Beruflich Dokumente
Kultur Dokumente
397 - 407
Since I(T ) = 0, we get neglecting terms of order higher than 0(2 ) the
following
z 2 T 2
Q = (1 2 )[z + ] + 2 [T + ] (5)
2 2
The average number of units AI(T ) in the inventory during the period
(0, T ) is
Z
1 T
AI(T ) = I(t) dt
T Z0 z Z T
1
= [ I(t)dt + I(t)dt]
T 0 z
Substituting the values of I(t) in the above integrals and then eliminating
Q using equation (5) and neglecting higher order terms in we find the
following :
1 z 2 z 3 T 2 T 3
AI(T ) =[(1 2 )( + ) + 2 ( + )] (6)
T 2 6 2 6
Therefore, aveage cost per unit time in (0, T ) is
K + cQ h z 2 z 3 T 2 T 3
c(T, z) = + [(1 2 )( + ) + 2 ( + )]
T T 22 6 22 6
z T
= K/T + c[(1 2 )(z + ) + 2 (T + )]/T
2 3
2 2 3
2
z z T T
+h[(1 2 )( + ) + 2 ( + )]/T
2 6 2 6
We now minimize the average cost per unit time c(T, z) under the fol-
lowing two situtations :
(i) z is a known point of time.
(ii) z is a random point of time.
Case I : z is a known point of time.
400 A.K. Pal and B. Mandal
(a) For z T .
In this case, the expression for the average cost c(T, z) reduces to the
following
T 2 T 2 T 3
c(T ) = K/T + c1 [T + ]/T + h1 [ + ]/T
2 2 6
which is the cost function of the EOQ inventory model for deteriorating
items.
If there be no deterioration i.e. = 0, the above expression further
simplifies to the following
c(T ) = K/T + c1 + h1 T /2
which is similar to the equation obtained by Harris and Wilson.
(b) For z < T .
Since z < T , we can take z = 0 T, 0 0 < 1.
Thus from expression (6), we find
T 2 0 T 3 T 2 T 3
AI(T ) = [(1 2 )02 ( + ) + 2 ( + )]/T
2 6 2 6
Hence (7) reduces to the following
0 T T 0 T
c(T ) = K/T + c1 0 (1 + ) + c2 (1 0 )(1 + + )
2 2 2
hT 0 T T 2 T
+ [1 02 (1 + ) + 2 (1 0 ){(1 + 0 )(1 + ) + 0 }]
2 2 3 3
(8)
For minimum average cost, the necessary condition is c(T T
)
= 0.
This gives
h + c 2 h 2
K/T 2 + [0 1 + (1 02 )2 ] + T [1 03 + 2 (1 03 )] = 0
2 2 3
or, LT 3 + MT 2 + N = 0 (9)
where
2
L = h[1 03
+ 2 (1 03 )]
3 (10)
M = (h + c)[1 02 + 2 (1 02 )]
N = 2K
Since 0 0 < 1 and the last term of the cubic equation (9) is negative,
it has at least one positive real root.
An EOQ Model for Deteriorating Inventory 401
Let T be the positive real root of the equation (9), then T is the
optimum cycle time. It can also be seen that the sufficient condition for
minimum cost
2 c(T )
|T =T > 0 is satisfied.
T 2
From expression (5), we can determine the optimum value of Q as follows.
0 T T
Q = 1 0 T [1 + ] + 2 T (1 0 )[1 + ] (11)
2 2
The expression for the minimum cost c(T ) can be obtained by substi-
tuting T = T in the equation (8) where T is the optimum cycle time
to be obtained from equation (9).
Particular Cases :
(i) Absenece of deterioration :
In the absence of deterioration ( = 0), the expression for the minimum
cost c(T ) and the optimum ordering quantity Q reduce to the following
:
hT 2
c(T ) = [K/T + c0 (1 2 ) + c2 + {0 (1 2 ) + 2 }]
2
and Q = 1q 0 T + 2 T (1 0 )
where T = 2K/[1 02 + 2 (1 02 )]h
The optimum cycle time and order quantity are the following
q q
T = 2K/h1 , Q = 2K1 /h.
Case II : z is a random point of time.
From equation (7), we have
z 2 T 2
c(T, z) = K/T + c[(1 2 )(z + ) + 2 (T + )]/T
2 2
z 2 z 3 T 2 T 3
+h[(1 2 )( + ) + 2 ( + )]/T
2 6 2 6
In this case, the cost function c(T, z) is a random variable with respect
to z.
So expected total cost per unit time is
1 T 2
(T ) = [K + c(1 2 )[E(z) + E(z 2 )] + c2 (T + )]
T 2 2
23
h 1 2 h T T
+ (1 2 )[ E(z 2 ) + E(z 3 )] + ( + ) (12)
T 2 6 T 2 6
Let us now assume the distribution function of z to be Fz ().
Therefore we can take
"
FY (u), u < T
F z (u) =
0 uT
Substituting these integrals for E(z), E(z 2 ) and E(z 3 ) in the equation
(12), the expected total cost reduces to the following
RT R
(T ) = 1
T
[K (1 2 ){c F Y (u)du + (h + c) 0T uF Y (u)du
0
h Z T 2 cT 2 hT 2 hT 3
+ u F Y (u)du} + 2 (cT + + + )]
2 0 2 2 6
(13)
An EOQ Model for Deteriorating Inventory 403
(T )
Now equating T
to zero, we get the following
3
T + 2 (h+c)
2 h 3
2
T 2 [K + (1 2 ){c(a + b)
+ 2 (a + 2ab + 2b2 )(h + c) + h
1 2
6
(a3 + 3a2 b +
6ab2 + 6b3 )}] = 0.
(15)
Since 1 > 2 , the above cubic equation must have one positive real root
which is the optimum cycle time T .
Using (5), we get the expected optimum ordering quantity is
2
< Q >= (1 2 )[a + b + (a2 + 2ab + 2b2 )] + 2 [T + T ] (16)
2 2
Now we have studied the following three situations :
(i) when there is no deterioration ( = 0).
In this case equation (15) reduces to the following
2 h 2 h
T [K + (1 2 ){c(a + b) + (a2 + 2ab + 2b2 )}] = 0
2 2
Therefore the optimum cycle time is
h
T = [2K/2 h + 2(1 2 ){c(a + b) + (a2 + 2ab + 2b2 )}/2 h]1/2
2
Also from (16), the optimum ordering quantity is
< Q > = (1 2 )(a + b) + [2K2 /h + 22 (1 2 ){c(a + b)
h
+ (a2 + 2ab + 2b2 )}/h]1/2
2
(ii) when there is no change of demand rate i.e. uniform demand rate
(1 = 2 ).
In this case equation (15) gives
2 h 3 2
T + (h + c)T 2 K = 0
3 2
The optimum cycle time T can be obtained from the above cubic equa-
tion.
Also the optimum ordering quantity
< Q >= 2 [T + T 2 ].
2
(iii) when there is no deterioration and no change of demand rate ( =
0, 1 = 2 ).
Therefore from equation (15), the optimum cycle time is
T = [2K/1 h]1/2
An EOQ Model for Deteriorating Inventory 405
References
1. B.V. Dhandra and M.S. Prasad, A continuous review inventory model with al-
ternating demand rates , International journal of Management and Systems, Vol.
11(1995), No. 1, 71-80.
2. F.W. Harris, What quantity to make at once. The Library for Factory Management,
Vol. V, Operation and costs, A.W. Shaw, Chicago (1915), 47-52.
3. G. Hadley and T.M. Whitin, Analysis of inventory system , prentice - Hall Inc.,
1963.
An EOQ Model for Deteriorating Inventory 407