Upload
alicia-george
View
215
Download
3
Embed Size (px)
Citation preview
Chapter 7: Stochastic Inventory Model
Proportional Cost Models:
x: initial inventory,
y: inventory position (on hand + on order-backorder),
: random demand, () , (),
(y- )+: ending inventory position, N.B.L,
(y- ) : ending inventory position, B.L,
=1/(1+r) : discount factor,
ordering cost : c(y-x),
holding cost : h (y- )+
penalty cost : p( -y)+
salvage cost : - s(y- )+
Minimum cost f(x) satisfies:
L(y) convex, L’() < -c (otherwise never order) L′ eventually becomes positive
)2()()(min
)()(
)()()()(min)(0
yLxyc
dy
dyshxycxf
xy
y
pcp
y
xy
SxIfxSotherwisexq
Sxxy
cSL
,,0
*
*
)(
},max{)(
PolicyStockBase
)4(0)('
)5()()]1([
)1()(.
)5()()(
)(..
0))(1)(()()(
bcc
c
shccp
cpyLB
acc
c
shccp
cpyLBN
yyshc
ou
u
ou
u
pcp
Example
c=$1, h=1¢ per month, =0.99, p=$2(NBL), p=$0.25(BL),
s=50 ¢, c+h- s=51.5 ¢,
NBL: p-c = 100 ¢, BL: p-c(1- )=24 ¢,
)32.0(,32.05.5124
24)()(
)66.0(,66.05.51100
100)()(
1
1
yyii
yyi
Set up cost K
L(x) if we order nothing
K+c(S-x)+L(S) if we order upto S If we order, L’(S)+c=0.
Use the cheaper of alternatives L(x) and K+ c(S-x)+L(S)
)()()(min)( yLxycxyKxfxy
S x
cost
L(x)
KK
s S x
L(x)+cx
c
s
K+c(S-x)+L(S)
Two-bin or (s,S) policy
order S-x if x ≤ s
order nothing if x > s
Multiperiod models
Infinite Horizon (f1000 & f1001 cannot be different)
dyfyLxycxfxy
)()()()(min)(0
12
)9()()()()(min)(0
dyfyLxycxf
xy
Taking derivative of {}
)10()()(')('00
dSfSLc
If f convex, find S the base stock level, then for x ≤ S
)11()()()()()(0
dSfSLxScxf
We see from (11) thatf’(x)=-c for x ≤ S . (12)
)18(0))(1()('
N.B.Lfor Similarly
(13) B.L0)1()('
toreduces)(10
ScSL
cSL
Proportional costs:
So that
dypdyhyLy
y)()()()()(
0
policyS)(s,stilld,complicatemorecostSetuptime,Lead
:Remark
c-by replaced s- : (5) with cf.
)20()1()1(
)1()(:.
)20()1()(
)(:..
(18),and(13)into(19)Substitute
)19()()()('
bcc
c
chcp
cpSLB
acc
c
chcp
cpSLBN
pSphSL
ou
u
ou
u
Example 4:
1800
5
4
205
20
)(
5 ,20
:Solution
loaves. ofnumber daily optimal theFind loaf.per
cents 5 of loss aat sellsoutlet store a otherwise, on time; sold if loafper
20cents ofprofit Makes ].[1000,2000 Uniform~Demand Bread
S
cc
cS
cc
ou
u
ou
Example 5:
5.73.05.1882
)(25
1
2
3)(
25
1
10
3
)()()()()(
:Solution
0,2
3p(z):costshortage
0,z ,10
3)(:costHolding
,1,15,25
1)(:ondistributiDemand
25
252
25
0
0
25
2
ye
deydey
dypdyhyL
zz
zzh
cKe
y
y
y
y
y
Intuition: The current period would be a separate one
period if we know what the next period would be willing
to pay for our leftover inventory. Assuming we are not
“overstocked”, every unit leftover will mean the next period will
order one less, thus saving c. So the next period should be
willing to pay c per unit in salvage for one leftover inventory.
5.805.1010
2525
25
25
q
:policy optimal The
5.80
:ionapproximatSuccesive
5.73.05.18825.101155.73.05.1882
)()(
5.101
025.753.01)(
3.025.75)(
xifxotherwise
Ss
S
Sy
y
s
Seses
SLcSKsLcs
S
edy
ydLc
edy
ydL
Multiperiod models: No Setup Cost Begin with two periods
Demand D1, D2, i.i.d
Density: ()
L(y) = expected one period holding+ shortage penalty cost;
strictly convex with linear cost and () >0,
c purchase cost /unit
c1(x1) optimal cost with 1 period to go;
c+L’(S1)=0
while S1 is the optimal base stock level.
gotoperiods2withlevelbasestock
convexiswhich)]([)()(min)(
)()]()([)()(
)()())((
)()(
)(
2
11222222
12210
2
02111
)()()(
22111
)()()(11
22
12
12
12222
1221221
111
11111
S
xcEyLxycxc
dSLyScdyL
dycxcE
Dycxc
xc
xy
Sy
Sy
SDyifDyLSDyifSLDySc
SxifxLSxifSLxSc
Example: c=10, h=10, p=15 the demand density is
Solution:
100
10
1
0)(
if
otherwise
2
0
10
11
1
1
)4/5(1575
10
)(10
10
)(15)(
2,10
)(
5
1
1015
1015)(
zz
dz
dz
zL
SS
SSince
hp
cpS
z
z
3/3592/194/)(24/)(
)4/5(1575)(10min)(
3/3592/194/)(24/)(
10
1)](1070[
10
1]))(4/5()(1575[
10
1]2)4/5(2*1575)2(10[
10
1]))(4/5()(1575[)]([
22
23
2
222222
22
23
2
10
22
2
0
222
10
2
22
2
0
22211
222
2
2
2
2
yyy
yyxyxc
yyy
dy
dyy
dy
dyyxcE
xy
y
y
y
y
2201
5502
2
2222
2
222
2
2
11
22
:policy optimal The
5.withvalue
smalleratoleads)(xcinto6and5ngSubstituti
42.5
0])(8
122/29[
{}
zerotoequalitsetting,ytorespectwithderivativeTake
xifxotherwise
xifxotherwise
q
q
S
SS
S
SSdy
d
Multi-Period Dynamic Inventory Model with no Setup Cost
Cn(xn): n periods to go,
: discount factor.
DP equations:
SS
xcxc
dycyLxycxc
S
SSSSSS
xc
DycEyLxycxc
nn
nn
xy
nn
nnnnnnxy
nnnn
lim)3
)(lim)(bysatisfied
)()()()(min)(2)
optimal)horizon (Infinite 0;)-c(1)(L'
where,...................... 1)
:Properties
0)(
])([)()(min)(
0
1321
00
1
Multi-Period Dynamic Inventory Model with Setup Cost
nnnn
nn
nn
nn
nn
SxifxSSxifn
xyifKxyifnn
nnnnnnnxy
nn
q
xyK
dycyLxycxyKxc
0
nn
n
0
01
:)policyS,(soptimalThe
.Sfindthenconvex,isL(y)If
)(
)()()()()(min)(
Multi-Period Dynamic Inventory Model with Lead TimesLead time:
0)1()('
horizoninfinite
)()()(
:follows as timelead 0 toormCan transf
periods) 1-next in arrive order toon hand(on position inventoryu
)()()()()()(min)(
0
n
01
0
cS
dyLy
dyfdyLuycuyKuf nnnnnnnuy
nnnn
L
L