You are on page 1of 3

INVESTIGACION DE OPERACIONES II

PROGRAMACION DINAMICA PROBABILISTICA


ANALISIS RECURSIVO INCLUYENDO PROBABILIDADES

Instrucciones:
Resolver el siguiente problema considerando la condicin de
incertidumbre en el uso de la recursividad
de la programacin dinmica.

PROBLEMA 1
Solucin:

xn : nmero de unidades paralelas a instalar del componente n


pn(xn) : probabilidad de que el componente n funcione si se le instala xn
unidades paralelas
cn(xn) : costo de instalar xn unidades paralelas del componente n
sn : $ que quedan disponibles para gastar en componentes

fn(sn,xn) = max { pn(xn) fn+1*(sn - cn(xn))}

Etapa 4 (componente 4)
f4(s4,x4) = p4(x4) Solucin
ptima
s4
x4 = x4 = x4 = 3 f4*(s4) x4 *
1 2
10 - - - - -
0
20 0.5 - - 0.5 1
0
30 0.5 0.7 - 0.7 2
0
40 0.5 0.7 0.9 0.9 3
0
50 0.5 0.7 0.9 0.9 3
0
60 0.5 0.7 0.9 0.9 3
0

Etapa 3 (componente 3)
f3(s3,x3) = p3(x3) f4*(s3 - c3(x3))} Solucin
s3 ptima
x3 = 1 x3 = 2 x3 = 3 f3*(s3) x3 *
20 - - - 0.35 1
0
30 (0.7) - - 0.35 1
0 (0.5)=0.35
40 (0.7) - - 0.49 1
0 (0.7)=0.49
50 (0.7) (0.8) - 0.63 1
0 (0.9)=0.63 (0.5)=0.40
60 (0.7) (0.8) (0.9) 0.63 1
0 (0.9)=0.63 (0.7)=0.56 (0.5)=0.45
70 (0.7) (0.8) (0.9) 0.72 2
0 (0.9)=0.63 (0.9)=0.72 (0.7)=0.63

Ing. Manuel Snchez Tern


INVESTIGACION DE OPERACIONES II

Etapa 2 (componente 2)
f2(s2,x2) = p2(x2) f3*(s2 - c2(x2))} Solucin
s2 ptima
x2 = 1 x2 = 2 x2 = 3 f2*(s2) x2 *
70 (0.6) (0.7)
- 0.378 1
0 (0.63)=0.378 (0.35)=0.245
80 (0.6) (0.7) (0.8)
0.378 1
0 (0.63)=0.378 (0.49)=0.343 (0.35)=0.280
90 (0.6) (0.7) (0.8)
0.441 2
0 (0.72)=0.432 (0.63)=0.441 (0.49)=0.392

Etapa 1 (componente 1)
f1(s1,x1) = p1(x1) f2*(s1 - c1(x1))} Solucin
s1 ptima
x1 = 1 x1 = 2 x1 = 3 f1*(s1) x1 *
100 (0.5) (0.6) (0.8)
0.3024 3
0 (0.441)=0.2205 (0.378)=0.2268 (0.378)=0.3024

Solucin ptima: x1=3, x2=1, x3=1, x4=3,

Ing. Manuel Snchez Tern


INVESTIGACION DE OPERACIONES II

PROBLEMA 2
Solucin:

Objetivo: maximizar cantidad acumulada (esperada) despus de los 3 aos


Monto a invertir: solamente $5000
Estados: $ acumulados (no necesariamente se invierte todo)
0.3 fn+1*(sn -5000) + 0.7
fn(sn,xn) =f *(s +5000)
n+1 n
0.9 fn+1*(sn) + 0.1 fn+1*(sn+5000)

Etapa 3
A B Solucin
ptima
0.3 (0) + 0.7 0.65 fn+1*(sn) + 0.35 f3*(s3) x3 *
s3 fn+1*(sn+5000) fn+1*(sn+5000)
0 0 0 0 0
5000 (0.30)(0)+(0.70)(10000)=7000 (0.65)(5000)+(0.35)(10000)=6750 7000 A
1000 5000+(0.30)(0)+(0.70)(10000)=12000 5000+(0.65)(5000)+(0.35)(10000)=11750 12000 A
0
1500 10000+(0.30)(0)+(0.70) 10000+(0.65)(5000)+(0.35) 17000 A
0 (10000)=17000 (10000)=16750

Etapa 2
A B Solucin
ptima
0.3 fn+1*(sn -5000) + 0.7 0.65 fn+1*(sn) + 0.35 f2*(s2) x2 *
s2 fn+1*(sn+5000) fn+1*(sn+5000)
0 0 0 0 0
5000 (0.30)(0)+(0.70)(12000)=8400 (0.65)(7000)+(0.35)(12000)=8750 8750 B
1000 (0.30)(7000)+(0.70)(17000)=14000 (0.65)(12000)+(0.35)(17000)=13750 14000 A
0

Etapa 1
A B Solucin
ptima
0.3 fn+1*(sn -5000) + 0.7 0.65 fn+1*(sn) + 0.35 f1*(s1) x1 *
s1 fn+1*(sn+5000) fn+1*(sn+5000)
500 (0.30)(0)+(0.70)(14000)=9800 (0.65)(8750)+(0.35)(14000)=10587.50 10587.50 B
0

Ing. Manuel Snchez Tern

You might also like