Documente Academic
Documente Profesional
Documente Cultură
'
= +
=
= + +
Thus, the optimum inventory policy for the
model is to order
*
2
o
h
DC
Q
C
=
units for every t
0
*
=Q
*
/D time units.
Total cost T(Q
*
).
14
EOQ with planned shortages.
Costs are incurred because of lost business, the cost of
placing special orders, loss of future goodwill and so on.
Q/D
Q
Q-Dt
M
Q - M
O A
B
15
M
OA
D
=
Q
OB
D
=
( )
Q M
AB
D
=
2
2 2
h
h
M C M M
C
D D
=
| || |
| |
\ .\ .
( )
2
2 2
b
b
Q M C
Q M Q M
C
D D
=
| || |
| |
\ .\ .
o
C cQ +
Assuming that C
b
be the cost of shortage,
Holding
Cost/cycle :
Shortage
Cost/cycle :
Order
Cost/cycle :
16
Total cost per unit time:
( )
( )
( )
( )
2
2
2
2
2 2
,
/
2 2
b
h
o
b
o h
C Q M
M C
C cQ
D D
T Q M
Q D
C Q M
M C DC
cD
Q Q Q
| |
| |
|
|
|
\ .
\ .
+ + +
=
= + + +
17
Minimum occur at
0
T T
M Q
c c
= =
c c
18
( )
1 1
0
h b
b
b h
T
MC Q C Q M Q
M
C
M Q
C C
| |
|
\ .
c
= =
c
=
+
19
( )
( )
2 2 2 1
0
2
2
2
0
0
1
2
1
0
2
2
2
h b
b
b h
h b
b h
h b
T
DC Q M C Q C Q M Q
Q
C Q M Q
DC C C
Q
C C
DC C C
Q
C C
c
= +
c
=
| || |
+
=
| |
\ .\ .
| || |
+
=
| |
\ .\ .
20
1/ 2
1/ 2
*
2
o b b
h h b h b
DC C C
M EOQ
C C C C C
(
| | | |
= =
(
| |
+ +
\ . \ .
1/ 2
*
h b
b
C C
Q EOQ
C
+
=
| |
|
\ .
Optimal policy:
Notes: May use different value of lead time.
21
If then
T T
L C s
e T
L L =
If then
T T
L C >
T
e T T
T
L
L L C
C
=
(
(
lead time cycle time
T T
L C
22
Example 1: Bart's Barometer Business (BBB) is
a retail outlet which deals exclusively with
weather equipment. Currently BBB is trying to
decide on an inventory and reorder policy for
home barometers. Barometers cost BBB $50
each and demand is about 500 per year
distributed fairly evenly throughout the year.
Reordering costs are $80 per order and holding
costs are figured at 20% of the cost of the item.
BBB is open 300 days a year (6 days a week and
closed two weeks in August). Lead time is 60
working days.
23
c = RM50 per barometer
D = 500 of barometers per year
C
0
= RM80
L = 60 working days
Open = 300 days a year
C
h
= 20% of the cost of item
24
( )
0
500
500 50
2
h
Q
C C
Q
| |
| |
+ +
|
|
\ .
\ .
500
80 0.2 50 25000
2
Q
Q
| |
| |
= + +
|
|
\ .
\ .
Number of cycles per year:
500
Q
Total cost per year:
25
( )
*
0
2 2(500)(80)
0.2 50
80000
8000 89.44
10
h
DC
Q
C
= =
= = =
Optimal Reorder Quantity
Approx. 90 barometers
26
Number of Orders Per Year
Number of reorder times per year is
(500/90) = 5.56
or
once every, (300/5.56) = 54 working days
or about every 9 weeks.
Cycle time is 54 working days.
27
Daily demand:
500
1.6667
300
=
Reorder point:
500
60 100
300
=
Bart should reorder 90 barometers when
his inventory position reaches 100 (that
is 10 on hand and one outstanding order).
!!! Lead-time (60 days) is longer than
cycle time (54 days).
28
If then
T T
L C s
e T
L L =
If then
T T
L C >
T
e T T
T
L
L L C
C
=
(
(
lead time cycle time
T T
L C
29
Since 60 54
T T
L C = > =
60
60 54
54
60 (1 54)
60 54 6 days
e
L
(
=
(
=
= =
then
Reorder point:
500
6 10 barometers
300
e
L demand = =
!!! Needs to make one reorder 60 days before
start-up.
30
If lead time is less than cycle time,
lets say 45 days, what will be the
policies?
500
45 75
300
=
Bart should reorder 90 barometers
when his inventory position reaches 75.
45
e T
L L = =
31
Example 2:
Hervis Rent-a-Car EOQ with Planned Shortages
Model
Hervis Rent-a-Car has a fleet of 2,500 Rockets serving
the Los Angeles area. All Rockets are maintained at a
central garage. On the average, eight rockets per month
require a new engine. Engines cost $850 each. There is
also a $120 order cost (independent of the number of
engines ordered). Hervis has an annual holding cost
rate of 30% on engines. It takes two weeks to obtain the
engines after they are ordered. For each week a car is
out of service, Hervis loses $40 profit.
32
Question:
1.How many days after receiving an order
does Hervis run out of engines?
2. How long is Hervis without any
engines per cycle?
33
Optimal Order Policy:
* * *
h
h b
C
Q M Q
C C
| |
=
|
+
\ .
*
2 96 120 255 2080
10.07 10
0.30 850 2080
Q
+
= = ~
* *
255
10 1.09 1
255 2080
Q M
| |
|
\ .
= = ~
+
Needs to order 10 units. 1 unit backorder.
34
Lead-time:
2
T
L =
2 2 4 =
4 1 3 =
2 weeks x 2 rockets per week:
Demand during lead-time minus back order:
Hence, the order should be placed
when there are 3 engines remaining in
inventory.
35
GAME THEORY
Two-person zero-sum game
Saddle point: 1
Pure strategy: (A plays II, B plays I)
Value of the game: v = 1
0 10
1 2
| |
|
\ .
Player A
Player B
I II
I
II
Max{min} = 1
Min{max} = 1
36
A two-person zero-sum game is a two-player
game in which, for any choice of both players
strategies, the row players reward and the
column players reward add up to zero.
37
If some entry a
ij
of the matrix A has the
property that
(1)a
ij
is the minimum of the i
th
row, and
(2) a
ij
is the maximum of the j
th
column,
then we say a
ij
is a saddle point. If a
ij
is a
saddle point, then Player I can then win at
least a
ij
by choosing row i, and Player II can
keep her loss to at most a
ij
by choosing
column j. Hence a
ij
is the value of the game.
38
2 by 2 game:
Two-person zero-sum game
Since two values not equal then no saddle point
Needs, mixed strategy
2 3
3 4
| |
|
\ .
Player A
Player B
I II
I
II
Max{min}=-2
Min{max}=3
39
2 3
3 4
| |
|
\ .
Player A
Player B
I II
I
II
a b
d c
| |
|
\ .
40
Assume no saddle point.
Assume a > b.
If b > c then b is the saddle point, b < c
If c < d then c is the saddle point, c > d
If d > a then d is the saddle point, a > d
a > b,b < c ,c > d and a > d (test 1)
Testing for saddle point
41
Assume no saddle point.
Assume a < b.
If a > d then a is the saddle point, a < d
If c > d then d is the saddle point, c < d
If b < c then c is the saddle point, b > c
a < b ,b > c , c< d and a < d (test 2)
Testing for saddle point
42
Equalizing strategies:
Player A
Player B
q 1 - q
p
1 p
a b
d c
| |
|
\ .
43
For p:
pa + (1 p)d = pb + (1 p)c
(a b)p = (c d)(1 p)
[(a b) + (c d)]p = (c d)
p = (c d)/[(a b) + (c d)]
For q:
qa + (1 q)b = qd + (1 q)c
(a d)q = (c b)(1 q)
[(a d) + (c b)]q = (c b)
q = (c b)/[(a b) + (c d)]
44
( ) ( )
c d
p
a b c d
=
+
( ) ( )
c b
q
a b c d
=
+
Player A
Player B
Mixed strategy: Player A: (p,1-p)
Player B: (q,1-q)
where
45
46
( ) ( )
(1 )
( )
1
( ) ( ) ( ) ( )
v pa d p
a
ac bd
a b c d
c d c d
d
a b c d a b c d
= +
=
+
| | | |
+
| |
+ +
\ . \ .
=
Value of the game:
p = q = 7/12 and v = 1/12
Many larger games can be reduced to 2 by 2 games.
Definition:
Let A = (a
ij
) be an m by n matrix.
(i) Row i dominates row j if a
ik
a
jk
for all k.
(ii) Column i dominates column j if a
ki
a
kj
for all k.
47
Removal of a dominated row or column
does not change the value of a game
2 0 4
1 2 3
4 1 2
A =
| |
|
|
|
\ .
2 0
1 2
4 1
| |
|
|
|
\ .
1 2
4 1
| |
|
\ .
Column 2 dominates
column 3
Column Deleted:
Choose larger
Row 3 dominates row 1
Row Deleted:
Choose smaller
48
A row (column) may also be removed if it is
dominated by a probability combination of
other rows (columns).
49
Steps:
Reduce the matrix (dominance)
Check for saddle point
Equalizing strategy
(use formula for 2 x 2)
Solving 2 x n and m x 2 games
with the aid of graphical interpretation
50
2 3 1 5
4 1 6 0
| |
|
\ .
Consider 2 x 4 game:
p
(1 p)
51
( )
( )
( )
= +
= +
= + =
=
Column 1 (expected payoff) : 1 2 4 1- =4-2p
Column 2 (expected payoff) : 2 3 1- =2p+1
Column 3 (expected payoff) : 3 6 1- 6 - 5
Column 4 (expected payoff) : 4 5
y p p
y p p
y p p p
y p
52
p = 5/7
0
1
y3
y1
y2
y4
{Intersection of y2 and y3}
53
Strategy for player A: (5/7,2/7)
Strategy for player B: (0,5/7,2/7,0)
Value of the game is v = 17/7.
From the graph, column 1 is dominated by column
2 and column 3 taken with probability each.
Delete column 1.
54
3 1
1 6
| |
|
\ .
Reduced to 2 x 2 game:
x1
x2
y2 y3
Player A
Player B
Strategy for player A: (5/7,2/7)
Strategy for player B: (0,5/7,2/7,0)
Value of the game is v = 17/7.
( ) ( )
6 1
5/ 7
2 5
c b
q
a b c d
=
+
= =
+
55
19 6 7 5
7 3 14 6
12 8 18 4
8 7 13 1
| |
|
|
|
|
\ .
larger
Left with column 2 and column 4.
56
6 5
3 6
8 4
7 1
| |
|
|
|
|
\ .
6 5
3 6
8 4
| |
|
|
|
\ .
smaller
57
( )
( )
( )
= +
= +
= + = +
Row 1 (expected payoff) : x1 6 5 1- = 5 + q
Row 2 (expected payoff) : x2 3 6 1- = 6 - 3q
Row 3 (expected payoff) : x3 8 4 1- 4 4
q q
q q
q q q
58
{Intersection of x2 and x1}
5 + q = 6 - 3q
q = 1/4
59
q = 1/4
0
5
3
6
8
4
1
6
6 5
3 6
| |
|
\ .
( ) ( )
6 3
3/ 4
1 3
c d
p
a b c d
=
+
= =
+
60
( ) ( )
36 15
21/ 4
4
ac bd
v
a b c d
=
+
= =
Game value:
Optimal strategies are:
Player A (3/4,1/4,0,0)
Player B (0,1/4,0,3/4)
61
62