Problem 1

1. Formulate the problem as a non-preemptive Goal Programming.

Parameters :
Capi : the production capacity of the 3 sites
Dj: the demand of the 4 cities
Cijk: transport costs from site Si to city Cj using the means of transport k.
Variable:
Xijk: The quantity of the minerals transported from site Si to city Cj using the means of transport k.
i=1,2,3, j=1,2,3,4, k=1,2

2. Reformulate the problem as a Preemptive GP knowing that the order of priority is as follows:
objectives 1, 2 and 4 (same priority), objective 5, then objective 3.
Problem 2

1) Primal LP

Max Z= 2000X1+3000X2+5000X3+4000X4

s.t.

100X1+100X2+100X3+100X4 ≤ 4000

10X1+10X2+20X3+20X4 ≤ 600

20X1+20X2+30X3+20X4 ≤ 900

20X1+10X2+30X3+30X4 ≤ 700

X1,X2,X3,X4 ≥0

Dual LP

Min W = 4000Y1+600Y2+900Y3+700Y4

s.t.

100Y1+10Y2+20Y3+20Y4≥ 2000

100Y1+10Y2+20Y3+10Y4 ≥ 3000

100Y1+20Y2+30Y3+30Y4 ≥ 5000

100Y1+20Y2+20Y3+30Y4 ≥ 4000

Y1,Y2,Y3,Y4 ≥ 0

2) Primal optimal solution

X1*=0,X2*=25,X3*=10,X4*=5,X5*=0,X6*=50,X7*=0,X8*=0,Z*=145000
 Dual optimal solution

X1 X2 X3 X4 X5 X6 X7 X8 Z*

∆j -1500 0 0 0 -5 0 -100 -50 145000

E1 E2 E3 E4 Y1 Y2 Y3 Y4 W*

1500 0 0 0 5 0 100 50 145000

E1*=1500,E2*=0,E3*=0,E4*=0,Y1*=5,Y2*=0,Y3*=100,Y4*=50,W*=145000

3) Sensitivity analysis to b2

X1 X2 X3 X4 X5 X6 X7 X8 R.H.S

0.5 1 0 0 0.015 0 0 -0.05 25

-5 0 0 0 -0.05 1 0 -0.5 50+∆b2

0 0 1 0 -0.02 0 0.1 0 10

0.5 0 0 1 0.015 0 -0.1 0.5 5

∆j -1500 0 0 0 -5 0 -100 -50

To preserve feasibility and hence optimality of the current basis 50+∆b2≥0. Therefore, ∆b2≥-
50. Thus, b2≥-50+600. ∆b2≥550.

As well as etching hours are at least 550, optimality of the table is preserved.

Sensitivity analysis to c1

As X1 is a non-basic variable in the optimal tableau, only its cbar value will change.
Therefore, to preserve optimality of the current basis -1500+∆c1 ≤ 0. Hence, ∆c1 ≤ 1500. That is,
c1 ≤ 3500.

Sensitivity analysis to c2

X1 X2 X3 X4 X5 X6 X7 X8 R.H.S

0.5 1 0 0 0.015 0 0 -0.05 25

-5 0 0 0 -0.05 1 0 -0.5 50

0 0 1 0 -0.02 0 0.1 0 10
 0.5 0 0 1 0.015 0 -0.1 0.5 5

∆j -1500- 0 0 0 -5- 0 -100 -

0.5∆c2 0.015∆c2 50+0.05∆c2

To preserve optimality of the current basis

−1500 − 0.5∆c2 ≤ 0
{ −5 − 0.015∆c2 ≤ 0
−50 + 0.05∆c2 ≤ 0

−0.5∆c2 ≤ 1500
{ −0.015∆c2 ≤ 5
0.05∆c2 ≤ 50

∆c2 ≥ −3000
{∆c2 ≥ −333,33
∆c2 ≤ 100

-333,33≤∆c2≤100

-333,33+3000≤c2≤100+3000

2666,67≤c2≤3100

4) X3 is a basic variable at the optimal tableau. Therefore, it has a positive optimal value
equals to 10. To remove it, we have to add the constraint X3=0 to the original LP. Hence,
the optimal solution will change and it will degrade given that we are adding a constraint
to the LP.
5) The optimal solution says that product 1 will not be produced as X1=0. Therefore, by
removing it, the optimal solution will not change.
6) Adding a new variable to the primal is equivalent to adding a new constraint to the dual.
This latter is given by the following expression: 100Y1+20Y2+10Y3+10Y4≥p. In order to
introduce this fifth product to the production plan, the dual constraint should be violated
(not satisfied by the current dual optimal solution). Therefore,
100*5+20*0+10*100+10*50≥p
500+0+1000+500≥p
2000≥p
That is, the selling price of product 5 should be at least 2000 Dinars to be profitable.