CHAPTER 1

INTRODUCTION
LITERATURE SURVEY ORGANISATION OF THE THESIS

INTRODUCTION
Since an engineer is always concerned with the cost of products and services, the efficient optimum economic operation and planning of electric power generation system have always occupied an important position in the elctric power industry. With large interconnection of the electric networks , the energy crisis in the world and continous rise in prices , it is very essential to reduce the running charges of electric energy. A saving in the operation of the system of a small percent represents a significant reduction in operating cost as well as in the quantities of fuel consumed. The classic problem is the economic load dispatch of generating systems to achieve minimum operating cost.

1.1 LITERATURE SURVEY :

In order to have better understanding about the project the literature survey has been done. As part of literature survey, several books, magazines, journals, websites and technical papers are studied. The economic dispatch problem is one of the fundemental issues in power systems. ELD problem using Lambda iteration method is discussed in detail by D.P.Kothari and J.S.Dhillon[1]. The optimal value of lambda for a given power demand is calculated which is then used to compute the economic generations iteratively and few overall computations are involved in this approach. And an IEEE paper by Dr.Maheswarapu Sydulu has been referred to solve economic load dispatch problem neglecting losses [2]. And for the method of solving economic load dispatch problem by Particle Swarm Optimization method is discribed in detailed by M.Sudhakaran, et,al. [3].

1.2 ORGANISATION OF THE THESIS:

In chapter 2 review of economic load dispatch has been discussed. Different traditional methods applied to find out solution to economic load dispatch problem have been discussed. In chapter 3 Economic Operation of Power system, Optimum load dispatch, Cost functions and System constraints are discussed in detail. In chapter 4 the previous approaches available for solving the economic load dispatch problem were discussed and brief introduction was given for the available methods. In chapter 5 Lagrange method has been explained while considering as well as neglecting losses. The problem formulation, algorithms and flow charts for the Lagrange method have been presented. In chapter 6 A Very Fast and Effective Non- Iterative Lambda Based Method has been discussed and the algorithm and flow chart are presented. In chapter 7 we have clearly explained the concept of Particle Swarm Optimization method. Basic parameters of PSO are explained. And the, algorithm and flowchart have been represented. In chapter 8 the results obtained by solving the ELD problem for the 3 generator system and six generator system for different load demands are presented and duly discussed. In chapter 9 the conclusions deduced for the results obtained are presented along with the future scope of the project.

CHAPTER 2

ECONOMIC LOAD DISPATCH

INTRODUCTION

ECONOMIC LOAD DISPATCH

2.1 INTRODUCTION :
The economic load dispatch is an important function in modern power system like unit commitment, load forecasting, available transfer capability calculation, security analysis, scheduling of fuel purchase etc. ELD is solved traditionally using conventional optimisation technique such as Lambda iteration, non iterative lambda based approach, gradient method. Lately, ELD is also addressed by intelligent methods like genetic algorithm(GA)[6],[7] evolutionary programming(EP)[8],[9]. Dynamic programming(DP)[10], tabu search[11], Particle Swarm Optimisation(PSO)[4] etc. for calculation simplicity , existing methods use second order fuel cost functions which involve approximations and constraints are handled seperately , although sometimes valve-point effects are considered. Lamda iteration , gradient method[12],[13],[14] can solve simple ELD calculations and they are not sufficient for non linear optimization problems with several constraints, There are several intelligent methods, among them genetic algorithm is proved to be more effective in solving complex non linear problems. Lately, it has beed proved that Particle Swarm Optimization method yields better optimal solution compared to GA [6] in solving non linear optimization problems. For a practical problem like ELD, the intelligent methods must be modified accordingly so that they are suitable to solve the problem with more accurate multiple fuel cost functions and constraints and they can reduce randomness. Intelligent methods are iterative techniques that can search optimal solution depending on the problem domain and execution time limit. They are general purpose searching techniques based on principles inspired from the genetic and evolution mechanisms observed in natural systems and populations of living beings. These methods have the advantage of searching the solution space more thoroughly . The main difficulty is their sensitivity to the choice of parameters. Among intelligent methods, PSO is simple and promising. It requires less computation time and memory. It has also standard values for its parameter.

CHAPTER 3
5

ECONOMIC OPERATION OF POWER SYSTEM

INTRODUCTION OPTIMUM LOAD DISPATCH COST FUNCTION SYSTEM CONSTRAINTS

ECONOMIC OPERATION OF POWER SYSTEM

6

3.1 INTRODUCTION:
The engineers have been very successful in increasing the efficiency of boilers, turbines and generators so continously that each new added generating plants of a system operates more efficiently than any older unit on the system. In operating the system for any load condition the contribution from each plant and from each unit within a plant must be determined so that the cost of delivered power is minimum. Any plant may contain different units such as hydro, thermal, gas etc. These plants have different characteristic which gives different generating cost of any load. So that there should be a proper scheduling of plants for minimisation of cost of operation . The cost characteristic of each generating unit is non-linear. So the problem of achieving the minimum cost becomes a non-linear problem and also difficult.

3.1.1 OPTIMUM LOAD DISPATCH :

The optimum load dispatch problem involves the solution of two different problems. The first of these is unit commitment or pre dispatch problem wherein it is required to select optimally out of available genrating souces to operate to meet the expected load and provide specified margin of operating reserve over a specified period of time. The second aspect of economic dispatch is the online economic dispatch wherein it is required to distribute load among the generating units actually paralleled with the system in such manner as to minimise the total cost of supplying the minute to minute requirements of the system. The objective of this work is to solve a nonlinear economic load dispatch problem.

3.2 MATHEMATICAL MODELLING:

7

The problem of Economic Dispatch can be mathematically stated in the form of an optimization Problem with minimizing the total cost of the system as the objective function with the system limits as the constraints. The cost and the system limits can be mathematically modeled as follows.

3.2.1 COST FUNCTION :

Let C(i) mean the cost expressed in $/hr , producing energy in generator unit i, the total controllable system production cost therfore will be $/hr The genrated real power PG(i) accounts for major influence on c(i). The individual real generation are raised by increasing the prime mover torques and this requires an increased expenditure of fuel. The reactive generation QG(i) donot have any measurable influence on c(i) because they are controlled by controlling the field current. The individual production cost c(i) of generatora unit i is therefore for all practical purposes a function only of PG(i) , and for the overall controllable production cost ,thus

When the cost function C can be written as sum of terms where each term depends only upon independent variable.

3.2.2 SYSTEM CONSTRAINTS :

Broadly speaking there are two types of constraints i) ii) Equality constraints Inequality constraints

The inequality constraints are of two types (i) hard type and (ii) soft type. The hard type are those which are definite and specific like tapping range of an on-load tap changing transformer where as soft type are those which have some flexibility associated with them like the nodal

voltages and phase angles between the nodal voltages etc. soft inequality constraints have been very efficiently handled by penalty function methods.

i)

EQUALITY CONSTRAINTS :
From observation we can conclude that cost function is not affected by the reactive power

demand. So the full attention is given to real power balance in the system. Power balance requires that the controlled generation variables PG(i) obey the constraint equation

ii) i)

INEQUALITY CONSTRAINTS : Generator constraints :

The KVA loading in a generator is given by P2 +Q2 and this should not exceed a specified value of power because of the temperature rise conditions. The maximum active power generation of a source is limited again by thermal consideration and also minimum power generation is limited by the flame instability of a boiler. If the power output of the generator for optimum operation of the system is less than a specified value Pmin, the unit is not put on the busbar because it is not possible to generate that low value of power from the unit. Hence the generator power P cannot be outside the range stated by the inequality
Pmin

P Pmax

Similarly the maximum and minimum reactive power generation of a source is limited. The maximum reactive power is limited because of overheating of rotor and the minimum is limited because of the stability limit of the machine . Hence the powers of the generators Pp cannot be outside the range stated by the inequali

Qmin

Q Qmax

ii)

Voltage constraints :
9

It is essential that the voltage magnitudes and phase angles at various nodes should vary within certain limits . The normal operating angle of transmission lies between 30 to 45 degrees for transient stability reasons. A lower limit of delta assures proper utilisation of transmission capacity.

iii)

Running spare capacity constraints :

These constraints are required to meet a) the forced outages of one or more alternators on the system and b) the unexpected load on the system The total generation should be such that in addition to meeting the load demand and losses a minimum spare capacity should be available i.e, G Pp+Pso Where G is the total generation and Pso is some pre-specified power. A well planned system is one in which this spare capacity Pso is minimum.

iv)

Transmission line constraints :

The flow of active and reactive power through the transmission line is limited by thermal capability of the circuit expressed as Cp Cp max Where Cp max is the maximum loading capacity of the Pth line.

v)

Transformer taps settings:

If an auto transformer is used , the minimum tap setting could be zero and the maximum one i.e, 0 t 1.0 Similarly for the two winding transformer if the tappings are provided on the secondary side, 0tn

10

CHAPTER 4

LAGRANGE METHOD
INTRODUCTION ECONOMIC LOAD DISPATCH PROBLEM (WITHOUT AND WITH LOSSES) PROBLEM FORMULATION ALGORITHM FLOW CHART

11

LAGRANGE METHOD 4.1 INTRODUCTION : In this chapter Lagrange method has been discussed and the
algorithms are developed to solve the ELD Problem with and without losses [1]. The problem formulation for economic load dispatch problem using lambda iterative method has been given. And the algorithm and flowchart to corresponding problem are presented. The economic load dispatch problem deals with the minimization of cost of generating the power at any load demand. The study of this economic dispatch can be classified into two different ways. One is economic load dispatch without the transmission line losses and other is economic load dispatch with transmission line losses. In Lambda iteration method lambda is the variable introduced in solving constraint optimization problem and is called Lagrange multiplier. Since all the inequality constraints to be satisfied in each trial the equations are solved by the iterative method

4.2 ECONOMIC LOAD DISPATCH PROBLEM NEGLECTING LOSSES:

4.2.1 PROBLEM FORMULATION :

Economic dispatch (ED) is the scheduling of generators to minimise the total operating cost depending on the various constraints. This optimization problem can be mathematically stated as Fi(Pi) = ai + biPi + ciPi2 $/hr subjected to constraints Pgi = Pd + PL Pgmin Pg(i) Pgmax Where, i=1,2,3.....ng
12

------------------------- (1)

--------------------------- (2) ----------------------------- (3)

ai bi ci are cost coefficients Pd = Given load demand Pgi= Real power generation ng = number of generation buses Pl = Transmission power loss The system losses are obtained by Krons loss formula PL = Pgi * Bij * Pgj Where, i=j=1,2ng B loss coefficient matrix.
----------------------------------- (4)

The following formulae are used to solve the ELD problem using the Lambda iterative method --------------------------------------- (5) and Pgi is given by ---------------------------------------- (6)

and

P = Pd + PL - Pgi

-----------------------------------------

(7)

Equation describing Pgi including Loss coefficients is given by

Pgi =

------------------------------------- (8)

13

4.2.2 ALGORITHM FOR SOLVING ELD BY LAGRANGE METHOD NEGLECTING LOSSES :

STEP 1: Read data,namely cost coefficients, ai, bi, ci (i=1,2,.NG); convergence tolerance, ; step size ; and maximum allowed iterations,ITMAX,etc. STEP 2: Compute the initial values of Pgi (i=1,2,NG) and by assuming that PL =0.Then the problem can be started by Eqs (1) and (2) and the values of and Pgi (i=1,2,,NG) can be obtained Directly using Eqs.(5) and (6),respectively. STEP 3: Assume no generator has been fixed at either lower limit or at upper limit. STEP 4: Set iteration counter,IT =1. STEP 5: Compute Pgi (i=1,2,R) of generators which are not fixed at either upper or lower limits, using Eq (8),wher R is the number of participating generators. STEP 6: Compute P=PD- Pgi . STEP 7: Check P, if yes then go to STEP 11.(it means the program moves Forward without obtaining required convergence.) STEP 8: Modify new ==P,where is the step-size used to increase or decrease the value of in order to meet the STEP 7. STEP 9: IT=IT+1, =new and GOTO STEP 5 and repeat. STEP10: Check the limits of generator,if no more violations then GOTO STEP13,
14

Else fix as following. If Pgi < Pgimin then Pgi =Pgimin If Pgi>Pgi max then Pgi =Pgimax STEP11: GOTO STEP 4. STEP12: Compute the optimal total cost from Eq.(1) STEP13:Stop

15

4.2.3 FLOW CHART FOR ELD USING LAGRANGE METHOD NEGLECTING LOSSES
Start Read the values of ai, bi, ci, , ,ITMAX,Pd

Compute initial values of Pg(i), using eqs 5&6 , take Pl=0

Set IT=1

Compute P given by eq 7

If |P| new + *P IT IT+1

Pgi= Pgmin

Pgmax<Pgi<Pgmi , n

Pgi = Pgmax

Pgi

Stop

FIG 4.1 : FLOWCHART FOR ELD USING LAGRANGE METHOD(NEGLECTING LOSSES)

16

4.2.4 ALGORITHM FOR ECONOMIC LOAD DISPATCH BY LAGRANGE METHOD INCLUDING LOSSES AND CONSIDERING LIMITS:
STEP 1: Read data,namely cost coefficients, ai, bi, ci , B- coefficients, Bij, Bi0, B00 (i=1,2,.NG; j=1,2,.NG);convergence tolerance, ;step size ; and maximum allowed iterations,ITMAX,etc. STEP 2: Compute the initial values of Pgi (i=1,2,NG) and by assuming that PL =0.Then the problem can be started by Eqs.(1) and (2) and the values of and Pgi (i=1,2,,NG) can be obtained Directly using Eqs.(5) and (6) respectively. STEP 3: Assume no generator has been fixed at either lower limit or at upper limit. STEP 4: Set iteration counter,IT =1. STEP 5: Compute Pgi (i=1,2,R) of generators which are not fixed at either upper or lower limits, using Eq (8),wher R is the number of participating generators. STEP 6: Compute the transmission loss using Eq(3.14) STEP 7: Compute P=PD+PL- Pgi . i=1,2,3..NG

STEP 8: Check P, if yes then go to STEP 11.(it means the program moves Forward without obtaining required convergence.)

STEP 9: Modify new ==P,where is the step-size used to increase or decrease the value of in order to meet the STEP 7.

17

STEP 10: IT=IT+1, =new and GOTO STEP 5 and repeat. STEP11: Check the limits of generator,if no more violations then GOTO STEP13, Else fix as following. If Pgi < Pgimin then Pgi =Pgimin If Pgi>Pgi max then Pgi =Pgimax STEP12: GOTO STEP 4. STEP13:Compute the optimal total cost from (1) and transmission loss from (4) STEP14:Stop

4.2.5 FLOWCHART FOR SOLVING ELD USING LAGRANGE METHOD ( INCLUDING LOSSES)

18

Start

Read the values of ai, bi, ci, ,,ITMAX,Pd

Set IT=1

Compute initial value of using eq 6

Compute Pgi , PL given by eq 8&4 Compute P = Pd + PL - Pgi

If |P| & IT>ITM AX

new + *P

IT=IT+1, = new if Pgi<Pg Pgi= Pgmin

min

Pgi= Pgmax

Pgi>Pg 19
max

Pgi

Compute the optimum cost using eq 1

Stop

FIG 4.2 : FLOWCHART FOR ELD USING LAGRANGE METHOD(INCLUDING LOSSES)

CHAPTER 5

20

A VERY FAST AND EFFECTIVE NON-ITERATIVE -LOGIC BASED METHOD INTRODUCTION PROBLEM FORMULATION ALGORITHM FLOW CHART

A VERY FAST AND EFFECTIVE NON-ITERATIVE -LOGIC BASED METHOD

21

5.1 INTRODUCTION : In this chapter we have solved for the problem of economic load
dispatch neglecting losses by non-iterative lambda based method. The advantage of using this method is been clearly explained and the problem formulation and algorithm and flowchart have been represented. This is a direct or non iterative method it does not demand any initial guess value of for ED of units for a given P
demand

. Many of the existing conventional methods fail to impose P-

limits at voilating units. Improper selection of initial value of may cause slow convergence or at atimes leads to divergence for conventional algorithms. Further , each specified P D the problem is to be attempted afresh with new guess value of . If the load varies from P D min to PD
max

on the plant having Kunits ,the solution time becomes significantly large for higher values of

K. Unlike to this, this proposed method always offers the solution in an non-iterative mode with very low solution time as it is computationally very fast[2]. The proposed -logic method is a new contribution in the area of economic dispatch. It has two stages at first , preprepared power demand(PPD) is to be calculated. At the second stage solution vector (P1,P2,P3PK) is calculated with very less computation as is directly calculated for specified PD without any iterative approach .

5.2 ELD BY NON-ITERATIVE METHOD 5.2.1 PROBLEM FORMULATION :

This method for solving ED consists of two stages i.e, (i) Pre Prepared Power Demand Data (ii) Calculation of Solution P1,P2..Pk for specified power demand Pd. The ED condition is given as dFl/dPI = dFz/dPZ = ... ....= dFk/dPk Fi(Pi) = ai + biPi + ciPi2 $/hr. .... dFi(Pi)/dPi = = bi+2*ci*Pi --------------------------- (1)

min and max are calculated by the following eqns

22

= min at Pi =Pmin = max at Pi =Pmax

and

Pi = [-bi ] / [2*ci ] Then PPD = Pi i=1,2,K

-------------------------------------------------------------

(2) (3)

And the slope is given by Slope = [i+1-i] / [PPDi+1 PPDi ] -----------------------(4)

Scan PPDi and identify the interval i and ( i+1)

For given power demand, Pd P= Pd - PPD Then find new = slope*(P) + i --------------------------------(5)

5.2.2 ALGORITHM FOR SOLVING ELD USING NON-ITERATIVE METHOD:

Step1:Start Step2:Calculate prepaid data using i min = dFi / dPi at Pi=Pi min i max = dFi / dPi at Pi=Pimax
23

Step3: Arrange obtained

values in ascending order use following fundamental observation

Step4:Now calculate total power demand at each of ED condition. (i) min(i) then P(i) =Pmin(i)

(i) max(i) then P(i) = Pmax(i) min(i) (i) < max(i)

Step5:Calculate slope between any two intervals of power demand and Step6: Calculate P1,P2.Pk = new new = slope (P)+ j Step7:Repeat the same process for different power demands. Step8:Stop. for specified pd using equation 1 where

vector.

5.2.3 FLOWCHART FOR SOLVING ELD PROBLEM USING NONi gi ITERATIVE LAMBDA BASED METHOD i=1,2ng

Read

F (P ),

Pd

I=

Start

and Pd, i=1.ng 24 and Pd, i=1.ng

Differentiate Fi(Pgi) and identify a i, bi

Calculate min=dFi /dPgi at Pi=Pmin(i) max= dFi /dPg(i) at Pi =Pmax(i) Arrange lambda vector in ascending order If jmin(i Pi =Pmax(i)
)

Pi=Pmin(i)

j>max(
i)

Pi =[-bi ]/[2*ai ]

25

Find the sum demands at PPDj Calculate eqn(4)

of power each j as

slope

from

Scan PPD(j) and identify interval

P=Pdnew

PPDj,

If min(i) PI = Pmax(i) >max(i)

Pi = Pmin(i)

and Pd=PPDi

stop

FIG 5.1 : FLOWCHART FOR ELD USING NON-ITERATIVE METHOD

26

CHAPTER 6

PARTICLE SWARM OPTIMIZATION

INTRODUCTION ALGORITHM FLOWCHART

27

PARTICLE SWARM OPTIMIZATION

6.1 INTRODUCTION : In this chapter the new approach of solving economic load
dispatch problem has been discussed. The basic parameters and how the swarm searches for the food are clearly explained. Particle swarm optimization (PSO)[4] is a population based stochastic optimization technique developed by Dr.Ebehart and Dr.Kennedy in 1995,inspired by social behavior of bird flocking or fish schooling. Pso shares many similarities with evolutionary computation techniques such as Genetic Algorithms(GA).The system is initialized with a population of random solutions and searches for optima by updating generations. However, unlike GA,PSO has no evolution operators such as crossover and mutation In PSO, the potential solutions, called particles ,fly through the problem space by following the current optimum particles. The detailed information will be given in following sections. Compared to GA the advantages of PSO are that PSO is easy to implement and there are few parameters to adjust.PSO has been successfully applied in many areas: function optimization, artificial neural network training, fuzzy system control, and other areas where GA can be applied. PSO simulates the behaviors of bird flocking. Suppose the following scenario: a group of birds are randomly searching food in an area. There is only one piece of food in the area being searched. All the birds do not know where the food is. But they know how far the food is in each iteration. So whats the best strategy to find the food? The effective one is to follow the bird, which is nearest to the food.PSO learned from the scenario and used it to solve the optimization problems. In PSO, each single solution is a bird in the search space. We call it particle. All of particles have fitness values, which are evaluated by the fitness function to be optimized, and have velocities, which direct the flying of the particles. The particles fly through the problem space by following the current optimum particles. PSO is initialized with a group of random particles (solutions) and then searches for optima by updating generations[3]. In every iteration, each particle is updated by following two best values.

28

The first one is the best solution (fitness) it has achieved so far.(The fitness value is also stored.) This value is called pbest. Another best value that is tracked by the particle swarm optimizer is the best value, obtained so far by any particle in the population. This best value is a global best and called g-best. When a particle takes part of the population as its topological neighbors, the Best value is a local best and is called p-best. After finding the two best values, the particle updates its velocity and positions with following equation.

In the above equation, The term rand()*(pbest i -Pi(u)) is called particle memory influence. The term rand()*( gbesti -Pi(u)) is called swarm influence. Vi(u) which is the velocity of ith particle at iterations u must lie in the range Vmin Vi(u) Vmax The parameter Vmax determines the resolution, or fitness, with which regions The are to be searched between the present position and the target position. If Vmax is too high, particles may fly past good solutions. If Vmin is too small, particles may not explore sufficiently. In many experiences with PSO, Vmax was often set at 10-20% of the dynamic range on each dimension. The constants C1 and C2 pull each particle towards pbest and gbest positions.

Low values allow particles to roam far from the target regions before being tugged back. On the other hand, high values result in abrupt movement towards, or past, target regions. The acceleration constant C1 and C2 are often set to be 2.0 according to past experiences.

29

Suitable selection of inertia weight provides a balance between global and local explorations, thus requiring less iteration on average to find a sufficiently optimal solution.

In general,the inertia weight w is set according to the following equation, Where w -is the inertia weighting factor Wmax - maximum value of weighting factor Wmin - minimum value of weighting factor ITERmax - maximum number of iteration ITER - current number of iteration

6.2 SOLVING ELD BY PSO METHOD 6.2.1 PROBLEM FORMULATION :

The Economic Load Dispatch (ELD) is generating adequate electricity to meet the continuously varying consumer load demand at the least possible cost under a number of constraints. Practically, while the scheduled combination of units at each specific period of operation are listed, the ELD planning must perform the optimal generation dispatch among the operating units to satisfy the load demand, spinning reserve capacity, and practical operation constraints of generators. The objective of the ELD problem is to minimize the total fuel cost.

Where Ft = ai + biPi + ciPi2 $/hr The ELD problem is subjected to the following constraints, ----------------(1)

30

The power balance equation

gi

= Pd+ Pl

------------------- (2)

The total Transmission loss, Pl = Pm Bmn Pn --------------------- (3)

In addition, power output of each generator has to fall with in the operation limits of the Generator Pgimin Pgi Pgimax for i=1,2n Constraints Satisfaction Technique: To satisfy the equality constraint of equation (3), a loading of any one of the units is selected as the dependent loading Pd and its present value is replaced by the value calculated according to the following equation,

Pd=Pd+Pl-

Where Pd can be calculated directly from the equation (5.a) with the known power demand P D and the known values of remaining loading of the generators. Therefore the dispatch solution will always satisfy the power balance constraint provided that P d also satisfies the operation limit constraint as given in equation (5). An infeasible solution is omitted and above procedure is repeated until Pd satisfies its operation limit. Because Pl also depends on Pd, we can substitute an expression for Pl in terms of P1,P2Pd..Pn and Bmn coefficients. After substituting it in the equation (5.a), separate the independent and dependent generator terms to obtain a quadratic equation for Pd. Solving the quadratic equation for Pd, the power balance equality condition is exactly satisfied.

31

6.2.2 ALGORITHM FOR SOLVING ELD USING PSO METHOD :

STEP 1: Initialize randomly the individuals of the population according to the limit of each unit including individual dimensions, searching points, and velocities. These intial individuals must be feasible candidate solutions that satisfy the practical operation constraints. STEP 2: To each chromosome of the population the dependent unit output p d calculated from the power balance equation and Bmn Coefficient matrix. STEP 3: Calculate the evaluation on value of each individual P g1 , in the population using the elevation function f given by FT=i =1 Fi(Pi) STEP 4: Compare each individuals evaluation value with its pbest. The best evaluation value among the pbest is denoted as gbest. STEP 5: Modify the member velocity v of each individual pg, according to will be

Where 1=1,2,n

d=1,2,..m

STEP 6: Check the velocity components constraint occurring in the limits from the following conditions, If V id (u+1) > Vdmax then Vid (u+1)=Vdmax If V id (u+1) > Vdmin then Vid (u+1)=Vdmin Where Vd min = -0.5 Pgmin

32

Vd max= +0.5 Pgmax

STEP 7: Modify the member position of each individual pg according to

STEP 8: If the evaluation value of each individual is better than previous pbest, the current value is set to be pbest. If the best pbest is better than gbest, the value is set to be gbest. STEP 9: If the number of iterations reaches the maximum, then go to step10. Otherwise, go to step2 . STEP 10: The individual that generates the latest gbest is the optimal generation power of each unit with the minimum total generation cost.

33

6.2.3 FLOWCHART FOR SOLVING ELD PROBLEM USING PARTICLE SWARM OPTIMIZATION TECHNIQUE

Start Initialize the particles with random position and velocity vectors

For each particle position (p)evaluate the fitness If fitness (p) is better than fitness of pbest then

Set best of pbest as gbest

Update particle velocity and position

If Gbest is the optimal solution solution

Stop

34

FIG 6.1: FLOWCHART FOR ELD USING PSO

CHAPTER 7

RESULTS AND DISCUSSIONS

ECONOMIC LOAD DISPATCH OF THREE UNIT SYSTEM ECONOMIC LOAD DISPATCH OF SIX UNIT SYSTEM

35

RESULTS AND DISCUSSIONS

Programs are coded for the developed algorithms in MATLAB language. The developed programs are used to solve the problem of ELD for a system with three generators and a system with six generating units and the data is given in Appendix-I. Results obtained by using Particle Swarm Optimization (PSO) are compared with conventional lambda iteration method including the losses. Also, the programs developed for lambda iterative method and noniterative lambda method are tested on the systems mentioned above while neglecting losses. The results obtained are tabulated below.

7.1 ECONOMIC LOAD DISPATCH OF THREE UNIT SYSTEM 7.1.1 LAMBDA ITERATIVE METHOD (WITH LOSSES)
In this method initial value of lambda is guessed in the feasible region that can be calculated from derivative of cost function. For the convergence of the problem the delta lambda should be selected small. Here delta lambda is selected 0.0001 and the value of lambda must be chosen near the optimum point. S.NO Power demand (MW) 450 585 700 800 900 P1(MW) P2(MW) P3(MW) Loss(MW) Cost Time (Rs/Hr) (sec) 4665.1 5844.7 6872.2 7783.3 8711.8 10.52 6.67 7.06 7.71 5.07

1 2 3 4 5

184.80 241.29 289.20 330.70 372.22

198.36 255.30 304.02 364.59 389.39

68.16 90.78 110.13 127.03 144.01

1.36 2.33 3.36 4.41 5.06

36

TABLE 7.1 : RESULTS OBTAINED FOR LAMBDA ITERATIVE METHOD (WITH LOSSES) 7.1.2 PARTICLE SWARM OPTIMIZATION METHOD In this method the initial particles are randomly generated within the feasible range. The parameters c1 and c2 and inertia weight are selected for best convergence characteristics. Here c1=2 and c2=2. Here maximum value of is chosen 0.9 and minimum value is 0.4. The velocity limits are selected as Vmax=0.5*Pmax and minimum velocity is selected as Vmin= -0.5*Pmin. There are ten number of particles selected in the population. Power demand (MW) 450 585 700 800 900

S.NO

P1(MW)

P2(MW)

P3(MW)

Loss(MW)

Cost Time (Rs/Hr) (sec) 4664.1 5842.7 6868.9 7779.3 8705.8 12.58 6.31 8.06 8.81 6.30

1 2 3 4 5

204.71 268.19 322.35 369.51 416.70

188.59 241.60 286.90 326.29 365.90

58.06 77.54 94.13 108.50 122.90

1.37 2.35 3.38 4.44 5.64

TABLE 7.2 : RESULTS OBTAINED FOR PSO METHOD (WITH LOSSES)

37

7.1.3 COMPARISION OF RESULTS OBTAINED IN LAGRANGE METHOD AND PSO METHOD ( WITH LOSSES)
It has been observed that when transmission line losses are included the minimum cost was found in PSO method and the execution time is minimum for Lagrange method. Power demand (MW) 450

Method PSO

P1(MW) 204.71 184.80

P2(MW) 188.59 198.36

P3(MW) 58.06 68.16

Loss(MW) 1.37 1.36

Cost (Rs/Hr) 4664.1 4665.1

Time (sec) 12.58 10.52

LAGRANGE 450

TABLE 7.3 : RESULTS OBTAINED IN LAGRANGE METHOD AND PSO METHOD ( WITH LOSSES) FOR Pd= 450

Method PSO LAGRANGE

Power demand (MW) 900 900

P1 (MW) 416.70 372.22

P2 (MW) 365.90 389.39

P3 (MW) 122.90 144.01

Loss (MW) 5.64 5.06

Cost Time (Rs/Hr) (sec) 8705.8 8711.8 6.30 5.07

TABLE 7.4 : RESULTS OBTAINED IN LAGRANGE METHOD AND PSO METHOD ( WITH LOSSES) FOR Pd= 900

38

7.2 ECONOMIC LOAD DISPATCH OF SIX UNIT SYSTEM 7.2.1 LAMBDA ITETARIVE METHOD (WITH LOSSES)
The initial value of lambda is guessed in the feasible region that can be calculated from the derivative of cost function. For the convergence of the problem must be selected small. The convergence is largely affected by the selection of lambda value. The time taken for convergence increases than the three unit system. It is also observed that the time taken for convergence of six unit with losses case is more than without losses case.

Power demand (MW) 600 700 800 850 950

P1 (MW)

P2 (MW)

P3 (MW)

P4 (MW)

P5 (MW)

P6 (MW)

Loss (MW)

Cost (Rs/Hr)

Time (sec)

23.86 28.30 32.11 34.74 39.03

10 10 14.22 17.45 23.97

95.62 118.96 141.60 152.78 175.30

100.69 118.68 136.09 144.67 161.95

202.8 230.76 257.7 2 270.97 297.57

181.17 212.75 243.0 9 257.96 287.77

14.23 19.43 25.33 28.56 35.64

32132.1 36914.1 41927. 1 44452.1 49683.1

6.84 6.74 9.90 6.74 9.9

TABLE 7.5: RESULTS OBTAINED IN LAGRANGE MEHTOD ( WITH LOSSES)

39

7.2.2 PARTICLE SWARM OPTIMIZATION

. Power demand (MW) P1 (MW) P2 (MW) P3 (MW) P4 (MW) P5 (MW) P6 (MW) Loss (MW) Cost Time (Rs/Hr) (sec)

600 700 800 850 950

23.80 28.20 31.95 30.18 39.05

10 10 10.8 10 24.40

95.70 118.53 153.20 143.82 191.80

100.00 118.53 151.80 147.70 172.56

202.6 230.20 247.3 0 324.90 294.50

181.2 214.16 229.6 9 222.50 262.40

14.24 19.40 25.95 29.24 34.90

32091.6 8 36912.2 41896. 2 44452.0 8 49681.3 8

7.30 8.70

8.09 9.80 9.40

TABLE 7.6: RESULTS OBTAINED IN PSO MEHTOD ( WITH LOSSES)

7.2.3 COMPARISION OF RESULTS OBTAINED IN LAGRANGE METHOD AND PSO METHOD ( WITH LOSSES)

It has been observed that when transmission line losses are included the minimum cost was found in PSO method compared to conventional method. The performance depends on randomly generated particle in PSO

40

For Power Demand of 600MW

Method

P1 (MW) 23.80 23.86

P2 P3 (MW (MW)
)

P4 (MW) 100.00 100.69

P5 (MW) 202.6 202.8

P6 (MW) 181.2 181.17

Loss (MW) 14.24 14.23

Cost Time (Rs/Hr) (sec) 32091.6 8 32132.1 7.30 6.84

PSO LAGRANG E

10 10

95.70 95.62

TABLE 7.7: RESULTS OBTAINED IN LAGRANGE METHOD AND PSO METHOD ( WITH LOSSES) FOR Pd= 600 For Power Demand of 950MW Method P1 (MW) 39.0 5 39.03 P2 (MW) 24.40 23.97 P3 (MW) 191.8 0 175.30 P4 (MW) 172. 56 161.95 P5 (MW) 294.5 0 297.57 P6 (MW) 262. 40 287.77 Loss (MW) 34.9 0 35.64 Cost (Rs/Hr) 49681. 38 49683.10 Time (sec) 9.40 9.9

PSO
LAGRANGE

TABLE 7.8: RESULTS OBTAINED IN LAGRANGE METHOD AND PSO METHOD ( WITH LOSSES) FOR Pd= 950

41

7.3 ECONOMIC LOAD DISPATCH OF THREE UNIT SYSTEM 7.3.1 LAMBDA ITERATIVE METHOD (NEGLECTING LOSSES)
The initial value of lambda is guessed in the feasible region that can be calculated from the derivative of cost function. For the convergence of the problem must be selected small. The convergence is largely affected by the selection of lambda value.

S.NO

Power demand (MW) 600 800 950

P1(MW)

P2(MW)

P3(MW)

Cost (Rs/Hr) 5692.0 7341.5 8657.4

Time (sec) 0.1880 0.1920 0.2030

1 2 3

418.16 516.83 590.85

138.54 210.77 264.95

50.00 72.38 94.00

TABLE 7.9: RESULTS OBTAINED IN LAGRANGE MEHTOD ( NEGLECTING LOSSES)

7.3.2 NON-ITERATIVE LAMBDA BASED METHOD

This is a direct or non iterative method it does not demand any initial guess value of for ED of units for a given P
demand

. many of the existing conventional methods fail to impose P-limits at

voilating units. Improper selection of initial value of may cause slow convergence or at atimes leads to divergence for conventional algorithms
42

S.NO

Power demand (MW) 600 800 950

P1(MW)

P2(MW)

P3(MW)

Cost (Rs/Hr) 5701.0 7351.3 8669.8

Time (sec) 0.1720 0.1870 0.1560

1 2 3

424.25 522.94 596.96

132.70 204.94 259.11

50.00 72.11 93.91

TABLE 7.10 : RESULTS OBTAINED IN NON ITERATIVE MEHTOD ( NEGLECTING LOSSES)

7.3.3 COMPARISION OF RESULTS OBTAINED IN LAGRANGE METHOD AND NON ITERATIVE METHOD ( NEGLECTING LOSSES)
It has been observed that when transmission line losses are not included the minimum cost was found in lambda based method when compared to non-iterative lambda based method but the time of execution is minimum for non iterative method. Power demand (MW)

Method

P1(MW) 418.16 424.25

P2(MW) 138.54 132.70

P3(MW) 50.00 50.00

Cost (Rs/Hr) 5692.0 5701.0

Time (sec) 0.1880 0.1720

LAGRANGE 600 NON ITERATIVE 600

43

TABLE 7.11: RESULTS OBTAINED IN LAGRANGE METHOD AND NON ITERATIVE METHOD (NEGLECTING LOSSES) FOR Pd= 600

Method

Power demand (MW)

P1(MW) 590.85 596.96

P2(MW) 264.95 259.11

P3(MW) 94.00 93.91

Cost (Rs/Hr) 8657.4 8669.8

Time (sec) 0.2030 0.1560

LAGRANGE 950 NON ITERATIVE 950

TABLE 7.12 : RESULTS OBTAINED IN LAGRANGE METHOD AND NON ITERATIVE METHOD (NEGLECTING LOSSES) FOR Pd= 950

7.4 ECONOMIC LOAD DISPATCH OF SIX UNIT SYSTEM 7.4.1 LAMBDA ITERATIVE METHOD (NEGLECTING LOSSES)
The initial value of lambda is guessed in the feasible region that can be calculated from the derivative of cost function. For the convergence of the problem must be selected small. The convergence is largely affected by the selection of lambda value.

Power demand (MW)

P1(MW)

P2(MW)

P3(MW)

P4(MW)

P5(MW)

P6(MW)

Cost (Rs/Hr)

Time (sec)

550 600 700

109.56 119.51 125.00

150 150 150

78.96 86.14 100.49

44

66.67 72.73 84.85

130 130 130

125 125 125

8056.7 8459.0 9079.5

0.1870 0.1880 0.2030

TABLE 7.13 : RESULTS OBTAINED IN LAGRANGE MEHTOD ( NEGLECTING LOSSES)

7.4.2 NON-ITERATIVE LAMBDA BASED METHOD (NEGLECTING LOSSES)

This is a direct or non iterative method it does not demand any initial guess value of for ED of units for a given P
demand

. many of the existing conventional methods fail to impose P-limits at

voilating units. Improper selection of initial value of may cause slow convergence or at atimes leads to divergence for conventional algorithms
Power P1(MW) demand (MW) P2(MW) P3(MW) P4(MW) P5(MW) P6(MW) Cost (Rs/Hr) Time (sec)

550 600 700

103.46 114.18 135.62

173.49 191.46 227.41

74.57 82.30 97.75

62.97 69.49 82.53

130 130 130

125 125 125

8171.2 8878.6 10503.2

0.1720 0.1870 0.1410

TABLE 7.14 : RESULTS OBTAINED IN NON ITERATIVE MEHTOD ( NEGLECTING LOSSES)

7.4.3 COMPARISION OF RESULTS OBTAINED IN LAGRANGE METHOD AND NON ITERATIVE METHOD ( NEGLECTING LOSSES)
It has been observed that when transmission line losses are not included the minimum cost was found in lambda based method when compared to non-iterative lambda based method but the time of execution is minimum for non iterative method.
45

Method

Power demand (MW)

P1 (MW)

P2 (MW)

P3 (MW)

P4 (MW)

P5 (MW)

P6 (MW)

Cost (Rs/Hr)

Time (sec)

LAGRANGE

550
NON ITERATIVE

109.56 103.46

150 173.49

78.96 74.57

66.67 62.97

130 130

125 125

8056.7 8171.2

0.1870 0.1720

550

TABLE 7.15 : RESULTS OBTAINED IN LAGRANGE METHOD AND NON ITERATIVE METHOD (NEGLECTING LOSSES) FOR Pd= 550

Method

Power demand (MW)

P1 (MW)

P2 (MW)

P3 (MW)

P4 (MW)

P5 (MW)

P6 (MW)

Cost (Rs/Hr)

Time (sec)

LAGRANGE

700 700

125.00 135.62

150 227.41

100.49 97.75

84.85 82.53

130 130

125 125

9079.5 10503.2

0.2030 0.1410

NON ITERATIVE

TABLE 7.16 : RESULTS OBTAINED IN LAGRANGE METHOD AND NON ITERATIVE METHOD (NEGLECTING LOSSES) FOR Pd=700

46

CHAPTER 8

47

CONCLUSION SCOPE OF FUTURE WORK

CONCLUSION
In both study cases of three generating and six generating systems considering losses, PSO method gives the better cost than Lambda iterative method. In Lambda iterative method the number of iterations to converge increases. But in PSO method the no of iterations are not affected. In PSO method the selection of parameters c1,c2 and w are very important in the convergence of the system. And for the case of neglecting losses though Lambda iterative method gives the best result but the non-iterative method provides the result in less execution time.

8.1 SCOPE OF FUTURE WORK

Here the loss coefficients are prescribed in the problem the work may be extended for the problem when transmission loss coefficients are not given. In that case the loss coefficients can be calculated by solving the load flow problem. In PSO method selection of parameters are important. So, the parameters may be optimized by using ANN method. Any other method can be applied with PSO to improve the performance.

48

REFERENCES/ BIBILOGRAPHY
[1] D.P.Kothari , J.S.Dhillon Power System Optimization Published by Prentice Hall of India Pvt.Ltd , NewDelhi. [2] D.P.Kothari , I.J.Nagrath Modern Power System Analysis Published by Mc.Graw-Hill [3] Dr.Maheswarapu Sydulu A Very Fast And Effective Non-Iterative -Logic Based Algorithm For Economic Dispatch Of Thermal Units 1999 IEEE TENCON [4] [5] [6] [7] [8] [9] M.Sudhakaran , P.Ajay-D-Vimal Raj and T.G.Palanivelu Application Of Particle Swarm Optimization For Economic Load Dispatch Problems S.S.Rao Power System Optimization D.C.Waltersw and G.B.Scheble , Genetic Algorithm Solution of Economic Dispatch with valve point loading, IEEE Trans. Power system J.Tippayachai, W.Ongsakul and I.Ngamroo, Parallel micro genetic algorithm for constrained economic dispatch IEEE Trans Power system. N.Sinha, R.Chakrabarthi and P.K.Chattopadhyay, Evolutionary programming techniques for economic load dispatch IEEE Evol. Comput.,7(February(1))( 2003). H.T.Yang, P.C.Yang and C.L.Huang, evolutionary programming based economic dispatch For units with nonsmooth fuel cost functions,IEEE trans.power system. [10] A.J.Wood and B.F.Wollenberg, Power Generation, Operation, and Control (2nd ed.), Wiley, New York(1996). [11] W.M.Lin,F.S.Cheng and M.T.Tsay, An improved Tabu search for economic dispatch with multiple minima, IEEE Trans , Power systems [12] A.J. Wood and B.F. Wollenberg, Power Generation, Operation, and and Sons., New York (1984). [13] P. Aravindhababu and K.R. Nayar, Economic dispatch based on optimal lambda using
49

Control, John Wiley

radial basis function network, Elect. Power Energy Syst,. 24 (2002), pp. 551556. [14] IEEE Committee Report, Present practices in the economic operation IEEE Trans. Power Appa. Syst., PAS-90 (1971) 17681775. of power systems,

APPENDIX- I
CASE STUDY-1: THREE UNIT SYSTEM
The three generation units considered are having different characteristics. Their cost function characteristics are given by the following equations F1=0.00156P12+7.92P1+561 Rs/Hr F2=0.00194P22+7.85P2+310 Rs/Hr F3=0.00482P32+7.97P3+78 Rs/Hr According to the constraints considered in this work among inequality constraints only active power constraints are considered. Their operating limits of maximum and minimum power are also different. The unit operating ranges are 100MWP1600MW 100MWP2400MW 50MWP3200MW The transmission line losses can be calculated by knowing the loss coefficient. The B mn loss coefficient matrix is given by Bmn = 0.000075 0.000005 0.0000075 0.001940 0.000015 0.0000100
50

0.004820 0.000100 0.0000450

CASE STUDY-2: SIX UNIT SYSTEM

The cost function of six units are given as follows F1= 0.15240P12+38.53P1+756.79886 Rs/Hr 2 F2= 0.10587P2 +46.15916P2+451.32513 Rs/Hr F3= 0.02803P32+40.39655P3+1049.9977 Rs/Hr F4= 0.03546P42+38.30553P4+1243.5311 Rs/Hr F5= 0.02111P52+36.3278P5+1658.5596 Rs/Hr F6= 0.01799P62+38.27041P6+1356.6592 Rs/Hr The unit operating ranges are 10 MW P1 125 MW 10 MW P2 150 MW 35 MW P3 225 MW 35 MW P4 210 MW 130 MW P5 325 MW 125 MW P6 315 MW Bmn = 0.000022 0.000020 0.000019 0.000025 0.000032 0.000085 0.000026 0.000015 0.000024 0.000030 0.000069 0.000032 0.000019 0.000016 0.000017 0.000071 0.000030 0.000025 0.000015 0.000013 0.000065 0.000017 0.000024 0.000019 0.000017 0.000060 0.000013 0.000016 0.000015 0.000020 0.000140 0.000017 0.000015 0.000019 0.000026 0.000022

51

CASE STUDY-3: THREE UNIT SYSTEM ( NON- ITERATIVE LAMBDA BASED METHOD)
The cost function of three units are F1=0.00142P12+7.20P1+510 Rs/Hr F2=0.00194P22+7.85P2+310 Rs/Hr F3=0.00482P32+7.97P3+78 Rs/Hr According to the constraints considered in this work among inequality constraints only active power constraints are considered. Their operating limits of maximum and minimum power are also different. The unit operating ranges are 150MWP1600MW 100MWP2400MW 50MWP3200MW

52

53