages
You must possess (or purchase) a list of email addresses.
Some people will respond several times or pass questionnaires along to friends
to answer. Many programs have no check to eliminate people responding multiple
times to bias the results. The Survey Systems Email Module will only accept one reply
from each address sent the questionnaire. It eliminates duplicate and pass along
questionnaires and checks to ensure that respondents have not ignored instructions
(e.g., giving 2 answers to a question requesting only one).
Many people dislike unsolicited email even more than unsolicited regular mail.
You may want to send email questionnaires only to people who expect to get email
from you.
You cannot use email surveys to generalize findings to the whole populations.
People who have email are different from those who do not, even when matched on
demographic characteristics, such as age and gender.
Email surveys cannot automatically skip questions or randomize question or
answer choice order or use other automatic techniques that can enhance surveys the
way Web page surveys can.
8.5.6 Internet/Intranet (Web Page) Surveys

Web surveys are rapidly gaining popularity. They have major speed, cost, and
flexibility advantages, but also significant sampling limitations. These limitations
make software selection especially important and restrict the groups you can study
using this technique.
Advantages
Web page surveys are extremely fast. A questionnaire posted on a popular
Web site can gather several thousand responses within a few hours. Many people
who will respond to an email invitation to take a Web survey will do so the first day,
and most will do so within a few days.
There is practically no cost involved once the set up has been completed. Large
samples do not cost more than smaller ones (except for any cost to acquire the
sample).
You can show pictures. Some Web survey software can also show video and
play sound.
Web page questionnaires can use complex question skipping logic,
randomizations and other features not possible with paper questionnaires or most
email surveys. These features can assure better data.

Web page questionnaires can use colors, fonts and other formatting options
not possible in most email surveys.
A significant number of people will give more honest answers to questions
about sensitive topics, such as drug use or sex, when giving their answers to a
computer, instead of to a person or on paper.
On average, people give longer answers to open-ended questions on Web
page questionnaires than they do on other kinds of self-administered surveys.
Some Web survey software, such as The Survey System, can combine the
survey answers with pre-existing information you have about individuals taking a
survey.
Disadvantages
Current use of the Internet is far from universal. Internet surveys do not reflect
the population as a whole. This is true even if a sample of Internet users is selected to
match the general population in terms of age, gender and other demographics.
People can easily quit in the middle of a questionnaire. They are not as likely
to complete a long questionnaire on the Web as they would be if talking with a good
interviewer.
If your survey pops up on a web page, you often have no control over who
replies - anyone from Antartica to Zanzibar, cruising that web page may answer.
Depending on your software, there is often no control over people responding
multiple times to bias the results.
8.5.7 Scanning Questionnaires
Scanning questionnaires is a method of data collection that can be used with paper
questionnaires that have been administered in face-to-face interviews; mail surveys
or surveys completed by an Interviewer over the telephone. The Survey System can
produce paper questionnaires that can be scanned using Remark Office OMR
(available from CRS). Other software can scan questionnaires and produce ASCII Files
that can be read into The Survey System.
Advantages
Scanning can be the fastest method of data entry for paper questionnaires.
Scanning is more accurate than a person in reading a properly completed
questionnaire.
Disadvantages
Scanning is best-suited to "check the box" type surveys and bar codes.
Scanning programs have various methods to deal with text responses, but all require
additional data entry time.

Scanning is less forgiving (accurate) than a person in reading a poorly marked
questionnaire. Requires investment in additional hardware to do the actual scanning.
8.5.8 Summary of Survey Methods
Your choice of survey method will depend on several factors. These include:
Table 8-1 Comparison of survey methods
8.6 Household data
To understand the behavior and factors affecting the travel, one has got the origin of
travel when the decision for travel is made. It is where people live as family which is
the household. Therefore household data is considered to be the most basic and
authentic information about the travel pattern of a city.
Ideally one should take the details of all the people in the study to get complete
travel details. However, this is not feasible due to large requirement of time and
resources needed. In addition this will cause difficulties in handling these large data
in modeling stage. Therefore, same sample households are randomly selected and
survey is conducted to get the household data. Higher sample size is required fro
large population size, and vice-versa. Normally minimum ten percent samples are
required for population less than 50,000. But for a population more than one million
require only one percent for the same accuracy.
8.6.1 Questionnaire design

The next step in the survey is the questionnaire design. A good design will ensure
better response from the respondent and will significantly improve the quality of
data. Design of questionnaire is more of an art than a science. However few guiding
principles can be laid out. The questionnaire should be simple, direct, should take
minimum time, and should cause minimum burden to the respondent. Traditional
household survey has three major sections; household characteristics, personal
characteristics, and trip details.
Household characteristics This section includes a set of questions designed to obtain

socioeconomic information about the household. Relevant questions are: number of
members in the house, no. of employed people, number of unemployed people, age
and sex of the members in the house etc., number of two-wheelers in the house,
number of cycles, number of cars in the house etc., house ownership and family
income.
Personal characteristics This part includes questions designed to classify the

household members(older than 5) according to the following aspects: relation to the
head of the household (e.g. wife, son), sex, age, possession of a driving license,
educational level, and activity.
Trip data This part of the survey aims at detecting and characterizing all trips made
by the household members identified in the first part. A trip is normally defined as
any movement greater than 300 meters from an origin to a destination with a given
purpose. Trips are characterized on the basis of variables such as: origin and
destination, trip purpose, trip start and ending times, mode used, walking distance,
public-transport line and transfer station or bus stop (if applicable).
8.6.2 Survey administration

Once the questionnaire is ready, the next step is to conduct the actual survey with
the help of enumerators. Enumerators has to be trained first by briefing them about
the details of the survey and how to conduct the survey. They will be given random
household addresses and the questionnaire set. They have to first get permission to
be surveyed from the household. They may select a typical working day for the
survey and ask the members of the household about the details required in the
questionnaire. They may take care that each member of the household should
answer about their own travel details, except for children below 12 years. Trip details
of children below 5 years are normally ignored. Since the actual survey may take
place any time during the day, the respondents are required to answer the question
about the travel details of the previous day.
8.7 Data preparation
The raw data collected in the survey need to be processed before direct application in
the model. This is necessary, because of various errors, except in the survey both in
the selection of sample houses as well as error in filling details. In this section, we will

discuss three aspects of data preparation; data correction, data expansion, and data
validation.
8.7.1 Data correction

Various studies have identified few important errors that need to be corrected, and
are listed below.
Household size correction It may be possible that while choosing the random
samples, one may choose either larger or smaller than the average size of the
population as observed in the census data and correction should be made
accordingly.
Socio-demographic corrections It is possible that there may be differences
between the distribution of the variables sex, age, etc. between the survey,
and the population as observed from the census data. This correction is done
after the household size correction.
Non-response correction It is possible that there may not be a response from
many respondents, possible because they are on travel everyday. Corrections
should be made to accommodate this, after the previous two corrections.
Non-reported trip correction In many surveys people underestimate the non-
mandatory trips and the actual trips will be much higher than the reported
ones. Appropriate correction need to be applied for this.
8.7.2 Sample expansion

The second step in the data preparation is to amplify the survey data in order to
represent the total population of the zone. This is done with the help of expansion
factor which is defined as the ratio of the total number of household addressed in the
population to that of the surveyed. A simple expansion factor for the zone could be
of the following form.
where a is the total number of household in the original population list, b is the total
number of addresses selected as the original sample, and d is the number of samples
where no response was obtained.
8.7.3 Validation of results

In order to have confidence on the data collected from a sample population, three
validation tests are adopted usually. The first simply considers the consistency of the
data by a field visit normally done after data entry stage. The second validation is
done by choosing a computational check of the variables. For example, if age of a
person is shown some high unrealistic values like 150 years. The last is a logical check
done for the internal consistency of the data. For example, if the age of a person is
less than 18 years, then he cannot have a driving license. Once these corrections are
done, the data is ready to be used in modeling.
8.8 Other surveys
In addition to the household surveys, these other surveys are needed for complete
modeling involving four stage models. Their primary use is for the calibration and
validation of the models, or act as complementary to the household survey. These
include O-D surveys, road side interviews, and cordon and screen line counts.
8.8.1 O-D survey

Sometime four small studies, or to get a feel of the O-D pattern without doing
elaborate survey, work space interviews are conducted to find the origin-destination
of employers in a location. Although they are biased in terms of the destination, they
are random in terms of the mode of travel.
8.8.2 Road side interviews

These provide trips not registered in a household survey, especially external-internal
trips. This involves asking questions to a sample of drivers and passengers of vehicles
crossing a particular location. Unlike household survey, the respondent will be asked
with few questions like origin, destination, and trip purpose. Other information like
age, sex, and income can also be added, but it should be noted that at road-side,
drivers will not be willing to spend much time for survey.
8.8.3 Cordon and screen-line survey

These provide useful information about trips from and to external zones. For large
study area, internal cordon-line can be defined and surveying can be conducted. The
objective of the survey is primarily to collect the origin and destination zones and for
this many suitable methods can be adopted. It could be either recording the license
plate number at all the external cordon points or by post-card method.
Screen lines divide the study area into large natural zones, like either sides of a river,
with few crossing points between them. The procedure for both cordon-line and
screen-line survey are similar to road-side interview. However, these counts are
primarily used for calibration and validation of the models.
8.9 Summary
Data collection is one of the most important steps in modeling. Only if accurate data
is available, modeling becomes successful. Survey design is discussed in detail.
Household data gives important information required for data collection.

Questionnaire should be simple, less time consuming and should be designed such
that the required information is obtained with less burden on the respondent. Data
collected should be prepared well before application. Various corrections should be
made in data collection before they are used in modeling. Finally, other types of
surveys are also discussed.

Transport Planning and Modeling Introduction

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Transport Planning and Modeling Introduction

Uploaded by

Copyright:

Available Formats

CONTENTS

Transportation Systems Planning (ECIV 540) | INTRODUCTION 1

Transportation Systems Planning (ECIV 540) | INTRODUCTION 2

Transportation Systems Planning (ECIV 540) | INTRODUCTION 3

Transportation Systems Planning (ECIV 540) | INTRODUCTION 4

Transportation Systems Planning (ECIV 540) | INTRODUCTION 5

1.1 Transport planning and modeling

Transportation Systems Planning (ECIV 540) | INTRODUCTION 6

Transportation Systems Planning (ECIV 540) | INTRODUCTION 7

1.2.1 Characteristics of Transportation Demand

1.2.2 Characteristics of Transportation Supply

Transportation Systems Planning (ECIV 540) | INTRODUCTION 8

Figure 1-1 Congestion and its external effects

1.2.3 Equilibration of Supply and Demand

1.3 Modeling and Decision Making

1.3.1 Decision making Styles

1.3.1.2 Normative Decision Theory or Substantive Rationality

1.3.1.4 Group Decision Making

1.3.2 Choosing Modeling Approaches

Transportation Systems Planning (ECIV 540) | INTRODUCTION 11

1.4 Issue in Transportation Modeling

1.4.1 General Modeling Issues

1.4.1.2 Model calibration, validation, and use

Transportation Systems Planning (ECIV 540) | INTRODUCTION 12

1.4.1.3 Modeling, forecasting and judgment

1.5 The structure of the Classic Transportation Model

1.5.1 The Explicit Demand Models

Transportation Systems Planning (ECIV 540) | INTRODUCTION 14

1.5.2 The Implicit Demand Models

Transportation Systems Planning (ECIV 540) | INTRODUCTION 15

Transportation Systems Planning (ECIV 540) | INTRODUCTION 16

1.6 The Urban Transportation Planning Process

UTP = UT (studies) + UT (models)

1.6.1 Formulation of goals and objectives

1.6.3 Analytical methods

Transportation Systems Planning (ECIV 540) | INTRODUCTION 19

Transportation Systems Planning (ECIV 540) | INTRODUCTION 20

Zone Study area

Figure 1-7 Zoning of a study area

Figure 1-8 A road network coded as nodes and links

Transportation Systems Planning (ECIV 540) | INTRODUCTION 23

2.1 Basic Sampling Theory

2.1.1 Statistical Considerations

Transportation Systems Planning (ECIV 540) | DATA AND SPACE 24

Transportation Systems Planning (ECIV 540) | DATA AND SPACE 25

If a Normal distribution is assumed and a 95% confidence level is specified, this

2.1.1.3 Obtaining the Sample

Car ownership Household size % of population

With car Four or less 9

Without car Four or less 25

There are ways to proceed:

2.1.2 Conceptualization of the Sampling Problem

Transportation Systems Planning (ECIV 540) | DATA AND SPACE 26

2.1.2.2 Stratified Sample

2.1.2.3 Choice based Sampling

Low income High income Total

Bus user 0.45 0.15 0.60

Total 0.65 0.35 1.00

Transportation Systems Planning (ECIV 540) | DATA AND SPACE 27

Bus user 0.519 0.107 0.626

Total 0.750 0.250 1.000

Low income High income Total