Static and Dynamic Inventory Models Under Inflation, Time Value of Money

and Permissible Delay in Payment

Babak Khorrami
Thesis submitted to the
olle!e of "n!ineerin! and Mineral #esources
at $est Vir!inia University
in %artial fulfillment of the re&uirements
for the de!ree of
Master of Science
in
Industrial "n!ineerin!
$afik '( Iskander, Ph(D(, hair
)lan #( McKendall, *r(, Ph(D(
#al%h $( Plummer, Ph(D(
De%artment of Industrial and Mana!ement Systems "n!ineerin!
Mor!anto+n, $est Vir!inia
,--.
Key+ords/ Inventory ontrol, 0onlinear 1%timi2ation, Permissible Delay in Payment,
Inflation, Time Value of Money
Abstract
Static and Dynamic Inventory Models under Inflation, Time Value of
Money and Permissible Delay in Payment
Babak Khorrami
In this research a number of mathematical models were developed for
static and dynamic deterministic single-item inventory systems. Economic factors
such as inflation, time value of money and permissible delay in payment were
considered in developing the models. Nonlinear optimization techniques were
used to obtain the optimal policies for the systems.
First, a static single-item inventory model was considered in which
shortages are allowed and a delay is permitted in payment. In this case,
suppliers allow the customers to settle their accounts after a fied delay period
during which no interest is charged.
!n etension of the model was then considered in which all cost
components of the model are sub"ect to inflation and discounting, with constant
rates over the planning horizon. #he mathematical model of the system was
developed and a nonlinear optimization technique, $oo%e and &eeves search
method, was used to obtain the optimal policies for the system.
! dynamic deterministic single-item inventory model was also considered
in which the demand was assumed to be a linear function of time. 'uppliers
allow for a delay in payment and the cost components are sub"ect to inflation
and discounting with constant rates and continuous compounding. #he (olden
search technique was used to obtain the optimum length of replenishment cycle
such that the total cost is minimized.
)omputer applications using *isual +asic and ,athematica
were developed and several numerical eamples were solved.
iii
Dedication
I would like to dedicate this thesis to my parents. Without their support and
encouragement throughout my entire academic career I would not be where I am
today.
iv
Acknoled!ment
!s I loo% bac% over the course of my graduate study, I am truly than%ful to many
people who have each helped me during my time at -*..
I wish to than% my advisor, /r. -afi% Is%ander, for his advice, encouragement
and patience. I am grateful for the enormous time he spent reviewing every single page
of the thesis and providing constructive revision.
I also want to than% /r. !lan ,c0endall and /r. 1alph 2lummer for serving as
members of my committee and for their help and support.
I would also li%e to etend my appreciation to /r. ,a"id &araiedi for his support
during my graduate study. I always en"oyed his guidance and advice.
v
Table of "ontents3
!bstract....................................................................................................................ii
/edication ............................................................................................................... iii
!c%nowledgment......................................................................................................iv 4ist
of Figures3 .........................................................................................................i 4ist of
#ables3 ..........................................................................................................i )hapter
5ne ............................................................................................................ 6
Introduction ............................................................................................................. 6
6.6. Inventory 'ystems ......................................................................................... 7
6.7 ,easures of Effectiveness................................................................................ 8
6.9. ! review of the Inventory 'ystems !nalysis ................................................... 6:
6.;. /eterministic 'ingle Item ,odels with 'tatic /emand..................................... 6:
6.<. 'tatement of the 2roblem and 1esearch 5b"ective ......................................... 6=
)hapter #wo .......................................................................................................... 7:
4iterature 1eview ................................................................................................... 7:
7.6. 'tatic inventory systems and Inflation ........................................................... 7:
7.7. Inventory systems with time-varying demand >dynamic models? ..................... 79
7.9. Inventory models with dynamic demand and inflation .................................... 78
7.;. Inventory systems with permissible delay in payment..................................... 7@
)hapter #hree........................................................................................................ 96
Effects of Inflation, #ime *alue of ,oney and 2ermissible /elay in 2ayment on 5ptimal
2olicy of Inventory 'ystems .................................................................................... 96
vi
9.6. #he effect of inflation and time value of money on the E5A model ................. 97
9.7. ,athematical formulation for E5A model considering the time value of money 97
9.9.#he 5ptimal 'olution ..................................................................................... 9<
9.;. E5A under inflation and time value of money ................................................ 9@
9.<. !n inventory model with linear time-dependent demand rate under inflation and
time value of money ........................................................................................... 9=
9.<.6. ,athematical formulation of the Inventory system ................................... ;:
9.<.7. Numerical eample................................................................................. ;7
9.8. Economic order quantity under conditions of permissible delay in payment...... ;<
9.8.6. ,athematical Formulation....................................................................... ;<
9.8.7. Economic 5rder quantity when T t ...................................................... ;B
9.8.9. Economic 5rder Auantity when T < t ...................................................... <6
)hapter Four.......................................................................................................... <9
Inventory ,odels with !llowable 'hortages under inflationary conditions and
2ermissible /elay in 2ayment.................................................................................. <9
;.6. ! 'ingle Item Inventory ,odel >E5A? with 'hortages and 2ermissible /elay in
2ayment............................................................................................................. <9
;.6.6. 2ayment before the #otal /epletion of Inventory ..................................... <8
;.6.7. 2ayment after the #otal /epletion of Inventory........................................ <B
;.7. Effects of Inflation and #ime *alue of ,oney on a 'ingle Item Inventory ,odel
>E5A? with 'hortages and 2ermissible /elay in 2ayment....................................... 86
;.7.6. (race 2eriod Epires before the #otal /epletion of Inventory.................... 87
vii
;.7.7. (race 2eriod Epires after the #otal /epletion of Inventory ...................... 8@
;.9. 5ptimal 'olution for Inventory ,odel >E5A? with 'hortages and..................... 8B
2ermissible /elay in 2ayment .............................................................................. 8B
;.;. 5ptimal 'olution for Inventory ,odel >E5A? with 'hortages, Inflation, #ime
*alue of ,oney and 2ermissible /elay in 2ayment ................................................ @<
)hapter Five .......................................................................................................... ==
/ynamic 'ingle Item Inventory ,odels with Inflationary )onditions and 2ermissible
/elay in 2ayment. .................................................................................................. ==
<.6. !ssumptions and notations ........................................................................... =B
<.7. ,athematical Formulation of the 2roposed ,odel........................................... B7
<.7.6. 5rdering cost......................................................................................... B;
<.7.7. $olding )ost .......................................................................................... B;
<.7.9. 2urchasing )ost ..................................................................................... B<
<.7.;. )ost of interest charges for the items %ept in stoc% after the .................... B@
grace period .................................................................................................... B@
<.7.<. Interest earned during the grace period .................................................. BB
<.7.8. #he total cost function...........................................................................6::
<.9. 5ptimal 'olution and Numerical Eamples .................................................6::
)hapter 'i ...........................................................................................................6:9
)onclusions and Future 1esearch ...........................................................................6:9
1eferences............................................................................................................6:<
!22EN/I)E' .........................................................................................................6:B
viii
!ppendi !. $oo%e and &eeves, .nconstrained ,inimization 2rocedure .................66:
!ppendi +. )omputer program used to minimize the total cost function of the E5A
model with shortages and permissible delay in payment when inflation is not
considered3 .......................................................................................................669
!ppendi ). )omputer program used to minimize the total cost function of the E5A
model with shortages and permissible delay in payment when inflation is considered3
........................................................................................................................66B
ix
#ist of $i!ures%
Figure 6.63 !n inventory system................................................................................ ;
Figure 6.73 !verage Inventory................................................................................... =
Figure 6.93 Inventory behavior for a single item deterministic model with infinite input
rate. ...................................................................................................................... 67
Figure 6.;3 Inventory behavior for a single item deterministic model with finite input
rate. ...................................................................................................................... 69
Figure 6.<3 a cycle of an inventory model with infinite input rate and allowed 'hortage
............................................................................................................................. 6<
Figure 9.63 )ash Flow /iagram ............................................................................... 9;
Figure 9.73 (raphical representation of the inventory level........................................ 9B
Figure 9.93 #otal Inventory )ost and Number of 1eplenishment )ycles...................... ;9
Figure 9.;3 the inventory diagram when
T t
......................................................... ;@
# #ime........................................................................................................ ;B
Figure 9.<3 #he inventory diagram when T < t ......................................................... ;B
Figure ;.63 #he single item inventory system with shortages and permissible delay in
payment ................................................................................................................ <@
Figure ;.73 #he single item inventory system with shortages and permissible delay in
payment, when the grace period epires after depletion of the inventory. ................. 8:
Figure ;.;3 Eample of )omputer !pplication implementing $oo%e and &eeves !lgorithm
............................................................................................................................. @6
Figure ;.<3 +ehavior of #, with respect to length of grace period............................... @7
Figure ;.83 +ehavior of T
1
, with respect to length of grace period. ............................ @9
Figure ;.@3 +ehavior of the total cost of the system with respect to length of the grace
period.................................................................................................................... @9
Figure ;.=3 #otal )ost Function of the system TCP(T ,T
1
) .......................................... @;
Figure ;.B3 Eample of the )omputer !pplication developed to implement the $oo%e @8
and &eeves 'earch ,ethod...................................................................................... @8
Figure ;.6:3 1elationship between T and the length of grace period when 1C :.6. ... @=
Figure ;.663 1elationship between T
1
and the length of the grace period when 1C :.6:
............................................................................................................................. @=
Figure ;.673 1elationship between the total )ost of the 'ystem and the 4ength of the @B
(race 2eriod. ......................................................................................................... @B
Figure ;.693 #otal )ost Function, , C :.:6, 1 C :.6: ............................................... @B
Figure ;.6<3 1elationship between T
1
and the length of the grace period , -hen ..... =6
1C :.:<................................................................................................................. =6
Figure ;.7;3 1elationship between the total cost and - 1 when ,C :.:@.................... =8
Figure ;.783 1elationship between the total cost of the system and - 1, when ,C:.6<=@
Figure <.6, Inventory level as a function of time for the proposed model ................... B9
#ist of Tables%
#able 9.6. #otal cost vs. number of cycles ;7
#able 9.7. 5ptimal Inventory )ost and Number of 1eplenishment )ycles /ifferent
values of R ;;
#able ;.6. 5ptimal 5rdering 2olicies for /ifferent *alues of ,. @7
#able ;.7. 5ptimal 2olicy of the 'ystem when 1C :.6. @@
#able ;.93 5ptimal 5rdering 2olicy of the 'ystem, -hen 1 C :.:< =:
#able ;.;. 5ptimal 2olicy of the 'ystem when 1 C - :.6: =7
#able ;.<. #he optimal policy of the system when 1 C - :.9: and , varies from :.:6 to
:.6<. =9
#able <.6. 5ptimal number of replenishment cycles #otal cost of the system for
different value of 1, with respect to ,. 6:6
1
"ha&ter 'ne
Introduction
In todayDs highly competitive mar%et, production and financial decisions are
completely related to each other and must be made simultaneously. Everyday, new
products are introduced to the mar%et with different life cycles and particular demand
patterns. #his, together with sensitive customersD epectations, has forced
manufacturers, suppliers and business firms to focus on their supply chains. In a
classic supply chain, raw materials are ordered by the manufacturer and shipped to
temporary warehouses before further process. #he final products are then produced
and shipped
to intermediate warehouses in order to be shipped to customers. #o minimize the cost
of the system and maintain a high service level, an effective supply chain strategy must
ta%e into account the interrelation between financial and productionEinventory decisions
at different levels of the supply chain.
#he effect of inflation and time value of money on the determination of the total
cost of inventory systems cannot be ignored. Inflation has become an embedded aspect
of the economy, all over the world. #he inventories of raw materials and products are
capital investments of production plants and must compete from financial point of view,
with the other assets for the firmDs limited funds. #he effect of inflation and time value
of money should be considered in developing the proper mathematical representations
of production and inventory costs of the supply chain.
#here are other financial issues, involved with managing the flow of inventory
over the supply chain, which affect the optimal policy of production and procurement
2
for the system. For eample in developing the mathematical models for different
inventory systems, it is always assumed that the supplier is paid as soon as the items
are received by the system. In practice, however, this is not always true. #he suppliers
often offer their customers a fied period of delay in payment. #his grace period allows
the customer to settle the account for payment of the amount owed to the supplier
without charged interest. $owever, a relatively high interest rate is charged if the
payment is not settled by the end of the grace period.
From the customer standpoint, this grace period for settling the account can be
considered as a loan from the supplier without paying interest. 5n the other hand,
receiving this credit from the supplier will stimulate the demand, which is one of the
ma"or goals of the supplier.
In the net sections the methods developed for the analysis of inventory systems
are presented for several single item inventory models.
()() Inventory Systems
In this section, general methods used for the analysis of inventory problems are
reviewed. Inventory can be defined as the accumulation of a commodity that will be
used to satisfy some future demand for that commodity. #he commodity could be raw
material, purchased parts, semi-finished products, finished products in manufacturing,
spare parts in maintenance operations, purchased products in retailing, or purchased
supplies in service operations.
3
! schematic way to describe an inventory system is considered in this section.
)onsider the system shown in Figure6.6 where the inventory level of an item, is
affected by an input process and an output process. 4et P(t ) be the rate at which
material is added to inventory at time t and W (t)
be the rate at which material is
withdrawn from inventory. .sually, it is assumed that the output is being withdrawn to
satisfy a demand, with rate D(t) . It is assumed that D(t) is not a controllable variable.
#he output rate will equal the demand rate unless the inventory is depleted and in this
case it is said to be in an 5ut-of-stoc% or stoc%-out condition.
#he input process is partially under control in that it could be decided when and
how much to order from the sources of supply. +ecause of variable time delays in the
supplier filling the orders, the actual input rate, P(t ) , may be different from the desired
one.
#he state of the inventory system may be described by variables such as the
following3
I (t ) C the on-hand inventory level at time t
B(t ) C the bac%order level at time t
O(t) C the on-order position at time t
(t ) C the net inventory at time t
! (t ) C the inventory position at time t
#he on-hand inventory is the quantity of material in stoc% at a given time. -hen
the inventory is out of stoc%, that is I (t ) 0 , any demand that occurs is considered to
be
a shortage. 'ome of this demand may be >bac%ordered?, i.e. it is accumulated and is to
be satisfied as soon as possible.
/emand, D(t)
Input
P(t)
Inventory
'ut&ut W (t)
Figure 6.63 !n inventory system
#he net inventory is defined as the on-hand inventory level minus the bac%order level
(t) I (t) B(t
)
>6.6?
#he inventory position of the system is defined as the net inventory plus the on-
order quantity, O(t) .
! (t) (t) + O(t ) I (t) B(t ) +
O(t)
>6.7?
-ith the assumption that all shortages are bac%ordered, epressing the state of
the system in terms of the input, output and demand rate, one should have
t
(t) (0) +

[P(u)
D(u)]du
0
>6.9?
it should also be noted that
I (t ) max[0,
(t)]
>6.;?
B(t) max[0,
(t)]
>6.<?
and

D(t),

i" (t) > 0

W (t)
'
#in{D(t),
P(t)}

i" (t) 0 >6.8?

P(t) i" (t ) < 0

-hen to order and how much to order are the basic decisions in an inventory
system, and the answers to these questions determines the inventory policy of the
system.
#o define and solve inventory problems, the following issues must be
considered3
/efinition of the controllable decision variables in the system, which identify the
ordering policy. For eample, the desired policy might be to order $ units when the
inventory position drops to r units. #he decision variables are $ and r . In another
situation the desired policy might be to order $ units and have b as the maimum
bac%order level permitted.
#he effectiveness of the system must be measured based on an appropriate measure,
which is normally the relevant costs and revenues of the inventory system. !
mathematical model must be constructed to epress the measure of effectiveness
properly. #he value of the effectiveness measure usually varies with different
alternatives.
()* Measures of +ffectiveness
)ertain revenues and costs are involved in inventory systems. #he analysis
performed to determine a good policy involves identification of the relevant economic
factors and construction of a mathematical model to show how they are related to the
decision variables in the inventory policy. #he revenue and cost parameters then must
be estimated from accounting or other sources. For a given product, the revenue in a
period of time will be a function of the inventory provided and the demand realized.
#he ma"or cost components include procurement, inventory holding, shortage,
and system operating costs. 2rocurement costs, both in purchasing and production
situations, consist of a component that is independent of the procurement lot size and a
component that varies with the lot size. #he former is the fied cost per lot, sometimes
called the Fordering costG in purchasing or the Fsetup costG in production. #o purchase a
lot requires processing of purchase orders, receiving reports, accounting records, and
so on, as well as fied epenses in physically transporting, receiving and storing the lot.
If the lot is to be produced instead of purchased by the firm, then the setup cost
includes the paperwor% processing costs and costs associated with preparing
machines, equipment, and wor%ers to produce this particular product and changing the
setup after the lot has been produced. #hese costs will be incurred regardless of the
size of the lot. #he variable costs per lot depend on the price schedule used by the
supplier or on the variable production costs in manufacturing.
In modeling the procurement costs, it is often assumed that the cost of a lot is
given by
% + " ($) , where % is the fied cost per lot and " ($) is the total variable

cost. In case it is appropriate to assume

" ($) to be a linear function of $ , one may
write,
" ($) C P$ where P is the constant unit cost.
Inventory holding costs result from out-of-poc%et losses such as inventory taes,
insurance, damage, deterioration, handling, and storage space requirements. In
addition, there are opportunity losses associated with the funds tied up in inventory.
Inventories are equivalent to sums of money that are unavailable for investment in
other opportunities open to the firm.
! common method of modeling inventory holding costs is to assume that they
are proportional to the average inventory. If I (t) is the inventory at time t , the average
inventory over a period
(0, T
)
is defined as
1
T
I I (t )
dt
T
0
>6.@?
!s it is shown in Figure 6.7, I is the area under the inventory curve divided by T . If h
is the cost to carry a unit of inventory for one unit of time, the average inventory
holding cost per unit time over the interval
(0, T
)
is equal to hI and the total inventory
carrying cost over (0, T
)
is ThI .
Inventory
I (t)
#ime, t T
Figure 6.73 !verage Inventory
#o determine the amount of inventory carrying cost per unit per year, it can be
assumed that h consists of a cost proportional to the dollar value of a unit of inventory
plus a cost that is independent of the dollar value, to incorporate costs such as those
associated with storage and handling. $ence one can write h &P H w , where & is the
cost of carrying I6 of inventory for one unit of time, P is the dollar value of a unit, and
w is the cost per unit of inventory per unit time. #he factor & is called the inventory
carrying cost rate. In many inventory models w is omitted, and the average inventory
holding cost per unit time becomes, h &PI . If a demand for an out-of-stoc% item
occurs, two types of loses must be considered, the bac%order and lost sale costs. If the
demands are bac%ordered there will be some added costs such as costs of epediting,
special handling and shipping of the bac%ordered items, information-processing costs
and loss of goodwill because of customer dissatisfaction. /epending on the situation,
some other costs may occur. For instance, if the item is to be used in a production

operation, the operation may have to shut down with loss of production while the
inventory is in a bac%order condition. 5ther parts of the bac%order costs may be the
cost of informing the customer about the out-of-stoc% situation and cost of routing the
information to find when the item will be available to ship.
-hen a customer cancels an order, there would be a loss of revenue since the
sale is not completed. !nd in some cases customers may ta%e their business
somewhere else, and the system loses these customer for good, which may have a
significant impact on future planning and forecasting. +ac%order and lost sales costs
are very difficult to measure, as they include many different components.
! good way to model the shortage cost is to consider two cost components. !
constant loss, , associated with each unit demanded when the inventory is out of
stoc%. #his would be more appropriate for the lost sale case. If the item is bac%ordered
it is usually assumed that the loss is proportional to the time required to fill the
bac%order. #hus, a cost, , is defined as the cost of carrying a bac%order of one unit for
one unit of time. #he total shortage loss over a period (0, T ) , can hence be defined as
b + TB
>6.=?
-here b is the total number of shortages and B is the average bac%order position
during (0, T ) . #he average bac%order position is equal to3
1
T
B B(t)dt
T
0
>6.B?
-here B(t) is the bac%order position at time t .
(),) A revie of the Inventory Systems Analysis
In this section, some basic Inventory ,odels are reviewed and a common
approach for analyzing inventory systems is revisited.
In the first step of the problem analysis, the structure of the system must be
understood. #he ob"ective, constraints, variables and parameters of the system need to
be determined. ! mathematical model is then constructed to measure the effectiveness
of specific choices of the decision variables. #hese decision variables may include the
ordering lot size, amount of bac%order, and reorder point. !fter constructing the
ob"ective function, the constraints of the system need to be mathematically epressed
as functions of the decision variables. #o complete the analysis one should determine
the values of the decision variables, which optimize the ob"ective function sub"ect to the
constraints. In the net section, some deterministic single item models with static and
dynamic demands are revisited.
()-) Deterministic Sin!le Item Models ith Static Demand
If the demand rate for a commodity is %nown with certainty and is constant, the
system is called deterministic with static demand. It is assumed that all shortages are
bac%logged and are satisfied when the orders are received. It is also assumed that the
procurement lead time is constant and equal to , and the entire order is delivered as a
single pac%age.
#he main ob"ective of this system is to determine when to release an order and
how much should be ordered. 'ince the demand rate is constant, the policy will have
equal size lots. #he order is placed when the inventory level drops to a certain level,
called the reorder point.
!s mentioned earlier, the input >production? rate, amount of bac%order permitted
and the reorder point are the controllable variables in the system. +ased on this
assumption, four different models for the single item deterministic problem can be
considered. #hese models are shown in Figures. 6.9 and 6.;.
-hen the bac%order cost is considered to be very high or literally infinity, no
bac%logging is allowed.
Inventory 4evel
:
#ime
a? ,odel I. Infinite Input rate, bac%logging
not allowed.
Inventory 4evel
:
#ime
b? ,odel II. Infinite Input rate, bac%logging
allowed
Figure 6.93 Inventory behavior for a single item deterministic model with infinite input
rate.
Inventory 4evel
:
#ime
a? ,odel III. Finite Input rate, bac%logging
not allowed
Inventory 4evel
:
#ime
b? ,odel I*. Finite Input rate , +ac%logging
!llowed.
Figure 6.;3 Inventory behavior for a single item deterministic model with finite input
rate.
#he mathematical model for case >II?, Infinite Input rate and bac%logging allowed is
presented. #he mathematical model for case >I? can be obtained as a special case of
this model. )ases >III? and >I*? will not be considered in this research. #he following
assumptions are made in order to construct the model3
a? #he demand rate of each item is %nown and constant.
b? #he unit cost of each item is constant.
c? #he replenishment lead-time is constant.
d? 'hortages are permitted.
#he following notation will be used3
% C fied ordering cost associated with a replenishment.
p Cunit cost of production > or purchase?
h C inventory carrying cost per unit per year, h C &p , where & is the annual
inventory carrying rate.
C shortage cost per unit short
C shortage cost per unit year of shortage
$ C the order quantity
I
max
C maimum on-hand inventory level
b C maimum bac%order level permitted
T C time interval between replenishments
' C average annual cost
In this model it is assumed that a lot size of $ units is received at one particular
time. #he lot size is fied and the demand is constant, so the cycle length is constant
and equal to T $ / D . #he decision variables in this model are $ and b . #he $ units
satisfy the shortage and also build up the inventory at the warehouse up to point
I
max
. >Figure 6.<.?
I
max
T
1
T
2
Figure 6.<3 a cycle of an inventory model with infinite input rate and allowed 'hortage
#he best approach to obtain the optimal order quantity and bac%order is to
construct a mathematical model for the average cost per cycle and then minimize the
average cost per year. #he average cost per cycle is the sum of procurement cost,
inventory cost and shortage cost. #hese costs are identical for all cycles during the
planning horizon.
#he average inventory is the area under the inventory graph during which the
system is carrying stoc% divided by T >Figure 6.<?. 4et T
1
be the portion of cycle time
2
2 2
during which the system is carrying inventory and is equal to T


$ b

I
max
. #he
1
D D
average inventory level, I , is equal to3 I C
($ b)($ b)
2D
($ b)
2
T
C
2$
.
#he average bac%order position over the cycle, is the area under the bac%order
triangle divided by T . 4et T
2
be the part of cycle time during which a bac%order
position of b is built up, and is equal to T C
b
. #he average bac%order position over
2
D
the cycle is3 B
b
.b
b
.
2D
T
2$
!t this point, the average cost per cycle is developed as the sum of ordering,
purchasing, inventory and shortage costs, and is equal to3
% + p$ + hTI + TB + b
#o obtain the average annual cost, the average cost per cycle must be multiplied
by the number of the cycles per year, D / $ .
#he annual cost is denoted by ' ($,
b)
which is a function of two decision variables, the
order quantity $ , and the shortage position b .
' ($, b)
%D
+ pD
+
&p($ b)
+
(2Db + b
)
>6.6:?
$ 2$ 2$
#he optimal value for the order quantity and maimum bac%order level can be
obtained by solving the following equations3
' ($, b)
0
$
' ($, b)
0
b
!fter solving those equations the optimal order quantity and bac%order level are
obtained as follows3
$
*
2 %D
(D)
2
_
*
&p +

>6.66?


&p

&p(&p + )
,

b
*

(&p$
*
D)
&p +

>6.67?
If we consider model I in Figure 6.9>a?, the cost equation would be3
' ($)
%D
+ pD + &p
$
>6.69?
$ 2
and the optimal solution is obtained by solving the equation
d' ($, b)
0
d$
which results in
$
*

2 %D
&p
>6.6;?
#his model is the fundamental model in inventory theory %nown as the
Economic 5rder Auantity >E5A? model.
().) Statement of the Problem and /esearch 'b0ective
#his research is concerned with developing mathematical models for static
and dynamic deterministic single item inventory systems, considering economic
factors such as inflation, time value of money and permissible delay in payment.
In the first part of the research, a static single item inventory model is
considered in which shortages are allowed and a delay in payment is permitted.
#he appropriate mathematical model presenting the present value of the total
annual cost of the system is developed which is a nonlinear function of two
decision variables3 >6? the length of replenishment cycle, and >7? the length of
the period during which the inventory level is positive. In order to minimize the
total annual cost of the system, a nonlinear optimization procedure is applied. !
computer software is provided to implement the procedure.
In the second part of the research a dynamic deterministic single item
inventory model is considered in which the demand is assumed to be a linear
function of time. #he supplier allows a grace period to pay for the goods
acquired. #he grace period is a fraction of the replenishment cycle. #he
mathematical model presenting the present worth of the total cost of the system
during the planing horizon will be developed. #he ob"ective of the model is to
minimize the total cost of the system, by obtaining the optimal number of
replenishment cycles during the planning horizon. ! numerical eample will be
provided. In developing the mathematical model related to this problem, a
software called ,athematica was used which is a product of -olfram 1esearch
Institute. ,athematica is an outstanding tool for mathematical modeling,
analysis, and large-scale programming.
)hapter two of this document presents the literature related to the
problem. )hapter three gives an introduction to inventory problems that ta%e into
account the effects of inflation, discounting, and permissible delay in payment.
)hapter four gives a detailed formulation of the static single item models.
5ptimal solutions and numerical eamples are presented as well. )hapter five is
concerned with developing the mathematical formulations related to dynamic
models. )hapter si presents conclusions and suggestions for future research.
"ha&ter To
#iterature /evie
#his chapter reviews the literature available in four main areas3
6. Inventory control models that ta%e into account the effect of inflation and time value
of money with constant demand rate >static models?.
7. Inventory control models with time-varying demands >dynamic models?
9. Inventory control models which ta%e into account the effect of inflation and time
value of money, with time-varying demand rate.
;. Inventory control models with permissible delay in payments
*)() Static inventory systems and Inflation
5ne of the pioneering wor%s in this field is by &. !. +uzzacott J6B@<K. !s stated in
his paper, the annual inflation rates in most western countries ranged from = to 7:
percent and it was no longer reasonable to use the classical E5A model without
investigating its modification, when inflation results in cost increasing with time. In his
paper a classical E5A model was investigated, in which he assumed that there is a
constant inflation rate of k IEIEunit time, i.e. if the cost at time t is b(t) , at time
t + the cost would be3
b(t + ) b(t) + kb(t)
db(t ) / dt kb(t)
!s 0
0
which results in
b(t ) b e
kt
-here b
0
is the cost at time zero.
#he three basic costs of the system, >6? Inventory carrying cost, >7? purchase
cost and >9? setup or ordering cost were treated by ta%ing into consideration the effect
of inflation, and the optimal order quantity was calculated.
1. +. ,isra J6B@<K considered a general E5A model and investigated the effect
of inflation and time value of money. $e considered, different inflation rates for various
costs associated with the inventory system. ,odels that consider separate inflation
rates for each cost component of the system are more general and more realistic.
+uilding a mathematical model for this system is fairly straightforward, but the
optimization of the model is very difficult. ,isra classified the costs in two separate
categories. #he first category consists of all costs, which follow are inflation rate that is
effective inside the company. #he second category includes the costs that follow the
inflation rate of the general economy. #hese are called the internal and eternal
inflation rates.
#he replenishment cost is increased at the internal inflation rate while the
purchasing cost is increased at the eternal inflation rate. #he inventory carrying cost,
out-of poc%et costs such as costs of insurance, taes, etc., and the amount of capital
tied up in inventory would change with the eternal inflation rate. #he storage cost
depends on whether the company owns the warehouse or rents it.
#he total cost equation was developed for the system to minimize the present
worth of all future costs. #he epression for the optimal order quantity was then
obtained.
,. &. )handra and ,. 4. +ahner J6B=<K considered more general inventory
models. In the E5A model studied by +uzzacott and ,isra, it was assumed that no
shortage is permitted and the input >replenishment? rate is close to infinity, but those
assumptions were relaed in the two models studied by )handra and +ahner. In the
first model, shortage was allowed and in the second one a finite replenishment rate was
assumed. #he approach of ,isra J6B@BK was used to develop the cost epression
for each of the two systems, and the optimal values of the decision variables were
obtained.
+haba 1. 'ar%er and $aiu 2an J6BB;K studied the effect of inflation and time
value of money on the optimal ordering quantities and the maimum allowable shortage
in a finite replenishment inventory system. #he earlier studies by ,isra J6B@BK and
)handra and +ahner J6B=<K dealt with either a classical E5A model or with models in
which shortage and finite replenishment are considered separately. 'ar%er and 2an
studied the effect of inflation and time value of money in a model with shortage and finite
replenishment rate, using the same approach of ,isra. #hey developed the
present value of the total cost incurred during the planning horizon and then obtained
the optimal order quantity and maimum allowable shortage using a direct search
optimization technique.
#. 0. /atta and !. 0. 2al J6BB6K, studied the effect of inflation and time value of
money on an inventory model with dynamic demand and allowed shortage. #he
demand was assumed to be linear and time-dependent. In this paper the approach of
,isra J6B@BK was used to evaluate the present value of the total cost during the
planning horizon, T . #he total cost included cost components for replenishment
shortage and purchasing. #he number of replenishment cycles in the planning horizon,
m , and the fraction of each cycle length during which the inventory is carried for the
cycle, k were considered as decision variables.
*)*) Inventory systems ith time1varyin! demand 2dynamic models3
-. ! /onaldson J6B@@K analyzed an inventory system in which the demand for
an item in the time interval >t, tHdt? is given by f>t?dt. #he instantaneous demand rate at
time t is thus be represented by the continuous function f>t?. In his model, no shortages
were allowed, inventory at time tC : was assumed to be zero, and
replenishment lead-time was assumed to be zero. ! planning horizon equal to $ was
used. #he ob"ective was to determine the optimal times
t
i
, i 0,1,2,..., n 1 , at which to
reorder, and the number of orders, n , so that the total cost over >:, $? is minimized
and inventory is zero again at time
t
n
H .
Edward !. 'ilver J6B@BK considered the situation of a deterministic demand
pattern having a linear trend. $is ob"ective was to select the timing and sizes of
replenishments so as to %eep the total value of replenishment and carrying costs as low
as possible. $e considered an approimate solution procedure, %nown as the 'ilver-
,eal heuristic, which was developed for the general deterministic case with time-
varying demand pattern. ! special case of a positive linear trend was also
considered, which resulted in a very simple decision rule.
1. I. 2helps J6B=:K considered the classical deterministic inventory model for the
case of constant time between replenishments and a linear trend in demand. #he
optimum policy was derived and shown to apply to both positive and negative trends.
#his policy was applied on two eamples considered in earlier papers by /onaldson
J6B@@K and 'ilver J6B@BK. #he 2helps method is computationally easier than any of
earlier methods and does not require any heuristic ad"ustment. It has the operationally
desirable feature of constant intervals between replenishments.
E. 1itchie J6B=;K derived a simple optimal policy for the case of linear increasing
demand, which is analogous to the E5A for constant demand. #he eact solution for
linear increasing demand was published earlier by /onaldson J6B@@K, but that solution
does not have the simplicity of the E5A formula, which had led to development of
heuristic methods such as the 'ilver-,eal heuristic. !n eact solution, which has the
simplicity of the E5A formula, was needed which has been the ob"ective of 1itchieLs
paper.
,aitreyee /eb and 0. )haudhuri J6B=@K considered the inventory replenishment
policy for an item having a deterministic demand pattern with a linear >positive? trend
and shortages. #hey developed a heuristic to determine the decision rule for selecting
the times and sizes of replenishments over a finite time-horizon so as to minimize the
total cost. #he model of /onaldson J6B@@K was modified by introducing the concept of
inventory shortage which ma%es the problem much more mathematically comple. !
heuristic was therefore developed, by following 'ilver, to determine the decision rule for
selecting the time and sizes of replenishments and %eeping the carrying and shortage
costs as low as possible.
#. ,. ,urdeshwar J6B==K derived an analytical procedure for the above problem.
#he ob"ective was to obtain the optimal number of replenishment points and the times
at which the inventory reduces to zero.
'. 0. (oyal J6B==K stated that the heuristic method and the total relevant cost
model given by /eb and )haudhuri J6B=@K for determining the economic replenishment
policy for an item having a deterministic demand with a positive linear trend and
shortages are incorrect. $e pointed out the error, presented a correct heuristic, and
developed a model for the replenishment interval, which is consistent with 'ilverLs
model J6B@BK when shortages are not permitted.
.pendra /ave J6B=B-aK developed a single item order-level lot-size-type
inventory model for items with a deterministic time-dependent demand. #he model,
which allows for shortages, was developed for a fied finite planning horizon for which
the initial and the final inventory levels are zero. #he optimal number of replenishments
to be made and the corresponding replenishment points during the given horizon were
determined in the model.
.pendra /ave J6B=B-bK also derived a heuristic decision rule for the
replenishment of items with a linearly increasing demand rate over a finite-planning
horizon during which shortages are allowed. $e stated that the eact total cost
epression given by /eb and )haudhuri J6B=@K is incorrect. $e corrected their error and
rederived the 'ilver-,eal heuristic for items with a linearly increasing demand pattern
and shortages. #he numerical results indicated that the use of a heuristic incurs
negligible cost penalties.
*),) Inventory models ith dynamic demand and inflation
#. 0. /atta and !. 0. 2al J6BB6K, studied the effect of inflation and time value of
money for an inventory model with dynamic demand. #hey assumed that the demand is
a linear function of time and allowed for shortages in the model. #heir approach is
similar to the one used by ,isra J6B@BK. #he present value of the total cost during the
planning horizon, T , which includes replenishment cost, shortage cost and purchasing
cost, was developed. In their cost epression, the number of replenishment cycles in
the planning horizon, m , and the fraction of each cycle length during which the
inventory is carried for each cycle, k , were considered as the decision variables.
,oncer !. $ariga J6BB;K developed dynamic programming models for three
commonly used replenishment policies with time varying demand and shortages. In the
case of linear time dependent demand, a sensitivity analysis was performed. /atta and
2al assumed that the replenishment cycles are equal in length and the inventory
carrying times of the cycles are identical, but $ariga did not put any restriction on the
replenishment intervals and the length of the inventory carrying time of each cycle. !ll
models in his paper were formulated using a dynamic programming approach. #his
approach could be applied to any type of demand function and would provide a ready-
made sensitivity analysis for the length of the planning horizon, which is usually difficult
to estimate in practice.
$ariga J6BB<K also redeveloped the models of /atta and 2al by relaing the
assumption that inventory carrying times during replenishment cycles are equal. $is
model is applicable to both growing and declining mar%ets, with general continuous
time-dependent demand rates.
,. !. $ariga and ,. +en-/aya J6BB8K dealt with the inventory replenishment
problem over a fied planning horizon for items with linearly time-varying demand and
inflationary conditions. #hey developed models and optimal solution procedures with
and without shortages. #hey did not put any restriction on the length of the
replenishment cycles, thus ma%ing their proposed methods the first optimal solution
procedures for this problem.
&. 1ay and 0. '. )haudhuri J6BB@K developed a finite time-horizon deterministic
economic order quantity >E5A? inventory model with shortages, where the demand rate
at any instant depends on the on-hand inventory >stoc%-level? at that instant. #he effect
of inflation and time value of money was ta%en into account. #he model deals with the
inventory-level-dependent demand rate with shortages and effect of inflation and time
value of money.
*)-) Inventory systems ith &ermissible delay in &ayment
'. 0. (oyal J6B=<K derived mathematical models for obtaining the economic
order quantity of an item for which the supplier permits a fied delay in settling the
amount owed. $e solved some eamples to illustrate his method. !s (oyal states,
in
practice a supplier permits a fied period for settling the account and usually there is no
charge if the outstanding amount is settled within the allowed fied period. #his means
that the supplier is actually giving the customer a loan without interest during the grace
period.
#he reason that a supplier offers this grace period to the customer is that he
tries to stimulate the demand for his product. #he supplier usually epects that the
profit increases due to rising sales volume, which can compensate for the capital loss
incurred during the grace period.
/uring the period before the account has to be settled the customer can sell or
use the items and earn revenue. $ence, logically the customer would li%e to delay the
settlement of the account up to the last moment of the permissible period allowed by
supplier. !s (oyal concludes, the economic replenishment interval and order quantity
generally increase, although the annual cost decrease considerably. #he saving in
cost is a result of the permissible delay in settling the replenishment account and it
comes from the ability to delay the payment without paying any interest.
'uresh )hand and &ames -ard J6B=@K analyzed the same problem under the
assumptions of the classical E5A model, which are different from (oyalLs
assumptions, and obtained different results.
$ar% $wang and 'eong -han 'hinn J6BB8K eamined the problem of
determining the retailerLs optimal price and lot-size simultaneously when the supplier
permits delay in payment for an order of a product whose demand rate is represented
by a constant price elasticity function. #hey assumed that inventory is depleted not only
by customersL demand but also by deterioration. In their model, replenishment is
instantaneous, and shortage is not allowed. #hey developed a mathematical model for
the system as well as a solution procedure.
0un-&en )hung J6BB=K studied the problem of the economic order quantity for an
item for which the supplier permits a grace period in settling the account. First, he
showed that the total annual variable cost function is conve. #hen, with conveity, he
developed a theorem to determine the economic order quantity. #he theorem also
revealed that the economic order quantity, with permissible delay in payments, is
generally higher than the economic order quantity given by the classical economic order
quantity model.
$ung-)hang 4iao, )hih-$ung #sai and )hao-#on 'u J7:::K developed an
inventory model for initial-stoc%-dependent consumption rate when a delay in payment
is permissible. In their model, shortages are not allowed. #he effect of the inflation
rate, deterioration rate, initial-stoc%-consumption rate and delay in payment are
discussed.
+haba 'ar%er, !. ,. ,. &amal and 'hao"un -ang J7:::K developed a model to
determine an optimal ordering policy for deteriorating items under inflation, permissible
delay in payment, and allowable shortage. #he purpose of their research was to aid the
retailers in economically stoc%ing the inventory under the influence of different decision
criteria such as time value of money, inflation rates, purchase price of the product, and
deterioration rate.
In the net chapter an introduction is given for the inventory problems that ta%e
into account the effects of inflation, time value of money and permissible delay in
payment. #his will be followed by the development of new models as defined in the
research ob"ectives.
"ha&ter Three
+ffects of Inflation, Time Value of Money and Permissible Delay in Payment on
'&timal Policy of Inventory Systems
!s mentioned earlier, economic factors such as inflation and interest rate may
have substantial impact on the economic order quantity and time interval between
consecutive orders. 'everal studies have eamined the inflationary effect on the
optimal policy of the inventory systems. In this chapter the effects of time value of
money and inflation on a classical E5A model are investigated, then an inventory model
will be considered in which the demand is a linear function of time, and the effects of
inflation and time value of money on the model are presented.
'pecial types of financial contracts between the supplier and the customer may
have a significant effect on the optimal policy of the inventory system. ! situation with
a permissible delay in payment in which the supplier allows some grace period before
the customer settles the account to pay for the goods bought, has received some
attention. #he mathematical formulation of the economic order quantity with
permissible delay in payment as presented by '. 0. (oyal J6B=<K, will also be
reviewed in this chapter.
,)() The effect of inflation and time value of money on the +'4 model
In this section the effect of inflation and time value of money on the classical
E5A model is discussed. #he discussion and mathematical formulation are based on
the research wor% of 1am +. ,isra J6B@<K and ,. &aya )handra and ,ichael 4. +ahner
J6B=<K.
!s mentioned earlier, several studies investigated the inventory models with
inflationary conditions. #his chapter reviews the most frequently used inventory model,
the general E5A, under conditions of constant inflation rate, and time discounting.
First the notations and assumptions are presented, then the E5A model with the
effect of time discounting is reviewed, and finally the cost model of a classical E5A
model under inflation and time value of money is developed and the optimal order
quantity is obtained.
,)*) Mathematical formulation for +'4 model considerin! the time value of
money
Notations used in the model are as follows3
D C demand rate, unitEyear
p C purchase cost, IEunit
% C replenishment cost, IEorder
h C inventory cost, IEunitEyr
" C inflation rate IEIEyr
r C discount rate, representing the time value of money IEIEyr
R C r "
& C annual inventory carrying cost. IEIEyr
T C replenishment period, yr
C number of periods during the planning horizon
P C the present value of the total cost of the inventory system for the
first period I
P
t
C the present value of total cost of all cycles over the time horizon I
#he total cost of a classical E5A model consists of3 >6? 5rdering costM >7?
purchasing cost and >9? inventory carrying cost. It is assumed that the ordering and
purchasing costs are paid at the beginning of each period, and the inventory carrying
cost is paid continuously during the period.
#he inventory level at time t is I (t ) D(T t) , and hence, the inventory carrying
cost at time t , is equal to
N &pI (t ) &pD(T t)
!ssuming continuous compounding, the present value of carrying cost at time t ,
is equal to
&pD(T t)e
rt
#he present value of the total cost of the inventory system for the first period
can be epressed as3
T
P % + pDT + &pD

(T t )e

rt
dt
0
>9.6?

t

#he time value of money eists in each cycle of replenishment so one has to
consider its effect over the time horizon, T .
Equation >9.6? represents the total cost at the beginning of the first cycle, the
present value of total cost of all cycles over the time horizon is assumed to be
P
t
.
Figure >9.6? shows that if the beginning of the first cycle is set as a reference point of
the present value, then P
t
is given by3
P
t
P(1 + exp(rT ) + exp(2rT ) + ... + exp[(
1)rT ])
>9.7?
1 exp( rT )
_
P
t

1 exp(rT )
,
>9.9?
for the infinite planning horizon, , it can be concluded that3
P P
1 _
>9.;?

1 exp(rT )
,
2eriods
6 7 9
5
P Pe

rT
Pe
2 rT
5 Pe
( 1) rT
cycle6 cycle7 cycle9 cycle N
Figure 9.63 )ash Flow /iagram
,),)The '&timal Solution
It is observed that the present value of the total cost function, P
t
, in equation
>9.;? is a function of T , the length of a period, and P
t
is minimum if the condition
dP
t
dT
0 holds.
!t this point, the present value of the total cost function of the system is
epanded and the optimal order quantity is calculated.
)onsider the present value of the total cost in one cycle3
T
P % + pDT + &pD

(T t )e

rt
dt
0
.pon integration by parts and after some simplifications3
P

%
+
pDT
+
&pD
1

+

1
(e

rT

1)
1

>9.<?
'

T
1
;

r

r
]
then
P
t

%
1 e
rT
+
pDT
1 e
rT
+
&pDT
r(1 e
rT
)

&pD
r
2
P

%r + (r + & ) pDT

&pD
>9.8? t
r(1 e
rT
) r
2
!fter ta%ing the first derivative of
P
t
with respect to T ,
dP
t

(r + & ) pDr (1
e
rT
) r
2
e

rT
[(r + & ) pDT + %r ]

0
>9.@?
dT r
2
(1 e
rT
)
2
it can be concluded that3
2
2 3
4
2
e
rT
{
(r + & ) pDr}
{(r
+ & )
pDr
+ r
2
[(r
+ & )
pDT
+ %r ]}
>9.=?
$ence3
e
rT

(r + T ) pDr
+
r (r + & ) pDT
+ r
2
%r
>9.B?
(r + & )
pDr
(r + & )
pDr
(r + & ) pDr
.pon simplification3
e
rT
1 + rT
+
r
2
%
>9.6:?
(r + & )
pD
5n the other hand3
e
rT
1 + rT
+
(rT
)
2!
+
(rT )
3!
+
(rT )
4!
+ ... >9.66?
5ne can ignore
(rT )
3
3!
and higher terms if rT is small enough > rT 0.1 ?3
e
rT
1 + rT
+
(rT
)
2!
>9.67?
+y equating the right hand sides of equations >9.6:? and >9.67?3
r
2
%
(r + & ) pD
(rT )
2
C
2!
>9.69?
#he optimal time interval between two replenishments would be3
T
*

2 %
(r + & )
pD
>9.6;?
and
$
*

2 %D
(r + & )
p
>9.6<?
'
T
;
,)-) +'4 under inflation and time value of money
In this section the effect of inflation and time value of money is considered with
an E5A model.
#he inflation rate, " , is considered to be constant. #o build the total cost
epression of the system, three components of the total cost of the first cycle are
considered3
>6? 5rdering cost, which is paid at the beginning of the period, and is equal to % .
>7? 2urchasing cost, which is paid at the beginning of the period, and is equal to
pDT
>9? Inventory carrying cost which is paid continuously during the period, and is
equal to
&pD(T t)e
rt
e
"t
For convenience, the difference between inflation rate and interest rate is set to
be R , R r " .
#he present value of the total cost of the system in the period is equal to
P % + pDT + &pD(
1
)

+
1
(e
RT

1)

>9.68?
R

R

!s mentioned before, inflation and time value of money eist all over the
planning horizon, hence to obtain the optimal cycle length, the present value of the
total inventory cost during the planning horizon should be developed. #o do so an
infinite planning horizon is considered.
#he present value of the total cost of the system during an infinite planning
horizon is as follows3

t
P
t

n0
e

nRT
>9.6@?
P P

1
_
RT

1 e
,
#o obtain the optimal ordering interval, T , one should solve the following
equation3
dP
t
dT
0 >9.6=?
Following the same procedure as the previous section leads to an optimal order
quantity equal to3
$
*

2
%D
1 / 2
_
>9.6B?

(R
+
& ) p
,
,).) An inventory model ith linear time1de&endent demand rate under
inflation and time value of money
In this section an inventory model with a linear time dependent demand rate is
considered and the effects of inflation and time value of money on the model are
discussed.
#.0. /atta and !.0. 2al J6BB6K developed an inventory model with linear time -
dependent demand rate considering the effects of inflation and time value of money. In
their model, shortage was allowed. In this section, a simplified version of the /atta and
2alLs problem is presented.
#he notations used are the same as before. #he model is presented under
the following assumptions3
>6? #he system operates for a prescribed time horizon, ( .
>7? 1eplenishment is instantaneous.
>9? #he demand rate, D(t) , is a linearly increasing function of time, t , i.e.,
D(t) a + bt , a, b > 0 and 0 t ( .
>;? 'hortages are not allowed during the planning horizon.
><? m replenishments are made during the planning horizon, ( , and the length of
each replenishment cycle is equal to ( / m . m is the decision variable.
#he graphical representation of the inventory system is given in Figure 9.7.
Inventory
4evel
#ime
t T
0
t T
1
t T
2
t T
m1
t T
m
(
Figure 9.73 (raphical representation of the inventory level
) 0
1
,).)() Mathematical formulation of the Inventory system
#he replenishments are made at
t T
0
, T
1
, T
2
,..., T
m1
. #he last replenishment
is
made at time T
m1
, which covers the demand during the last cycle. -e must have,
T

( * )
,
)
m
) 0,1,2,3,..., m .
#he total inventory cost of the system is comprised of three components3 >6?
ordering cost, >7? holding cost, and >9? purchasing cost. #he methodology used to
develop the total inventory cost is the discounted cash flow method, in which the
present value of the total cost of the system during the planning horizon is developed
first, then the optimal ordering policy of the system is determined. In order to develop
the mathematical formulations of the system ,athematica software is used.
#here are m replenishments during the planning horizon, hence the present
value of the total ordering cost during the planning horizon is given by3
m1
%(exp((R) 1)
CR

%

exp(

RT
)
) C
exp((R / m)
1
>9.7:?
#he present worth of the total purchasing cost during the planning horizon is
given by3
m1
T
) +1
CP p


(a + bt) exp(RT
)
)dt
) 0
T
)

_
(R

* (exp((R + )( (b( b exp((R) ( + b( exp((R / m)

2(exp((R / m)

1)m
2
m
,
b( exp((R +
(R
) + 2b exp((R)(exp(
(R
)
m
( 2am + 2a exp((R)m + 2a exp(
(R
)m
m m m

2a exp((R +
(R
)m 2b exp((R)[exp(
(R
)]
m
(m + 2b exp((R +
(R
)[exp(
(R
)]
m
(m) p
m m m m
>9.76?
2resent value of the inventory holding cost for m cycles, during the planning horizon is
the following3
m1
T
) +1
C( &p


[(t T
)
)(a + bt) exp(Rt)]dt
) 0
T
)
C

1
(&p(b(

exp(2(R / m)

(exp((R / m)
1+ m

exp((R / m) m exp((R / m)
_
_
+
m
2
R
3

(1 exp((R / m)
2

(1 exp((R / m)
2
m
_
exp((R / m) 1

,
* R(m m exp((R / m) + (R)

(exp((R /
m)

(2bm
2
+ 2bm
2
exp((R / m)

exp((R / m) 1 (exp((R / m) 1)

2b(mR am
2
R + a exp((R / m)m
2
R b(
2
R
2
a(mR
2
)))
>9.77?
$ence the present value of the total inventory cost of the system during the
entire planning horizon is equal to3
TC (m) C( + CP +
CR
>9.79?
#he above total inventory cost is a function of m , which is a discrete variable.
For a given positive integer m 1,2,3,... , etc, the value of total inventory cost is obtained
and a list of total cost values is prepared. #he minimum in the above list would be the
optimum total inventory cost and the corresponding m would be the number of equal
length replenishment cycles during the planning horizon.
,).)*) 6umerical e7am&le
!n eample is presented here to illustrate the application of the model developed
in the previous section. !ssume that, an inventory system with a linear increasing trend
in demand, has the following properties3
a 10 , b 1 , ( 10 , r 0.15 , " 0.5 , % 2 , & 0.32 , p 1
#he different values of m are substituted in equation >9.79? and the corresponding
TC (m)
values are obtained. #he results obtained are presented in table 9.6 and figure
9.9
#able 9.6. #otal cost vs. number of cycles
m TC (m)
in I
6 7=@.B<
7 6==.@:;
9 6<@.<8=
; 6;7.=:8
< 69;.<:8
8 67B.;6;
@ 678.6;=
= 67;.:6B
B 677.8;@
6: 676.=:;
66 676.9;8
67 676.6@<
*
69 676.778
6; 676.;<6
Ominimum value.
#
o
t
a
l

I
n
v
e
n
t
o
r
y

)
o
s
t
#otal )ost vs. Number of )ycles
9<:
9::
7<:
7::
6<:
6::
<:
:
: < 6: 6<
Number of 1eplenishment )ycles
Figure 9.93 #otal Inventory )ost and Number of 1eplenishment )ycles
#able 9.7 compares the optimal number of replenishment cycles and the
corresponding total inventory cost for the system used in the above eample with
values of 1 between -6:P and 6::P.
#he results show that as the inflation rate increases and the difference between
inflation rate and discount rate grows, the total cost of inventory system increases
significantly, which is epected.
#able 9.7.
5ptimal Inventory )ost and Number of 1eplenishment )ycles /ifferent values of R
R " r >P? No. of 1eplenishment #otal Inventory
)ycles )ost
-6: 67 676.6@<
< 6: 7<7.B:@
6: 6: 99<.8@=
9: @ 676<.B9
<: 8 <7<9.@8
=: < <@=77.:
6:: 8 96B;<B
,)8) +conomic order 9uantity under conditions of &ermissible delay in
&ayment
In deriving the Economic 5rder Auantity formula, it is assumed that the supplier
is paid for the items delivered as soon as the items are received. $owever, in practice
suppliers may allow for a certain fied period for settling the account and paying the
amount owed to them for the items supplied. #ypically there is no etra charge if the
outstanding amount is settled within the allowed grace period. +eyond this period an
interest will be charged.
-hen a supplier allows for a fied period for settling the account, he is actually
giving his customer a loan without interest during this period. #he customer should be
able to bring in revenue and also earn interest during the grace period, and logically the
customer should delay settling the replenishment account until the end of the
permissible grace period.
#he mathematical formulation of the inventory model is presented based on the
wor% of '. 0. (oyal J6B=<K.
,)8)() Mathematical $ormulation
First the assumptions and additional notations needed for model are introduced.
Notations 3
I
c
C interest charges per I investment in stoc% per year, IEIEyr
I
d
C interest that can be earned per I in a year, IEIEyr
% C 5rdering cost for one order, I
t C permissible delay in settling the account , yr
* (t) C total annual cost, I
!ssumptions 3
6. #he demand for the item is constant
7. 'hortages are not allowed
9. /uring the grace period, the revenue earned by sales is deposited in an account,
which brings in interest, and at the end of the grace period the customer starts
paying for the interest charges on the items in stoc%.
;. 2lanning horizon is infinite
#he annual variable cost of the system has four different components3
>6? 5rdering or setup cost which is equal to
%
.
T
>7? Inventory )arrying )ost >ecluding interest charges?. !s it can be seen
from Figure >9.;?, the average inventory is equal to
DT
, so the stoc%
2
holding cost per year is DTh / 2 .
>9? )ost of interest charges for the items %ept in the stoc%. #he sales revenue
during the grace period is used to earn interest, but after the end of fied
period the items still in the stoc% have to be financed at the interest rate
I
c
. #he inventory level at the time of settling the account is equal to
D(T t) , and the interest is payable during the time (T t ) .
c
Inventory
4evel
/#
/ >#-t?
/t
t >#-t? Time
#
Figure 9.;3 the inventory diagram when
T t
Interest payable in one cycle is equal to3
Dp(T t)
2
I
2
and the interest payable per year for the inventory system is equal to3
d

c
2
2
2

Dp(T t)
2
I

DpTI
c
+
Dpt I
c
DptI
2T 2 2T
c
>;? Interest earned during the grace period3
#he maimum amount of sale during the grace period is equal to Dtp if
T t , and if T < t , the maimum sale will be equal to DTp .
#he interests earned during the grace period for the two cases are as
follows3
>a? T t , 'ee Figure 9.;.
Interest earned in one cycle is equal to 3
Dpt I
d
2
Interest earned during one year is equal to 3
Dpt I
d
2T
>b? T < t , 'ee Figure 9.<.
In this case the interest earned during one cycle is equal to 3
DT
2
p
[
2
+ DTp(t T )]I
d
T
DTpI
d
(t )
2
and the interest earned in one year would be equal to 3
DpI

t
T _

2
,
5bviously the interest earned is subtracted from the total cost in order to obtain the net
total cost per year.
Inventory
4evel
>t-#?
/#
# #ime
t
Figure 9.<3 #he inventory diagram when T < t
In the net section the mathematical models to obtain the optimal order quantity for
the inventory systems, based on (oyal J6B=<K, are presented.
,)8)*) +conomic 'rder 9uantity hen T t
#he total variable cost for the system when T t , is given as3
% DTh
DTpI Dpt
2
I
Dpt
2
I
* (T ) + +
c
+
c

DptI

d
T 2 2 2T
c
2T
!fter some simplifications, one can obtain the annual cost as 3
1
1
2
c
c
* (T )
(2 % +
Dpt
(I
c
I
d
))
+
DT
(h +
pI
) DptI . >9.7;?
2T 2
c c
#o minimize the cost,
d* (T )

0 , and obtain T

T
*
dT
1
T
*

(2 % + Dpt
2
(I
D(h + pI
c
I
d
)
)
>9.7<?
It should be mentioned that
t > 0 and (I
c
I
d
) 0 .
$ence the economic order quantity is equal to3
D(2 % + Dpt
2
(I
I )
$(T
*
) DT
*

c d
>9.78?
1 1
(h + pI )
and the minimum annual cost is equal to3
* (T
*
)
D(2 % + Dpt
2
(I
c
I
d
))(h + pI
c
)
DptI
c
>9.7@?
!s (oyal states, the economic order quantity obtained under the condition of
permissible delay in payments is generally higher than the order quantity given by a
classical E5A model.
If I
d
I
c
, the economic order quantity given by >9.78? is equal to the order
quantity obtained by classical E5A model.
It rarely happens that
I
c
I
d
0 , and in this case the economic order quantity
calculated by >9.78? would be lower than the order quantity obtained by classical E5A
model.
d

2
1
1 1

,)8),) +conomic 'rder 4uantity hen T < t

In this case, there wouldnLt be any item %ept in the stoc% after the fied
period, t , hence there is no interest charge paid in this case.
#he total annual cost can be represented as follows 3
* (T )

%
+
T
DTh
2
DpI

t
T _

2
,
or
* (T )
%
+
T
DT
(h + pI
d
) DptI
d
2
>9.7=?
#o obtain the minimal annual cost,
d* (T )
0
dT
$ence the optimum interval between two successive orders is given by3
T
*

2 %
D(h + pI
d
)
>9.7B?
and the economic order quantity is given by3
$(T
*
) DT
*

2
%D
>9.9:?
2 2
(h + pI
d
)
(oyal points out that at T t , the total annual variable cost can be obtained
by substituting T t , in >9.7;? or >9.7=?3
* (T )
%
+
Dth

DptI
d
>9.96?
t 2 2
(oyal eplains the optimal operating policy as follows3
'tep 63 /etermine T
*
from >9.7<?. If T
*
t , obtain * (T
* )
from >9.7;?
1
2
1 2
'tep 73 /etermine T
*
from >9.7B?. If T
*
< t , obtain
* (T
*
)
from >9.7=?
2 2 2
'tep 93 If T
*
< t
and T
*
t , then evaluate * (t) from >9.96?.
'tep ;3 )ompare
* (T
*
)
,
* (T
*
)
and * (t) ) 'elect the replenishment interval and the
order quantity associated with the least annual cost value evaluated in steps 6, 7 and 9.
"ha&ter $our
Inventory Models ith Alloable Shorta!es under inflationary conditions and
Permissible Delay in Payment
!s mentioned earlier, it is usual nowadays to see that customers are allowed a
grace period before settling their accounts with the supplier and paying for the goods
bought. In chapter three, a classical E5A model was presented which considers the
conditions of permissible delay in payment, based on the wor% of '. 0. (oyal J6B=<K.
In this chapter, a single item inventory model with allowable shortages is
considered with permissible delay in payment. 'hortages are important, particularly in a
model that considers a delay in payment. 'hortages can affect the quantity ordered to
benefit from the delay in payment. #his chapter is concerned with determining an
optimal ordering policy for a deterministic single item model with allowable shortages
and a permissible delay in payment, with and without considering the effects of inflation
and time value of money. In the net section the first model disregarding the discount
and inflation rate is presented and the mathematical formulation of the inventory
system is developed. #he effects of discount and inflation rate will then be considered
in the following section.
-)() A Sin!le Item Inventory Model 2+'43 ith Shorta!es and Permissible
Delay in Payment
#he assumptions and notations used in the model are presented.
!ssumptions of the model3
>6? #he demand for the commodity is %nown with certainty and constant.
>7? 'hortages are allowed and are fully bac%logged and satisfied when
replenishment orders are received.
>9? #he revenue earned during the grace period is invested, hence brings in interest.
>;? !fter the grace period, if there are still items %ept in the stoc% >i.e. unpaid?, the
customer is charged a penalty >high interest? on the unpaid value.
><? 2lanning horizon is infinite.
>8? 'i cost component are considered3
>a? 'etup cost
>b? $olding cost
>c? 2urchasing cost
>d? 'hortage cost
>e? )ost of interest charges for the items %ept in the stoc% after the grace period
>f? Interest earned during the grace period
#he following notations are used throughout the chapter3
D C demand rate, unitsEyear
p C purchase cost, IEunit
% C setup cost, IEorder
h C inventory cost, IEunitEyr
& C annual inventory carrying cost. IEI.yr
C shortage cost per unit short per unit time, IEunitEyr
T C replenishment period, yr
T
1
C length of period with positive stoc% of the items, yr
I
c
C interest charges per I value in stoc% per year, IEI.yr
I
e
C interest that can be earned per I in a year, IEI.yr
" C inflation rate, E yr
r C discount rate representing the time value of money, E yr
R C r "
C number of replenishment cycles during the planning horizon >when it is
finite?
# C permissible delay in settling the account , yr
(C C present value of the inventory carrying cost for the first cycle, I
TC C present value of the total cost for the first cycle, I
PC
1
C present value of the first part of the purchasing cost in the first cycle, I
PC
2
C present value of the second part of the purchasing cost in the first cycle, I
PC C present value of the purchasing cost in the first cycle which is equal to
PC
1
H PC
2
+C C present value of the shortage cost during the first cycle, I
IC C present value of the cost of interest charges during the first cycle, I
,I C interest earned during the first cycle, I
TCP C present value of the total annual cost, I
2
In the net section the mathematical presentations of the different cost
components of the inventory system are developed. #he ob"ective of the proposed
model is to determine the length of the inventory replenishment cycle, T , and the
length of the period with positive stoc% of items, T
1
, such that the total cost of the
system is minimized. #here are two distinct cases in this type of inventory system3
)ase 63 2ayment >end of the grace period? is before the total consumption of inventory
> # T
1
?.
)ase 73 2ayment is after consumption of the inventory, > T
1
< # ?.
-)()() Payment before the Total De&letion of Inventory
#his is the case where the grace period epires before the total consumption of
the inventory >Figure ;.6?. #he total cost of the system is comprised of the si
aforementioned cost components.
>a? 'etup cost
'etup cost is paid at the beginning of each replenishment cycle and is equal to % .
>b? $olding cost
#he average inventory during the period : to T
1
is equal to DT
1
/ 2 , and the stoc%
holding cost per cycle is equal to
(C
hDT
1
2
Inventory 4evel
: # T
1
T #ime
Figure ;.63 #he single item inventory system with shortages and permissible
delay in payment
>c? 2urchasing cost
#otal purchasing cost during the first period is given by3
PC pD# + pD(T T
1
) H pD(T
1
# ) C pDT
>d? 'hortage )ost
#he inventory system bac%logs the shortages incurred from the end of positive
inventory level period until the end of replenishment cycle, t [T
1
,T ] .
#he shortage cost per cycle is given by,
D(T T
1
)
2
+C
2
>e? )ost of interest charges for items %ept in stoc% after the grace period
!n interest is charged for the goods %ept after the grace period, interest rate
I
c
,
the amount of interest charges is equal to3
1 1
1
2
CI
DpI
c
(T # )
2
.
2
>f? Interest earned during the grace period
#he customer earns interest during the grace period and the amount earned is
as follows3
,I I
e
Dp(T T
1
+
#
)#
2
#he total inventory cost of the system during one cycle is given by3
TC % + (C + PC + +C + CI ,I
>;.6?
#o obtain the average annual cost, TCP , the total cost per cycle should be
multiplied by the number of replenishment cycles per year,
1
. $ence the total
T
annual cost is given by
TCP
% hDT
2
D(T T )
2
H H pD H H
DpI
c
(T
1
# )
-
T 2T 2T
I
e
Dp(T
T
1
T
2T
+
#
)#
2
>;.7?
#he total annual cost of the inventory system is a nonlinear function of two
independent variables3 >6? the length of replenishment cycle, T , and >7? the length of
the period with positive stoc% of the items, T
1
.
#he optimal policy of the inventory system may be obtained by solving the
following equations3
TCP
0
T

2
TCP
0
T
1
It is not easy to obtain closed forms for T and T
1
from the above equations.
Instead the values may be obtained by trial and error. !s an alternative, a non-
derivative nonlinear optimization technique %nown as $oo%e and &eeves method will be
used to obtain the optimal policy of the system using equation >;.7?. /etails of the
technique are given in !ppendi !. ! numerical eample is provided in section ;.9.
-)()*) Payment after the Total De&letion of Inventory
#his section considers the case in which the grace period epires after the
complete depletion of the inventory >Figure ;.7.?. !s a result, the interest charges for
inventory %ept after the grace period are reduced to zero, because the supplier is paid
in full at the time, # . #he interest earned per cycle is the interest earned during the
positive inventory period plus the interest earned during the time period
(T
1
, # ) .
#he total cost of the system includes >a? setup cost, >b? holding cost, >c?
shortage cost, >d? purchasing cost and >e? interest earned during the grace period. #he
first four cost components are the same as the cost components of the previous case.
#he interest earned during the grace period is determined as follows3


DpT
1
2
_
,I I
e

+ DpT
1
(# T
1
) + Dp# (T T
1
)

,
1 1

1

2
Inventory 4evel
#
T
0
T
1
T
#ime
Figure ;.73 #he single item inventory system with shortages and permissible delay in
payment, when the grace period epires after depletion of the inventory.
#he annual total cost of the system is obtained by multiplying the total cost of
the system per cycle by the number of replenishment cycles per year and is given by3
TCP
% hDT
2
pDT D(T T )
2
H H H -
T 2T T 2T

DpT
2
_
I
e

+ DpT
1
(# T
1
) + Dp# (T T
1
)

T
,
>;.9?
!s mentioned in the previous section, the optimal ordering policy of the system
may be obtained by solving the following equations3
TCP
0
T
TCP
0
T
1
'ince no closed form equation could be obtained for the optimal values of T and
T
1
using the above equations, one may have to solve these equations by trial and error.
Instead a derivative free method, $oo%e and &eeves !lgorithm, is used to obtain the
optimal policy of the system.
In the net section the effects of inflation and time value of money are
considered on the single item deterministic system with shortages and permissible
delay in payment.
-)*) +ffects of Inflation and Time Value of Money on a Sin!le Item Inventory
Model 2+'43 ith Shorta!es and Permissible Delay in Payment
In this section the effects of inflation and time value of money on the inventory
models discussed in the previous sections are investigated. It is assumed that all cost
components of the inventory system are sub"ect to a constant inflation rate and both
inflation and discounting are sub"ect to continuous compounding.
#he present value of the total cost of the inventory system is developed first,
then the optimal lengths of the replenishment cycle and the period with positive stoc%
are obtained. ! finite planning horizon is considered in this case.
!gain, two distinct scenarios can be recognized3 >6? the case in which the grace
period epires before the total consumption of inventory > # T
1
?, and >7? the case in
which the payment is due after the depletion of the inventory, > T
1
# ?.
#he appropriate mathematical models are developed for both cases and presented
below.
-)*)() :race Period +7&ires before the Total De&letion of Inventory
In this case, the total cost of the inventory system is comprised of si
components3 setup cost, purchasing cost, inventory carrying cost, shortage cost,
interest payable and interest earned during the grace period. 2resent valuess of the
individual costs at the beginning of each cycle are presented below >Figure ;.6?.
>a? 'etup )ost
'etup cost is paid at the beginning of the cycle and is equal to % .
>b? $olding )ost
#he holding cost is continuously paid for the amount of inventory %ept during the
period in which the stoc% level is positive. #he amount of inventory in hand at time t
is 3
I (t) C D(T
1
t
)
0 t T
1
#he holding cost in the first cycle is given by,
(C C
T
1
&p

I (t )e

Rt
dt
0
>;.;?
T
1
C &pD

(T
1
t)e

Rt
dt
0
C &pD
e
RT
1
+ RT
1
1
R
2
>c? 2urchasing )ost
#he purchasing cost is comprised of two components, the first one is paid at the
end of the grace period, # , and covers the items that satisfy the demand during
the current period up to # , and the amount of bac%log from the preceding period.
#he second component is paid continuously after the first payment until the end of
period with positive inventory, T
1
.
#he epression for the present value of the first component of the payment is
given by3
PC
1
C
pD#e
R#
+ pD(T T
1
)e

R#
>;.<?
#he present worth of the second component of the purchasing cost, which is
paid continuously after the epiration of grace period, is given by3
PC
2
C
T
1
pD

e

Rt
dt
#
>;.8?
C
pD
(e
R#
R
e
RT
1
)
#he present worth of purchasing cost during the first cycle is given by3
PC PC
1
+ PC
2
>;.@?
>d? 'hortage )ost
#he shortage cost occurs if a demand for the commodity ta%es place when it is
out of stoc%. #he amount of shortage at time t is equal to3
B(t) C D(t T
1
)
T
1
t T
#he mathematical epression for the present value of the shortage cost during the
first cycle is given by3
T
+C C

D(t T
1
)e

Rt
dt
T
1
>;.=?

e
RT
1
+
e
RT
(R(

T
+
T )

1)
_

1
C D

R
,
>e? )ost of Interest )harges
#he epression for the present worth of the interest charges for the value of
goods paid after the grace period is as follows3
IC C
T
1
pI
c
D

(t # )e

Rt
dt
#
>;.B?

e
R#
+
e
RT
1
(#R

RT
1)
_

1
C pI
c
D

R
,
>f? Interest Earned /uring the (race 2eriod
#he customer pays no interest during the grace period. Instead he accumulates
revenues on the sale or use of the products, and earn interest on the revenue. #he
present value of the interest earned during the first replenishment cycle is given by3
,I C # #
I
e
pD

#
e

d
dt H I
e
pD(T T
1
)

e

Rt
dt
>;.6:?
0 t 0

1 e
R#
R
2
#e
R#
_

1 e
R#
_

C DI
e
p#

2


H I
e
Dp

R

(T T
1
)

R
, ,
#he present value of the total inventory cost during the first cycle is the
summation of aforementioned costs3
TC
1
% + (C + PC + +C + IC
,I
>;.66?
#he effects of inflation and time value of money eist in each cycle of
replenishment. In order to develop an appropriate model for the system, a finite time
horizon ( is considered. #he number of replenishment cycles during this time horizon
would be

(
. Equation >;.66? gives the total cost at the beginning of each cycle. If
T
TCP is considered as the present value of the total cost of the inventory system over
the planning horizon, then 3
TCP C TC
1
(1 + exp(RT ) + exp(2RT ) + ... + exp(( 1)
RT ))
>;.67?
1 exp( RT ) _
C TC
1

1 exp(RT )
,

e
RT
1
RT + 1 pD

1
R#
R#
R#
RT1
C

% + &pD
2

R
H pD#e + pD(T T
1
)
e
H (e
R
e )

e
RT
1
+
e
RT
(R(T

T )

1)
_

e

R#
+
e
RT
1
(#R

RT
1)
_

1

1

H pD

2


H pI
c
D

R
,
R
,

R#

R
2
#e

R#
_
(T T )
_
1 exp( RT ) _
- DI
e
p#
2

R

- I
e ,
Dp(1 e

R#
)
1

R
,

1
exp(R
T )
,
#wo different cases can be considered, >i?
r > " >i.e. R r " > 0 ?, and >ii?
r < " >i.e. R r " < 0 ?. Epressions of the total cost of the system for both
aforementioned cases are presented below.
>i? r > " > R r " > 0 ?
If r > " , when ( , and the total cost given by >;.67? will converge
to3
1 _
TCP C TC
1

>;.69?

1 exp(RT )
,
>ii? r < " > R r " < 0 ?
In this case, when , the total cost given by >;.67? will not be
convergent. Instead, a finite horizon would be considered. If one ta%es ( 1
year, then

1
T
and the present worth of the total cost would be3

TCP C TC
1

1 exp(R) _
>;.6;?

1 exp(RT )
,
#he values of T and T
1
, which minimize TCP , may be obtained by
simultaneously solving
TCP(T , T
1
) / T 0 ,
and
TCP(T ,T
1
) / T
1
0 . 'ince
TCP(T ,T
1
)
involves a complicated eponential function, it is not easy to calculate
the partial derivations of the total cost function and obtain closed form epressions
for the optimal values of T ,T
1
from the partial derivatives of the total cost function.
$ence the direct search method of $oo%e and &eeves is used again to obtain the
optimal ordering policy and minimum annual cost of the system. ! numerical
eample is provided in section ;.; to illustrate the procedure used and the general
behavior of the system with respect to different parameters of the system.
In the net section the case is considered in which the grace period epires after
the complete depletion of the inventory, T
1
< # .
-)*)*) :race Period +7&ires after the Total De&letion of Inventory
#he amount of order in each period can be chosen such that the inventory in
hand is consumed before the fied grace period epires. In this section, the
mathematical epressions, for the cost components of the system are developed.
#he inventory system in this situation includes five cost components, >a? setup cost,
>b? holding cost >c? purchasing cost, >d? shortage cost, and >e? interest earned
during the grace period. In this case, the customer pays for the entire order at the
end of the grace period, and no interest is charged for late payment. 'etup, holding,
and shortage costs are the same as in last case, but purchasing cost and interest
earned during the grace period will change in the new situation. #he mathematical
epressions for these cost components are developed below.
>i? 2urchasing )ost
#he total amount of purchasing cost in each replenishment cycle is paid at the
end of the grace period. $ence the present worth of the purchasing cost in the first
cycle is given by3
PC
1
pDTe

R#
>;.6<?

R R
>ii? Interest Earned /uring the (race 2eriod
T
1
# #
,I
1
C
I
e
pD

e

d dt H I
e
pD(T T
1
)

e

Rt
dt
>;.68?
0 t 0

1

e

RT
1

RT e
R#
_

1 e
R#
_

1
C I
e
pD

R
T
1
H I
e
Dp

,

(T T
1
)
,
#he present value of the total inventory cost during the first cycle is the summation
of >6? 'etup cost, >7? Inventory holding cost, >9? 2urchasing cost, >;? shortage cost,
and ><? Interest earned during the grace period.
TC % + (C + PC + +C ,I >;.6@?
For a planning horizon of ( with a number of cycles

(
, the present value
T
of the total cost of the system is given by3
TCP TC
(1 + exp(
RT ) + exp(2RT ) + ... + exp((
1)RT ))
TC
1 e
RT
1 e
RT
'imilar to the previous section, two cases are considered, >i? R > 0
1 _
for which TCP TC
1 e
, when
RT
,
( and >ii? R < 0 for which

1 e
R
_
TCP
TC

, when ( 1 .

1 e
RT
,
!s in the previous models, in order to obtain the optimal ordering policy of the
system, the following equations should be simultaneously solved.
TCP(T , T
1
) / T 0 , and
TCP(T ,T
1
) / T
1
0
$owever, the total cost function of the system involves complicated eponential
functions which ma%es the tas% of obtaining closed form solutions for the optimal values
of T ,T
1
very difficult, if not impossible. $ence the algorithm of $oo%e and &eeves is
used to obtain the optimal policy and minimum cost of the system.
-),) '&timal Solution for Inventory Model 2+'43 ith Shorta!es and
Permissible Delay in Payment
#he main ob"ective of the models considered in this research is.to minimize the
annual cost which is a nonlinear function of the length of the replenishment cycle and
the length of the period with positive inventory, T ,T
1
. 'ince it is very hard, if not
impossible, to derive closed formulae for the decision variables, some numerical
analysis method or alternatively a nonlinear optimization technique needed to be used.
#he direct search method of $oo%e and &eeves was applied to obtain the optimal
solutions. #he general idea of the search method is to change the value of one variable
at a time while %eeping the other>s? constant >eploratory search?. +y using the
information obtained at the eploratory search stage, a direction is defined to move
towards the minimum point >pattern search?. ! detailed eplanation of the $oo%e and
&eeves search algorithm is provided in !ppendi !. ! computer software was
developed to facilitate the implementation of the optimization algorithm. #he code is
given in !ppendi +. #he application is implemented using ,icrosoft *isual +asic 8. !s
,icrosoft premium programming language, *isual +asic 8 is the most popular 5b"ect
5riented 2rogramming approach for application development. It utilizes powerful
-indows
features as the front end. 5ne instance of the front end of the application is presented
in figure ;.;.
!n eample is presented here to illustrate the application of the models
developed with the following data3
D C 7:: unitsEyr, % C 6: IEorder, p C < IEunit, & C :.9 IEIEyr, C < IEunitEyr,
I
c
C :.6< IEIEyr, I
e
C :.6 IEIEyr, # C J:.:6, :.:9, :.:<, :.:@, :.6, :.67, :.6;K year.
#he behavior of the optimal policy of the system with respect to the length of
permissible delay in payment is investigated with the help of appropriate graphs and
discussions. 1elationship between the total annual cost of the system, the length of the
replenishment cycle, and length of the period with positive stoc% and different lengths
of grace period are also investigated.
Figure ;.;3 Eample of )omputer !pplication implementing $oo%e and &eeves
!lgorithm
#he optimal ordering policies obtained for different values of # for the above
eample are reported in #able ;.6.
/
e
&
l
e
n
i
s
h
m
e
n
t

"
y
c
l
e
2
y
r
3
Grace Period
,Year
Replenishment Cycle
(T), Year
Period With Positive
Stock (T1), Year
Optimal Order
(), !nits
Present
val"e o#
Total Cost,
:.:6 :.7<9== :.6@<;9 <:.@@8 I6:@@.;;B
:.:9 :.7<;6@ :.6@896 <:.=9; I6:@;.=9<
:.:< :.7<;8= :.6@@<9 <:.B98 I6:@7.7B@
:.:@ :.7<<9@ :.6@=<8 <6.:@; I6:8B.=9<
:.6: :.7<8B :.6=:8; <6.9= I6:88.7=6
:.67 :.7<=79 :.6=777 <6.8;8 I6:8;.::9
:.6; :.7<B=< :.6=;:7 <6.B@ I6:86.@B=
#able ;.6. 5ptimal 5rdering 2olicies for /ifferent *alues of ,.
!s shown in figure ;.<., the length of the replenishment cycle increases as the
length of grace period increases.
:. 7 8 6
:. 7 8
:. 7 < B
:. 7 < =
:. 7 < @
:. 7 < 8
:. 7 < <
:. 7 < ;
:. 7 < 9
: : .: < : .6 : . 6
<
1 ep le n is hm ent
)y c le
#en! t h o f : r a c e P e r i od 2 y r 3
Figure ;.<3 +ehavior of #, with respect to length of grace period.
#he length of period in which the inventory level is positive also increases as the
length of grace period increases, as shown in figure ;.8.
T
o
t
a
l

"
o
s
t

o
f

t
h
e
S
y
s
t
e
m
,

;
P
e
r
i
o
d

i
t
h

P
o
s
i
t
i
v
e

S
t
o
c
k
#
e
v
e
l

2
y
r
3
: . 6 = <
: . 6 = ;
: . 6 = 9
: . 6 = 7
: . 6 = 6
: . 6 =
: . 6 @ B
: . 6 @ =
: . 6 @ @
: . 6 @ 8
: . 6 @ <
: : . : < : . 6 : . 6 <
# e n ! t h o f : r a c e P e r i o d 2 y r 3
4 e n g t h o f
2 e r i o d w i t h
2 o s i t i v e
' t o c %
4 e v e l
Figure ;.83 +ehavior of T
1
, with respect to length of grace period.
#he total cost of the system decreases as the length of grace period
increases, as shown in figure ;.@.
6 : = :
6 : @ <
6 : @ :
# o t a l ) o s
t
6 : 8 <
6 : 8 :
: : . : < : . 6 : . 6 <
P e r m i s s i b l e D e l a y i n P a y m e n t 2 y r 3
Figure ;.@3 +ehavior of the total cost of the system with respect to length of the grace
period
#he total cost function of the system for the above eample is represented in
figure ;.=.,which shows the behavior of the total cost with respect to two decision
variables, 1eplenishment )ycle, T , and 2ositive 'toc% 4evel 2eriod T
1
. #he
figure
shows that the total cost function forms a conve surface.
TCP
1090
1085
0.2
1080
0.2
0.22
0.24
T
0.26
0.28
0.3
0.16
0.18
T1
Figure ;.=3 #otal )ost Function of the system TCP(T ,T
1
)
-)-) '&timal Solution for Inventory Model 2+'43 ith Shorta!es, Inflation, Time
Value of Money and Permissible Delay in Payment
In this section a numerical eample is provided to illustrate the application of the
inventory model developed in section ;.7. #he sensitivity of the present value of the
total cost of the system to the system parameters is also eamined. #he same
parameter values used in the previous eample are considered.
#he present value of the total cost of the inventory system is a function of two
continuous variable, length of replenishment cycle, T , and length of the period with
positive inventory level, T
1
. !s in the previous eample, the direct search procedure of
$oo%e and &eeves was applied to obtain the optimal policy of the system. /etailed
eplanations about the algorithm are given in !ppendi !. #he algorithm was
implemented using *isual basic 8. #he code is given in !ppendi ). !n instance of
the application developed is shown in figure ;.B.
#he application developed allows the user to change different parameters of the
system and obtain the optimal ordering policy and minimum cost.
Figure ;.B3 Eample of the )omputer !pplication developed to implement the $oo%e
and &eeves 'earch ,ethod
#he inventory parameters are as follows3
D C 7:: unitsEyr, % C 6: IEorder, p C < IEunit, & C :.9 IEIEyr, C < IEunitEyr, I
c
C
:.6< IEIEyr,
I
e
C :.6 IEIEyr. /ifferent values of # and R r " , are considered.
'everal tables are provided which show the length of replenishment cycles, length of
period with positive stoc%, optimal ordering quantities and present values of minimal
annual cost for # CJ:.:6, :.:9, :.:<, :.:@, :.6:, :.67, :.6<K, and R CJ:.6, :.:<, - :.6,
- :.9K.
'everal figures are also provided to illustrate the relationship between the
decision variables and the present value of the total cost of the system for different
parameter values. #he values of the length of the replenishment cycle, length of period
with positive stoc%, and present value of the total cost of the system are given for
different values of grace period at specific values of inflation rate. !dditional graphs are
provided to illustrate the relationship between the decision variables and the present
value of the total cost of the system for different values of inflation rate.
#able ;.7. shows the optimal policy of the system when R 0.1 .
$elay in Payment
(%),Year
Replenishment
Cycle (T), Year
Period With
Positive Stock
(T1), Year
Optimal Order
(), !nits
Total Cost, &
:.:6 :.7<@8; :.7:;6< <:.B<7:: 6:7<.:=<:
:.:9 :.7<87; :.7::=: <6.9;=:: 6:79.6;<:
:.:< :.7<9<7 :.6B86@ <:.BB=:: 6:7:.@BB:
:.:@ :.7;B79 :.6B:99 <:.;<7:: 6:6=.:98:
:.6: :.7;:69 :.6@B<7 ;B.988:: 6:69.:8;:
:.67 :.796@= :.6@:9; ;=.9=;:: 6::B.6<::
:.6< :.768:; :.6<;76 ;8.<=8:: 6::7.7<6:
#able ;.7. 5ptimal 2olicy of the 'ystem when 1C :.6.
Figure ;.6:. shows that the length of replenishment cycle generally decreases
as the length of grace period increases.
/
e
&
l
e
n
i
s
h
m
e
n
t
"
y
c
l
e

2
y
r
3
P
e
r
i
o
d

i
t
h
P
o
s
i
t
i
v
e

S
t
c
k
#
e
v
e
l
1C : .6
:
:.78 :::
:.7< :::
:.7; :::
:.79 :::
:.:: :.6: :.7:
: r ace P e r io d 2 y
r 3
4e ngth of
1 eple n is h m en t
)y c le s
Figure ;.6:3 1elationship between T and the length of grace period when 1C :.6.
In figure ;.66. the length of the period with positive stoc% level decreases as the
length of grace period increases.
1C :.6:
:.9::::
:.7::::
:.6::::
:.:::::
:.:: :.:< :.6: :.6< :.7:
:race Period 2yr3
2eriod with 2ositive
'toc% 4evel
Figure ;.663 1elationship between T
1
and the length of the grace period when 1C :.6:
Figure ;.67. shows that the present value of the total cost of the inventory
system decreases as the length of the grace period increases.
T
o
t
a
l

"
o
s
t

2
;
3
1C :.6:
6:9:.:
6:7<.:
6:7:.:
6:6<.:
6:6:.:
6::<.:
:.:: :.:< :.6: :.6< :.7:
:race Period 2yr3
#otal )ost >I?
Figure ;.673 1elationship between the total )ost of the 'ystem and the 4ength of the
(race 2eriod.
Figure ;.69. illustrates the shape of the total cost function of the system when
,C :.:6 and 1 C :.6, which shows that the function ma%es a conve surface.
TCP
1090
1085
1080
0.2
0.22
0.24
T
0.26
0.28
0.3
0.16
0.18
T1
0.2
Figure ;.693 #otal )ost Function, , C :.:6, 1 C :.6:
/
e
&
l
e
n
i
s
h
m
e
n
t
"
y
c
l
e

2
y
r
3
#he analysis presented in #able ;.7 and Figures ;.6: through ;.67 is
repeated below in #ables ;.9 through ;.< and Figures ;.6; through ;.7: with the
values of 1
changed from :.6: to :.:<, -:.6: and -:.9:.
$elay in
Payment
(%),Year
Replenishment
Cycle (T), Year
Period With
Positive Stock
(T1), Year
Optimal Order
(), !nits
Total Cost, &
:.:6 :.7<9:9 :.6=8@7 <:.=6= 6:<6.7<=:
:.:9 :.7<9@: :.6=8<6 <:.@B= 6:;=.=66:
:.:< :.7<68= :.6=;9; <:.88 6:;8.69::
:.:@ :.7;B7B :.6=68= <:.<7 6:;9.7:=:
:.6 :.7;;@B :.6@@7; <:.6@= 6:9=.989:
:.67 :.7;:97 :.6@96B ;B.B77 6:9;.=6::
:.6< :.796=7 :.688:6 ;B.;:= 6:7=.B<B:
#able ;.93 5ptimal 5rdering 2olicy of the 'ystem, -hen 1 C :.:<.
1 C : .: <
:.7 < 8
:.7 < ;
:.7 < 7
:.7 <
:.7 ; =
:.7 ; 8
: : .6 : .7
: r ace P e r io d 2 y
r 3
1 eplenishm ent
) y c le >y r?
Figure ;.6;3 1elationship between # and the length of the grace period ,, when
1 C :.:<.
T
o
t
a
l

"
o
s
t

2
;
3
P
e
r
i
o
d

i
t
h
P
o
s
i
t
i
v
e

S
t
o
c
k
#
e
v
e
l
1 C : . : <
: . 6 B
: . 6 = <
: . 6 =
: . 6 @ <
: : . 6 : . 7
: r a c e P e r i o d 2 y
r 3
2 e r i o d w it h
2 o s i t i v e ' t o c %
4 e v e l > y r ?
Figure ;.6<3 1elationship between T
1
and the length of the grace period , -hen
1C :.:<.
1 C :.:<
6:<<
6:<:
6:;<
6:;:
6:9<
6:9:
: :.:< :.6 :.6< :.7
:race Period 2yr3
#otal )ost
Figure ;.683 relationship between the total cost of the system and , -hen 1C :.:<.
T
,

T
(

2
y
r
3
$elay in
Payment
(%),Year
Replenishmen
t Cycle (T),
Year
Period With
Positive Stock
(T1), Year
Optimal
Order (),
!nits
Total Cost,
&
:.:6 :.78:<; :.68:: <7.:@= 669:.97B:
:.:9 :.78696 :.689<6 <7.7;= 667@.97;:
:.:< :.787:8 :.688== <7.=9= 667;.;B8:
:.:@ :.78;7< :.6@698 <9.;<; 6676.=;;:
:.6: :.78@<< :.6@@B6 <;.=68 666=.6=@:
:.67 :.7@:79 :.6=78@ <<.B:7 666<.B<@:
:.6< :.7@8;9 :.6B6:9 <@.B:; 6667.B66:
#able ;.;. 5ptimal 2olicy of the 'ystem when 1 C - :.6:.
:. ;
:. 9
#
:. 7
#6
:. 6
:
:. :: :. :< :. 6: :. 6< :. 7:
M 2 y r
3
Figure ;.6@3 1elationship between #, and #6 the length of the grace period,
when 1 C - :.6:.
T
o
t
a
l

"
o
s
t

2
;
3
T
,

T
(

2
y
r
3
669<
669:
667<
667:
666<
:.:: :.:< :.6: :.6< :.7:
M 2yr3
#otal )ost
Figure ;.6=3 1elationship between the total cost of the system and ,,
when 1C - :.6:.
$elay in
Payment
(%),Year
Replenishment
Cycle (T), Year
Period With
Positive Stock
(T1), Year
Optimal
Order (),
!nits
Total Cost,
&
:.:6 :.7B=8; :.6<6;8 <B.@7= 67;:.6@B:
:.:9 :.9:6<: :.6<B:9 8:.;77 679@.=<9:
:.:< :.9:<9@ :.68@;8 86.;7; 6798.:;6:
:.:@ :.966B@ :.6@8=; 89.:6= 679;.8:@6=
:.6: :.97<;9 :.6B7B< 88.988 6799.8;=:
:.67 :.9986: :.7:;;@ 8B.7:= 6799.;7B:
:.6< :.9<;B< :.779<7 @9.@B 6799.=9B:
#able ;.<. #he optimal policy of the system when 1 C - :.9: and , varies
from :.:6 to :.6<.
:.;
:.9
#
:.7
#6
:.6
:
:.:: :.:< :.6: :.6< :.7:
M 2yr3
Figure ;.6B3 1elationship between #, #6 and ,, when 1C - :.9:
T
o
t
a
l

"
o
s
t

2
;
3
T
,

T
(

2
y
r
3
67;7
67;:
679=
6798
679;
:.:: :.:< :.6: :.6< :.7:
M 2yr3
#otal )ost
Figure ;.7:3 1elationship between the total cost and , -hen 1 C - :.9:.
#able ;.8 summerizes the optimal solutions and shows the effects of inflation and
discounting on the system for ,CJ:.:6, :.:@, :.6<K. then effects are also illustrated in
Figures ;.76 through ;.78, which show the relationship between #, #6, and the total
cost, and - 1 for different values of the grace period ,.
:.9<
:.9:
:.7<
:.7:
#
:.6<
#6
:.6:
:.:<
:.::
-:.7 -:.6 : :.6 :.7 :.9 :.;
1 /
Figure ;.763 1elationship between #, #6 and - 1 when ,C :.:6.
T
o
t
a
l

"
o
s
t

2
;
3
T
,

T
(

2
y
r
3
6;::.:
67::.:
6:::.:
=::.:
8::.:
to ta l ) o s t
;::.:
7::.:
:.:
- : .7 - : .6 : : .6 : .7 : .9 : .;
1/
Figure ;.773 1elationship between the total cost and - 1 when , C :.:6.
:.9<
:.9:
:.7<
:.7:
#
:.6<
#6
:.6:
:.:<
:.::
- : .7 - : .6 : : .6 :.7 : .9 :.;
1 /
Figure ;.793 1elationship between #, #6 and -1 -hen ,C:.:@.
T
o
t
a
l

"
o
s
t
T
,

T
(

2
y
r
3
6;::.:
67::.:
6:::.:
=::.:
8::.:
#otal ) ost
;::.:
7::.:
:.:
-:.7 -:.6 : :.6 :.7 :.9 :.;
1 /
Figure ;.7;3 1elationship between the total cost and - 1 when ,C :.:@
:.;:
:.9<
:.9:
:.7<
#
:.7:
#6
:.6<
:.6:
:.:<
:.::
-:.7 -:.6 : :.6 :.7 :.9 :.;
1 /
Figure ;.7<3 relationship between #, #6 and - 1 when ,C:.6<.
T
o
t
a
l

"
o
s
t
6;::.:
67::.:
6:::.:
=::.:
8::.:
#otal )ost
;::.:
7::.:
:.:
-:.7 -:.6 : :.6 :.7 :.9 :.;
1/
Figure ;.783 1elationship between the total cost of the system and - 1, when ,C:.6<
From the information provided in #ables ;.7 through ;.< and Figures ;.76
through ;.78, it can be observed that the total cost of the system increases when the
inflation rate increases. #he length of the replenishment cycle generally tends to
increase when the inflation rate increases. #he length of the period with positive stoc%
level has a tendency to increase when the length of grace period increases.
"ha&ter $ive
Dynamic Sin!le Item Inventory Models ith Inflationary "onditions and
Permissible Delay in Payment)
In this chapter inventory models are developed for products with time-varying
demand rate. In these models, customers are allowed a period of time to pay bac% for
the goods bought without paying interest and all costs are sub"ect to a uniform inflation
rate and discounting. ! case is considered in which the grace period granted by the
supplier, is a fraction of the replenishment cycle. ! mathematical representation of the
model will be developed and the optimal policy will be presented for the system.
!s stated in previous chapters, it is common these days to see that customers
are allowed a grace period to settle the account with the supplier and pay for the goods
bought within that period. )ustomers pay no interest during that period and can
postpone the payment till the end of the grace period, but after that period, if the
customer has not paid for the goods delivered, an interest will be charged.
(ranting a delay period in payment to the customer can be considered as a
demand stimulating activity performed by the supplier to encourage the customer to
buy more. $ence an appropriate pattern should be considered which properly presents
the demand during the planning horizon. In the model presented in this chapter the
demand rate is considered as a linear function of time.
!s mentioned earlier, inflation is a fundamental feature of todayDs economy all
over the world and large-scale inflation rates are not uncommon in many countries. 5n
the other hand, inventory represents a capital investment of a firm and must compete
with other assets for the firmDs limited funds. #herefore the effects of inflation and time
value of money are eplicitly considered in analyzing the inventory system in this
chapter.
#he length of replenishment cycles are considered to be equal and the purpose
of the model is to determine the length of the replenishment period such that the total
inventory cost is minimized.
#he remainder of this chapter is organized as follows3 In the net section the
assumptions underlying the model are presented and for more convenience, the
notations used throughout the research are reproduced and new terms are added. In
the third section the model is developed. #he fourth section provides numerical
eamples to illustrate the application of the model developed.
.)() Assum&tions and notations
#he following notations are used throughout the chapter.
h C unit inventory-carrying cost per unit per year, IEunit.yr
I
c
C interest charges per I investment in stoc% per year, IEI.yr
I
d
C interest that can be earned per I in a year, IEI.yr
p C unit purchasing price, I
% C 5rdering cost for one order, I
# C permissible delay in settling the account as a fraction of the
replenishment cycle
T C time interval between two consecutive orders, yr
( C length of planning horizon, yr.
r C interest rate, Eyr.
" C inflation rate, Eyr.
R r "
CR C total setup >ordering? cost during the planning horizon
C(
)
C present worth of holding cost during the "-th cycle
C( C present worth of total holding cost during the planning horizon
CP
)1
C present worth of purchasing for the "-th cycle, for goods demanded
during the grace period
CP
) 2
C present worth of the purchasing cost for the goods demanded
after the grace period for the "-th cycle
CP C present worth of the purchasing cost during the planning horizon.
CI
)
C present value of interest charges during the "-th period
CI C present value of the total interest charges for the items %ept after
the grace period during the planning horizon
C,
)
C present value of the interest earned during the "-th cycle
C, C present value of the total interest earned during the planning
horizon
TC C present worth of the total cost of the inventory system
!ssumptions
#he mathematical model of the inventory replenishment problem is based on the
following assumptions3
!ll the cost components of the inventory system are sub"ect to an inflation rate,
which is constant over the planning horizon.
Inflation and discounting are sub"ect to continuous compounding.
#he replenishment rate is infinite.
'hortages are not permitted.
4ead time is zero
#he system operates for a prescribed planning horizon, ( .
#he demand rate is a linear function of the time, i.e.,
D(t) a + bt, a, b > 0 and 0 t ( .
m replenishments are made during the entire time horizon, ( , and the length of
each replenishment cycle is equal to ( / m .
! constant fraction of each replenishment cycle, # , is considered to be the grace
period granted by the supplier.
Five cost components will be considered3
>a? 5rdering cost
>b? $olding cost
>c? 2urchasing cost
>d? )ost of interest charges for the items %ept in stoc% after the grace period
>e? Interest earned during the grace period.
5rdering cost is paid at the beginning of each replenishment period and is sub"ect to
the constant inflation rate effective during the planning horizon.
#he holding cost is paid through each replenishment cycle and is affected by
inflation rate.
#wo components will be considered for purchasing cost
>a? #he first part is instantaneously paid at the end of each grace period,
#
)
,
during the replenishment cycles, and covers the cost of goods
demanded during the grace period.
>b? #he second part is continuously paid after the grace period until the
beginning of net replenishment cycle, for the goods demanded after the
grace period up to the end of the replenishment cycle.
+oth components are sub"ect to constant inflation rate and discounting during
the planning horizon.
!n interest is charged for items %ept in the stoc% after the grace period.
!n interest is earned during the grace period.
.)*) Mathematical $ormulation of the Pro&osed Model
In this section the mathematical formulation of the inventory model is presented.
#he ob"ective of the proposed model is to determine the optimal length of the
replenishment cycles during the planning horizon, ( , which minimizes the present
worth of the total inventory cost.
In this model, there are m equal length replenishment cycles during the planning
horizon. -here m is un%nown and needs to be determined. 1eplenishments occur at
the beginning of each cycle,
t T
0
, T
1
, T
2
,..., T
m1
,
and
T
)

( * )
,
m
) 0,1,2,3,..., m . !s
illustrated in figure <.6.,the grace period for the )th cycle is a constant fraction of that
cycle and starts at the beginning of the cycle, t T
) 1
, and continues up to time
t #
)
. # represents the constant fraction of each cycle length during which the
customer can settle the account and pay for the goods bought, but after that an
interest will be charged. $ence #
)

( (# + ) 1)
,
m
) 1,2,3,..., m .
#he approach used in developing a model for this problem starts by determining
the epressions for the present value of the various costs involved in a cycle.
T
0
#
0
T
1
T
m 1
#
m 1
(#
T
m
(
m
( / m
( (# + m
1)
m
Figure <.6, Inventory level as a function of time for the proposed model

.)*)() 'rderin! cost
)ustomers pay the ordering cost at the beginning of each period and the cost is
sub"ect to a constant inflation rate. #he present value of the total replenishment cost
incurred during the entire time horizon, ( , is given by >the following mathematical
equations were developed using a software called ,athematica?3
%

(R
+

(R
_
(exp((R 1)
m 1 exp


CR %

exp(RT
)
) C

m
,
, ><.6?

+

(R
_
) 0
1 exp


m
,
.)*)*) <oldin! "ost
'hortage is not allowed, therefore an inventory is %ept during the entire cycle.
#he present worth of the holding cost during the "-th cycle > ) 1,2,3,..., m ? is as follows3
C(
)
C
() / m
&p

(t ( ( ) 1) / m)(a + bt )
exp(Rt)dt
( ( ) 1) /
m
><.7?

exp(
R( )
1)(
)(amR + b(2m + ( ( ) 1)R
_

C(
)
C &pD

mR
3

,

exp(
()R
)(amR(m + (R) + b(2m
2
+ ( (1 + ))mR + (
2
_
)R

&pD

m

m
2
R
3

,
$ence the present worth of the total holding cost during the entire time horizon ( , is
the following3
m
C(

C(
)
) 1
><.9?
=
exp(
(R
)
exp(
(R
) exp(
(R
)
D&p(
1
(exp(
(R
)(bm(
m
(exp(
(R
)
1+ m
(
m
+
m
mR
3
m
exp(
(R
)m
(1 exp(
(R
))
2
m
1 (exp(
(R
))
1+
m
m
(1 exp(
(R
))
2
m
exp(
(R
)
1
m
exp(
(R
)
+
m
))R + (1 +
m
)(2mb b(R + amR)))
1
(b( (
m
exp(
(R
)
1
m
1 exp(
(R
)
m
m
2
R
3
(1 exp(
(R
))
2
m
exp(
(R
)
exp(
(R
)(1 +
m)
1 exp((R
(R
)
(exp(
(R
)
1+ m
(
m
+
m
))R(m
+ (R) + (1 +
m
)
m
(1 exp(
(R
))
2
m
exp(
(R
)
1
m
1 exp(
(R
)
m
m(2bm + b(R + amR + a(R
2
)))
.)*),) Purchasin! "ost
#he purchasing cost is comprised of two components. In each cycle the customer
pays for the goods demanded during the grace period at the end of the grace period,
#
)
. #he following is the present worth of the purchasing cost for the "-th cycle
> ) 1,2,3,..., m ?, for the goods demanded during the grace period3
CP
)1
p
#
)

(a + bt)
exp(R#
)
)dt ,
><.;?
T
) 1

b exp(
( ( ) + # 1)R
)(
2
)#
b exp(
( ( ) + # 1)R
)(
2
#
2
C p
m
+ p
m
m
2
a exp(
( ( ) + # 1)R
)
(#
2m
2
b exp(
( ( ) + # 1)
)(
2
#
+ p
m

m
,
m
m
2
) 1,2,3,..., m
#he customer starts paying for the second portion of the purchasing cost after the
grace period and continues till the beginning of the net period for the goods
demanded during that time,
t [#
)
, T
)
] .
Following is the present value of the purchasing cost for the goods demanded after the
grace period for the "-th cycle3
CP
) 2
p
T
)

(a + bt ) exp(Rt)dt , ><.<?
#
)

exp(
( (1 +
))R
)(amR + b(m + ()R))
_
C p


m

mR
2

,

exp(
( ( ) + #
1)R
)(amR + b(m + ( ( ) + # 1)R))
_
+ p

m

, ) 1,2,3,..., m

mR
2

,
$ence the present value of the purchasing cost during the planning horizon is3
m
CP

(CP
)1
+ CP
) 2
) C ><.8?
) 1
exp(
(R

(#R
)(1 (exp(
(R
)
m
)
1 exp((R
(R
)
1
( p(2amR((1 +
m m m

m
)m
2m
2
R
2
exp(
(#R
)(1 exp(
(R
))
m
exp(
(R
)
1
m
1 exp(
(R
)
m
exp(
(R

(R#
)
+ (
m m
)(mR) + b(2( (
m m
(exp(
(R
)
1+ m
exp(
(R
)
1
m
(1 exp(
(R
))
2
m
m
exp(
(R

(
m
(#R
)
m
exp(
(R

+
m
(#R
m
)(1 +
m) ))R((exp(
(R
) exp(
(R#
))m + (#R) +
(1 exp(
(R
))
2
m
exp(
(R
) 1
m m
m
exp(
(R
)(1 exp(
(R
))
m
(
m m
)(2 exp(
(R
)(m
2
(mR + (m#R) 2 exp(
(R#
)m
2
2(
2
#R
2
exp(
(R
) 1
m m
m
+ (
2
#
2
R
2
)))).
.)*)-) "ost of interest char!es for the items ke&t in stock after the
!race &eriod
!s mentioned, if the customer does not pay the supplier by the end of the grace
period, he will owe interest to the supplier. #he items still in stoc% have to be financed
at the interest rate
I
c
. #he present value of interest charges during the "th period is
given by3

CI
)
pI
c
T
)

(a + bt)(t #
)
)
exp(Rt)dt
#
)
><.@?

exp(
( ( ) + #
1)R
)(amR + b(2m + ( ( ) + # 1)R))
_
pI
c

m

mR
3

,

exp(
()R
)(amR(m ( (# 1)R) + b(2m
2
+ ( (1 + ) # )mR (
2
)(# 1)R
2
))
_
pI
c

m

m
2
R
3

,
) 1,2,3,..., m 1
!nd the present value of the total interest charges for the items %ept after the
grace period is given by3
I
c
p(
1
mR
3
m
CI

CI
)
) 1
((exp((R / m)
m
1
(b(
(
exp((R / m)
(1 exp((R / m)
2
(exp((R / m)
1
+ m
(
><.=?
exp((R / m)
+
(1 exp((R / m))
2
exp((R / m)(1 + m)
)(2bm b(R + b(#R + amR)))

1 + exp((R / m)
1
m
2
R
3
(b( (
exp((R / m)
(1 exp((R / m)
2
(exp(
(R
)
1+
m
m
(
exp((R / m)
+
exp((R / m)(1 + m)
))R(m (R + (#R) + (1 +
1 exp((R (R / m)
)
(1 exp((R / m))
2
(1 + exp((R / m))
1 exp((R / m)
* m(2bm + b(R b(R# + a(R
2
a(#R
2
))).
.)*).) Interest earned durin! the !race &eriod
#he customer earns money during the grace period. -hen the supplier allows
the customer to pay for the goods bought after a fied period of time, he is in fact
giving him a loan without interest during that period. #he customer can en"oy this
privilege and continue to accumulate profit and earn interest during the credit period
with rate
I
e
.
#he present value of the interest earned during the "th cycle > ) 1,2,3,..., m ? is given
by3
C,
)
pI
e
#
)
#
)

(a + bt ) exp(R )
ddt
><.B?
T
) 1
t
pI
e
(
1
2m
2
R
3
(exp(
( (1 + ) + # )R
)(2amR(m
+ ( (

1
+ )
+ # )R
+ b(2m
2
+ 2( (

1
+ )
+ # )mR
m
+ (
2
(1 + ) + # )
2
R
2
)))
+
1
(b exp(
( (1 + ) + # )
)(
2
(1 + ))
2
R
2
+ 2a exp(
( (1 + ) + #
)R
)
2m
2
R
3
m m
* ( (1
+
))mR
2
+ 2 exp(
( (1
+
m
))R
)m(amR + b(m + ( (1
+
))R)))),
) 1,2,3,..., m
#herefore the present value of the total interest earned during the planning
horizon is given by3
m
C,

C,
)
) 1
><.6:?

I p(
1
exp(
(R
)
((exp(
(R
)
(# 1)
(b( (
m
(exp((R / m))
1+ m
(
exp((R / m)
e
mR
3
m
(1 exp((R / m))
2
(1 exp((R / m))
2
+
exp((R / m)(1 +
m)
exp((R / m) 1
))R + (1 +
1 (exp((R / m)
1+
m
1 exp((R /
m)
)(2bm b(R + b(#R + amR)))

1
m
2
R
3
exp(

(R / m) exp(

(R / m)
exp(
(R
)(1 + m)
* (exp((R / m)(b(
(
(1 exp((R / m)
2
(exp((R / m)
1+ m
(
1+ m
(1 exp((R / m)
2
+
m
))
exp(
(R
)
1
m
* R(m + (#R) + (1 +
1 (exp((R /
m)
1 exp((R /
m)
)(2bm
2
b(mR b(#mR + am
2
R + b(
2
#R
2
a(#mR
2
))))
.)*)8) The total cost function
The total cost of the inventory system is comprised of the five aforementioned
components, and is given by3
TC CR + C( + CP + CI + C, , ><.66?
.),) '&timal Solution and 6umerical +7am&les
#he present value of the total cost of the system is a nonlinear function of one
variable, m. #he decision variable is not continuous and hence the optimal value can
not be found by ta%ing the derivative and equating it to zero. #o obtain the optimal
number of replenishment cycles during the planning horizon, a unidimensional
optimization technique >(olden search? was used. #o use the unidimentional search
method >(olden search?, one needs to specify an interval in which the optimum value
of decision variable, m , lies. #he interval is split into two segments according to what is
termed the Qgolden sectionQ, in which the ratio of the whole interval to the larger
segment is the same as the ratio of larger segment to the smaller one. #he algorithm
was implemented on a problem with the following parameters, a 10 , b 1 , ( 10 ,
R {0.10,0.10,0.20,0.40} , % 2
,
& 0.32 , p 1 , I
e
0.04 , I
c
0.15 , # {0.1,0.20} .
For more details about the golden search refer to !ppendi /. #he algorithm was
implemented on an interval for m between 6 and 7:.#he results obtained are reported
in #able <.6.
(race 2eriod 1 C r - f #otal )ost >I? Number of
)ycles
, C :.7 H:.6 67:.@:< 67
-:.6 9<B.@:B 6;
-:.7 8B;.= 6;
-:.; 9:@@.8< 6<
, C :.6 H:.6 676.=69 67
-:.6 987.:9= 6;
-:.7 8B=.=9< 6;
-:.; 9:B:.<B 6<
#able <.6. 5ptimal number of replenishment cycles #otal cost of
the system for different value of 1, with respect to ,.
Figure <.7. illustrates the relationship between the total cost of the system and
the inflation rate. #he figure shows the cost increases significantly with the increase in
inflation rate. #he total cost of the system also decreases when the length of the grace
period increases. #he number of replenishment cycles increases with the increase in
inflation rate.
#
o
t
a
l

)
o
s
t
#otal )ost
;:::
9:::
7:::
6:::
:
-:.8 -:.; -:.7 : :.7
1
#otal )ost, ,C :.7
Figure.<.73 relationship between the total cost of the system and
the inflation rate, when , C :.7.
"ha&ter Si7
"onclusions and $uture /esearch
In this research a number of static and dynamic inventory models were
developed in which the effects of economic factors such as inflation and time value of
money were ta%en into account. !llowing for a delay in payment is a common practice
among suppliers that lets customers pay for the goods bought within a certain period of
time. #he effects of permissible delay in payment in the inventory models were studied.
First a static single item model was considered in which the shortages were allowed and
the effect of a permissible delay in payment on the model was investigated. Net, the
same model was augmented by considering the effects of inflation and time value of
money. !ppropriate mathematical models were developed and a search method was
used to obtain the optimal policies of the inventory systems. #he ob"ective was to find
the optimal length of replenishment cycle and the optimal length of the period during
which the inventory level is positive.
! single item deterministic model was also considered in which the demand is a
linear function of time. Inflation, time value of money and permissible delay in payment
were considered in the development of the mathematical model representing the
system over a finite planning horizon. #he main ob"ective of the problem was to
determine the optimal number of replenishment cycles over the planning horizon.
'everal etensions can be made to this research. In the first problem in which an
E5A model was considered with shortages and permissible delay in payment, it was
assumed that the replenishment rate was infinite and goods were delivered
instantaneously as the order was released. In many practical cases, a finite
replenishment system ta%es place in which raw material is processed into products and
added to the inventory at a finite rate. #he problem of a single item inventory system
with finite input rate, no shortages and permissible delay in payment needs to be
investigated as an etension of the model developed in this research. 5ne needs to
etend the above problems while considering the effects of inflation and time value of
money.
In developing the dynamic inventory model with permissible delay in payment, it
was assumed that the length of the grace period was a fraction of replenishment cycle.
5ne may consider a case in which the length of the grace period is fied and does not
depend on the length of replenishment cycle. !lso in this research shortages were not
allowed in the dynamic model. !s an etension to the model one may consider a case in
which shortages are allowed. !lso in developing the dynamic model, replenishment
cycles were restricted to be equal in length. 5ne may want to rela this restriction and
allow for replenishment cycles with different lengths.
/eferences
+uzacott, &. !.M> 6B@< ?, Economic order quantities with inflation, 5perational 1esearch
Auarterly, *8 ,<<9-<<=
)hand, 'ureshM -ard, &amesM>6B=@?, Note on Economic 5rder Auantity under conditions
of permissible delay in payment, &ournal of the 5perational 1esearch 'ociety , ,> , =9-=;
)handra, ,. &eyaM +ahner, ,ichael 4.M>6B=<?, Effects of inflation and the time value of
money on some inventory systems, International &ournal of 2roduction 1esearch ,*, , @79-
@9:
)hung, 0un-&enM>6BB=?, #heorem on the determination of economic order quantity under
conditions of permissible delay in payments, )omputers R 5perations 1esearch ,*.,
6BB=,
;B-<7
/atta, #.0.M 2al, !.0.M >6BB6?, Effects of inflation and time value of money on an inventory
model with linear time-dependent demand rate and shortages, European &ournal of
5perational 1esearch ,.* , 978-999
/eb, ,aitreyeeM )haudhuri, 0.M>6B=@?, Note on the heuristic for replenishment of trended
inventories considering shortages, &ournal of the 5perational 1esearch 'ociety , ,>, ;<B-
;89
(oyal, '. 0.M >6B=<?, Economic order quantity under conditions of permissible delay
in payments, &ournal of the 5perational 1esearch 'ociety, ,8 , 99<-99=
$ariga, ,oncerM >6BB;?, Inventory lot-sizing problem with continuous time-varying demand
and shortages, &ournal of the 5perational 1esearch 'ociety , -. , =7@-=9@
$ariga, ,.!.M +en-/aya, ,.M>6BB8?, 5ptimal time varying lot-sizing models under inflationary
conditions, European &ournal of 5perational 1esearch ,>? , 969-97<
$immelblau, /avid ,.M >6B@7?, !pplied Nonlinear 2rogramming, ,c(raw-$ill +oo% )ompany
'hinn, 'eong -hanM $wang, $ar%M 2ar%, 'ung 'ooM>6BB8?, &oint price and lot size
determination under conditions of permissible delay in payments and quantity discounts for
freight cost, European &ournal of 5perational 1esearch ,?( , <7=-<;7
4iao, $ung-)hang, #sai, )hih-$ung and 'u, )hao-#ungM >7:::?, !n inventory model
with deteriorating items under inflation when a delay in payment is permissible, Int. &. of
2roduction Economics, 8,, 7:@-76;
,isra, 1am +.M>6B@B?, Note on optimal inventory management under inflation, Naval
1esearch 4ogistics Auarterly, *8 , 686-68<
,urdeshwar, #. ,.M >6B==?, Inventory replenishment policy for linearly increasing demand
considering shortages-an optimal solution, &ournal of the 5perational 1esearch 'ociety , ,? ,
8=@-8B7
2helps, 1. I.M>6B=:?, 5ptimal inventory rule for a linear trend in demand with a constant
replenishment period, &ournal of the 5perational 1esearch 'ociety , ,(, ;9B-;;7
1ay, &.M )haudhuri, 0.'.M >6BB@?, E5A model with stoc%-dependent demand rate, shortage,
inflation and time discounting, International &ournal of 2roduction Economics, ., , 6@6-6=:
1itchie, E.M >6B=;?, E5A for linear increasing demand3 a simple optimal solution, &ournal of
the 5perational 1esearch 'ociety , ,. , B;B-B<7
'ar%er, +haba , &amal, !.,., and -ang, 'hao"unM >7:::?, 'upply chain models
for perishable products under inflation and permissible delay in payment,
)omputers R 5perations 1esearch, *@, <B-@<
'ar%er, +haba 1.M 2an, $aiuM >6BB;?, Effects of inflation and the time value of money on
order quantity and allowable shortage, International &ournal of 2roduction Economics, ,- ,
8<-@7
'ilver, Edward !.M >6B@B?, 'imple inventory replenishment decision rule for a linear trend in
demand, &ournal of the 5perational 1esearch 'ociety , ,A, @6-@<
APP+6DI"+S
1
-
1 1
1
1 1
A&&endi7 A) <ooke and Beeves, Cnconstrained MinimiDation Procedure
.sing derivatives in solving unconstrained nonlinear programming problems
leads to quic%er solutions compared to direct search methods, but problems may
arise implementing such methods. In problems with a large number of variables, it
may be difficult if not impossible to derive close formulae for the variables.
In this section the algorithm proposed by $oo%e and &eeves for solving
unconstrained nonlinear problems is presented. #he algorithm is comprised of two
phases, first an FEploratory 'earchG is performed around a base point to find the best
direction to move, and second a F2attern 'earchG is used to minimize the function.
!ssume that the function " ( !
)
needs to be minimized. Elements of ! are the
decision variables. In order to implement the algorithm, the initial values of the decision
variables, elements of ! , must be provided as well as the initial incremental changes
! . !t the first step, the ob"ective function, " ( ! ) , is evaluated at the base point
provided by the user, then each variable is changed while %eeping all the others
unchanged. #o be specific
-
1
(0)
is changed by the amount of
+ -
(0)
, so that
(1)
1
-
(0)
+
-
(0)
. #he ob"ective function " ( !
)
is evaluated at the new pointM if there
is an improvement in the ob"ective function,
-
(1)
-
(0)
+
-
(0)
is considered as the
1 1 1
new value of
-
1
. If there is no improvement in the ob"ective function by increasing the
value of -
1
,

-
(0)
is subtracted from -
1
(0)
and again the ob"ective function is evaluated
at the new point,
-
(1)
-
(0)
-
1
(0)
.If the value of ob"ective function is not improved
1 1
-
by either (1)
1
-
(0)
t
-
(0)
,the value of -
1
is left unchanged. #hen -
2
is changed by
the amount of + -
2
(0)
and so on till all the decision variables have been changed and
the effects of their changes on the ob"ective function have been investigated. !fter
ma%ing one or two eploratory searches a pattern search is made. #hose variable
changes which improved the ob"ective function, form a vector which shows a direction
suitable to move along in order to decrease the value of " ( ! ) . ! series of movements
are made along this vector as long as the ob"ective function improves. #he etent of
the steps in the pattern search for each variable depends on the number of successful
steps previously made in each coordinate during the eploratory search in previous
cycles.
" ( !
)
if is not improved after the pattern search, a new eploratory search is made in
order to find a new direction to move. If the eploratory search does not give a new
successful direction, the amount of ! is reduced until a new direction can be defined
or each
-
i
becomes less than some predefined factor in order to stop the search. In
order to stop the algorithm there are three tests that need to be satisfied. #he first one
compares the change in the ob"ective function with a prescribed small number, after
each eploratory search and pattern search. If the ob"ective function does not change
by a value that eceeds the specified number, the eploratory search and pattern
search fail. #he second test is performed in the absence of the aforementioned failure
to determine if the ob"ective function increases >failure? or decreases >success? to
ensure that the value of the ob"ective function is always improving. #he third test
compares the amount of
-
i
, after an eploratory search failure, with some prescribed
numbers. If the amount of change in each variable is less than the specified number,
the test can be terminated.
A&&endi7 B) "om&uter &ro!ram used to minimiDe the total cost function of the
+'4 model ith shorta!es and &ermissible delay in &ayment hen inflation is
not considered%
5ption Eplicit
2rivate IterS5ccured !s 4ong
2rivate Iter !s Integer
2rivate optimalSpoint>7? !s *ariant
2rivate 5pt#ime>? !s *ariant
2rivate startpt>: #o 6? !s *ariant
2rivate delta>7? !s *ariant
2rivate prevbest !s 'ingle
2rivate startSpoint>? !s *ariant
2rivate iterationSma !s Integer
2rivate demand !s 'ingle
2rivate purchase !s 'ingle
2rivate 'ub cmd)alculateS)lic%>?
startpt>:? C *al>tt1eplenishment.#et?
startpt>6? C *al>tt2ositive.#et?
/im stpSshr% !s *ariant
/im eps !s *ariant
/im ttcost !s 'ingle
/im ii !s Integer, NoSofSIterations !s Integer
iterationSma C <:::
stpSshr% C *al>tt1ho.#et?
eps C *al>ttEpsilon.#et?
NoSofSIterations C $oo%eS&eeves>startpt>?, stpSshr%, eps, iterationSma?
/im tcost !s 'ingle
tcost C #otalS)ost>optimalSpoint>?? ttcost C tcost H
demand O purchase lbl#otal)ost.)aption C
Format>ttcost, QTT::.:::::Q?
lbl5ptimal1ep.)aption C Format>optimalSpoint>:?, QTT:.:::::Q?
lbl5ptimal2ositive.)aption C Format>optimalSpoint>6?, QTT:.:::::Q?
End 'ub
2ublic 'tatic Function #otalS)ost>5pt#ime>? !s *ariant? !s 'ingle
/im IterS5ccured !s 4ong
IterS5ccured C IterS5ccured H 6
/im setup !s 'ingle
/im F !s 'ingle
/im short)ost !s 'ingle
/im int)harges !s
'ingle /im intEarned !s
'ingle /im delay !s
'ingle
demand C *al>tt/emand.#et?
purchase C *al>tt2urchase.#et?
setup C *al>tt'etup.#et?
F C *al>tt)arry.#et?
short)ost C *al>tt'hortage.#et?
int)harges C *al>tt)harges.#et?
intEarned C *al>ttEarned.#et?
delay C *al>tt/elay.#et?
/im 1epS)ycle !s 'ingle, InvS$old !s 'ingle
1epS)ycle C 5pt#ime>:?
InvS$old C 5pt#ime>6?
#otalS)ost C >7 O setup H demand O purchase O F O InvS$old O InvS$old H short)ost O
demand O >1epS)ycle - InvS$old? O >1epS)ycle - InvS$old? H demand O purchase O
int)harges O >InvS$old - delay? O >InvS$old - delay? - 7 O demand O purchase O
intEarned O >1epS)ycle - InvS$old H >delay E 7?? O delay? E >7 O 1epS)ycle?
End Function
2rivate 'ub
cmdEitS)lic%>? End
End 'ub
2ublic Function eploratorySsearch>delta>? !s *ariant, basepoint>? !s *ariant, prevbest
!s 'ingle? !s 'ingle
/im newpoint>7? !s *ariant
/im mincost !s 'ingle
/im tmpcost !s
'ingle /im i !s
Integer mincost C
prevbest
For i C : #o 6
newpoint>i? C basepoint>i?
Net i
For i C : #o 6
newpoint>i? C basepoint>i? H delta>i?
tmpcost C #otalS)ost>newpoint>??
If tmpcost U mincost #hen
mincost C tmpcost
Else
delta>i? C : - delta>i?
newpoint>i? C basepoint>i? H delta>i?
tmpcost C #otalS)ost>newpoint>??
If tmpcost U mincost #hen
mincost C tmpcost
Else
newpoint>i? C basepoint>i?
End If
End If
Net i
For i C : #o 6
basepoint>i? C newpoint>i?
Net i
eploratorySsearch C mincost
End Function
2ublic Function $oo%eS&eeves>startSpoint>? !s *ariant, stpSshr% !s *ariant, epsilon
!s
*ariant, iterma !s Integer? !s Integer
/im del>7? !s *ariant /im
NewS)ost !s 'ingle /im
2revS)ost !s 'ingle /im
steplength !s *ariant /im
#empS2oint !s 'ingle
/im before>7? !s *ariant
/im new>7? !s *ariant
/im i !s Integer
/im " !s Integer
/im flag !s Integer
/im iters !s Integer
For i C : #o 6
new>i? C startSpoint>i?
before>i? C new>i?
del>i? C !bs>startSpoint>i? O stpSshr%?
If del>i? C : #hen
del>i? C stpSshr%
End If
Net i
steplength C stpSshr%
iters C :
2revS)ost C #otalS)ost>new>??
NewS)ost C 2revS)ost
/o -hile iters U iterma !nd steplength V epsilon
iters C iters H 6
For " C : #o 6
new>"? C before>"?
Net "
NewS)ost C eploratorySsearch>del>?, new>?, 2revS)ost?
flag C 6
/o -hile NewS)ost U 2revS)ost !nd flag C 6
For i C : #o 6
If new>i? UC before>i? #hen
del>i? C : - !bs>del>i??
Else
del>i? C !bs>del>i??
End If
#empS2oint C before>i?
before>i? C new>i?
new>i? C new>i? H new>i? - #empS2oint
Net i
2revS)ost C NewS)ost
NewS)ost C eploratorySsearch>del>?, new>?, 2revS)ost?
If NewS)ost VC 2revS)ost #hen
Eit /o
End If
flag C :
For i C : #o 6
flag C 6
If !bs>new>i? - before>i?? V :.< O !bs>del>i?? #hen
Eit For
Else
flag C :
End If
Net i
4oop
If steplength VC epsilon !nd NewS)ost VC 2revS)ost #hen
steplength C steplength O stpSshr%
For i C : #o 6
del>i? C del>i? O stpSshr%
Net i
End If
4oop
For i C : #o 6
optimalSpoint>i? C before>i?
Net
$oo%eS&eeves C iters
End Function
A&&endi7 ") "om&uter &ro!ram used to minimiDe the total cost function of the
+'4 model ith shorta!es and &ermissible delay in &ayment hen inflation is
considered%
5ption Eplicit
2rivate IterS5ccured !s Integer
2rivate optimalSpoint>7? !s *ariant
2rivate 5pt#ime>? !s *ariant
2rivate startpt>: #o 6? !s *ariant
2rivate delta>7? !s *ariant
2rivate prevbest !s 'ingle
2rivate startSpoint>? !s *ariant
2rivate iterationSma !s Integer
2rivate 'ub cmd)alculateS)lic%>?
startpt>:? C *al>tt1eplenishment.#et?
startpt>6? C *al>tt2ositive.#et?
/im stpSshr% !s *ariant
/im eps !s *ariant
/im ii !s Integer, NoSofSIterations !s Integer
iterationSma C *al>ttNoIterations.#et?
stpSshr% C *al>tt1ho.#et?
eps C *al>ttEpsilon.#et?
NoSofSIterations C $oo%eS&eeves>startpt>?, stpSshr%, eps, iterationSma?
/im NI !s Integer
NI C NoSofSIterations O 6:
prg6.,a C NI
prg6.,in C :
For ii C : #o NI
prg6.*alue C ii
Net ii
/im tcost !s
'ingle
tcost C #otalS)ost>optimalSpoint>?? lbl#otal)ost.)aption C
Format>tcost, QTT::.:::::Q? lbl5ptimal1ep.)aption C
Format>optimalSpoint>:?, QTT:.:::::Q? lbl5ptimal2ositive.)aption C
Format>optimalSpoint>6?, QTT:.:::::Q? End 'ub
2ublic 'tatic Function #otalS)ost>5pt#ime>? !s *ariant? !s
/ouble
/im / !s 'ingle
/im p !s 'ingle
/im ! !s 'ingle
/im F !s 'ingle
/im 2i !s 'ingle
/im Ic !s 'ingle
/im Ie !s 'ingle
/im , !s 'ingle
/im Inflation !s *ariant
/im Interest !s *ariant
/im 1 !s *ariant
/ C *al>tt/emand.#et?
p C *al>tt2urchase.#et?
! C *al>tt'etup.#et?
F C *al>tt)arry.#et?
2i C *al>tt'hortage.#et?
Ic C *al>tt)harges.#et?
Ie C *al>ttEarned.#et?
, C *al>tt/elay.#et?
Inflation C *al>ttInflation.#et?
Interest C *al>ttInterest.#et? 1
C Inflation - Interest
/im # !s *ariant
/im #6 !s *ariant
# C 5pt#ime>:?
#6 C 5pt#ime>6?
#otalS)ost C >! H F O p O / O >>Ep>1 O #6? - #6 O 1 - 6? E 1 W 7? H >>p O / O ># - #6
H ,? O Ep>1 O ,?? H >p O / O >#6 - ,? O >Ep>1 O #6???? H >2i O / O >Ep>1 O #6? H Ep>1
O #? O >1 O ># - #6? - 6?? E 1 W 7? H >/ O p O Ic O >Ep>, O 1? H Ep>1 O #6? O >1 O #6 - , O
1 - 6?? E 1 W 7? - >Ie O p O / O >>Ep>, O 1? - , O 1 - 6? E 1 W 7?? - >Ie O p O / O >>Ep>, O 1?
- 6? E 1? O ># - #6??? O >>6 - Ep>1?? E >6 - Ep>1 O #???
End Function
2rivate 'ub
cmdEitS)lic%>? End
End 'ub
2ublic Function eploratorySsearch>delta>? !s *ariant, basepoint>? !s *ariant, prevbest
!s 'ingle? !s 'ingle
/im newpoint>7? !s *ariant
/im mincost !s 'ingle
/im tmpcost !s
'ingle /im i !s
Integer mincost C
prevbest
For i C : #o 6
newpoint>i? C basepoint>i?
Net i
For i C : #o 6
newpoint>i? C basepoint>i? H delta>i?
tmpcost C #otalS)ost>newpoint>??
If tmpcost U mincost #hen
mincost C tmpcost
Else
delta>i? C : - delta>i?
newpoint>i? C basepoint>i? H delta>i?
tmpcost C #otalS)ost>newpoint>??
If tmpcost U mincost #hen
mincost C tmpcost
Else
newpoint>i? C basepoint>i?
End If
End If
Net i
For i C : #o 6
basepoint>i? C newpoint>i?
Net i
eploratorySsearch C mincost
End Function
2ublic Function $oo%eS&eeves>startSpoint>? !s *ariant, rho !s *ariant, epsilon !s
*ariant, iterma !s Integer? !s Integer
/im del>7? !s *ariant /im
NewS)ost !s 'ingle /im
2revS)ost !s 'ingle /im
steplength !s *ariant /im
#empS2oint !s 'ingle
/im before>7? !s *ariant
/im new>7? !s *ariant
/im i !s Integer
/im " !s Integer
/im flag !s Integer
/im iters !s Integer
For i C : #o 6
new>i? C startSpoint>i?
before>i? C new>i?
del>i? C !bs>startSpoint>i? O rho? If
del>i? C : #hen
del>i? C rho
End If
Net i
steplength C rho
iters C :
2revS)ost C #otalS)ost>new>??
NewS)ost C 2revS)ost
/o -hile iters U iterma !nd steplength V epsilon
iters C iters H 6
For " C : #o 6
new>"? C before>"?
Net "
NewS)ost C eploratorySsearch>del>?, new>?, 2revS)ost?
flag C 6
/o -hile NewS)ost U 2revS)ost !nd flag C 6
For i C : #o 6
If new>i? UC before>i? #hen
del>i? C : - !bs>del>i??
Else
del>i? C !bs>del>i??
End If
#empS2oint C before>i?
before>i? C new>i?
new>i? C new>i? H new>i? - #empS2oint
Net i
2revS)ost C NewS)ost
NewS)ost C eploratorySsearch>del>?, new>?, 2revS)ost?
If NewS)ost VC 2revS)ost #hen
Eit /o
End If
flag C :
For i C : #o 6
flag C 6
If !bs>new>i? - before>i?? V :.< O !bs>del>i?? #hen
Eit For
Else
flag C :
End If
Net i
4oop
If steplength VC epsilon !nd NewS)ost VC 2revS)ost #hen
steplength C steplength O rho
For i C : #o 6
del>i? C del>i? O rho
Net i
End If
4oop
For i C : #o 6
optimalSpoint>i? C before>i?
Net
$oo%eS&eeves C iters
End Function
1
1
A&&endi7 D) :olden Search Method
In this section details about a unidimensional optimization technique
called (olden 'earch ,ethod is presented. #o use the (olden search method, one
needs to specify an interval in which the optimum value of decision variable, - , lies.
#he interval is split into two segments according to Qgolden sectionQ, in which the ratio
of the whole interval to the larger segment is the same as the ratio of larger segment to
the smaller one. #he two ratios employed are3
&

3
1
2
5
0.38
and
&

5 1

0.62 . Note that & (& )

2
and & +
&
1. 4et the three - values
2
2
1 2 1 2
designated for the interval in which the optimum of - lies as
-
3
(0)
>the last point?,
-
2
(0)
,
-
1
(0)
, where " ( -
3
(0)
)
" ( -
2
(0)
) , and let the interval

(k )
-
3
(k )
-
(k )
. For the
%-th stage the >%H6?-th interval is computed as follows. /etermine
y
(k )
-
(k )
+ &
1

(k )
and y
2
(k )
-
1
(k )
+ &
2

(k )
-
3
(k )
&
1

(k )
.
If
" ( y
(k )
)
>
" ( y
2
(k )
)
1 1
then

(k +1)
(
y
(k )
-
(k )
)
and -
(k +1)
-
(k )
, -
(k +1)
y
(k )
. If " ( y
(k )
)
<
" (
y
(k )
) then
2 1 1 1 3 2 1 2

(k +1)
(
-
(k )
y
(k )
) and -
(k +1)
y
(k )
, -
(k +1)
-
(k )
. If " ( y
(k )
)

" (
y
(k )
) then
3 1 1 1 3 3 1 2

(k +1)
(
y
(k )
-
(k )
) C (
-
(k )
y
(k )
) and -
(k +1)
-
(k )
, -
(k +1)
y
(k )
or
2 1 3 1 1 1 3 2
-
(k +1)
y
(k )
,
-
(k +1)
-
>k ?
.
1 1 3 3