Scribd
Upload
266 views

Linear Programming 1

The document describes conducting sensitivity analysis on a linear programming model by making six changes individually and analyzing the impact. The three sentence summary is: The sensitivity analysis involved changing right-hand side values, objective function coefficients, and constraint coefficients individually and analyzing the impact on feasibility, optimality, and optimal solution values. Most changes did not impact feasibility or optimality, but changing basic variable coefficients could impact both and require reoptimizing the model.

Uploaded by

Özlem Yurtsever
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
266 views

Linear Programming 1

The document describes conducting sensitivity analysis on a linear programming model by making six changes individually and analyzing the impact. The three sentence summary is: The sensitivity analysis involved changing right-hand side values, objective function coefficients, and constraint coefficients individually and analyzing the impact on feasibility, optimality, and optimal solution values. Most changes did not impact feasibility or optimality, but changing basic variable coefficients could impact both and require reoptimizing the model.

Uploaded by

Özlem Yurtsever
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

6.6.

2
Reconsider the model of Prob. 6.6.1. You are now to conduct sensitivity analysis by independently investigating each of the following six changes in the original model. For each change, use the sensitivity analysis procedure to revise the given final set of equations (in tableau form) and convert it to proper form. Then test this solution for feasibility and for optimality (Do not reoptimize) a) Change the RHS of constraint 1 b1=10 a) Change the RHS of constraint 2 b2=10 c) Change the coefficient of x2 in the objective function to c2 = 3 d) Change the coefficient of x3 in the objective function to c3 = 2 e) Change the coefficient of x2 in constraint 2 to a22=2 f) Change the coefficient of x1 in constraint 1 to a11=8
Max Z = 3x1 + x2 + 4x3 St 6x1 + 3x2 + 5x3 25 3x1 + 4x2 + 5x3 20 x1 ,x2 ,x3 0 The optimal simplex tableau OPTIMAL SOLUTION Z x1 x3

x1 0 1 0

x2 2 - 1/3 1

x3 0 0 1

s1 1/5 1/3 - 1/5

s2 3/5 - 1/3 2/5

RHS 17 5/3 3

The optimal Solution X1*=5/3 X2*=0 X3*=3 Z*=17 y1*=1/5 , y2*=3/5, W*=17

a) Change the RHS of constraint 1 b1=10

b1 = 10 b2= 20 B-1=
1/3 - 1/5 b*= B-1 b 1/3 - 1/5 - 1/3 2/5 10 20 - 1/3 2/5

b*= -10/3 6

The current basic solution is infeasible

b) Change the RHS of constraint 1 b2=10

b1 = 25 b2= 10 B-1=
1/3 - 1/5 b*= B-1 b 1/3 - 1/5 - 1/3 2/5 25 10 - 1/3 2/5

b*= 5 -1

The current basic solution is infeasible

c) Change the coefficient of x2 in the objective function to c2 = 3


x2 is a non basic variable. If there is a change in the coefficient of a nonbasic variable this will show itself in the dual. PRIMAL Max Z = 3x1 +3 x2 + 4x3 St 6x1 + 3x2 + 5x3 25 3x1 + 4x2 + 5x3 20 x1 ,x2 ,x3 0 DUAL Min W = 25y1 + 20y2 St 6y1 + 3y23 3y1 + 4y23 5y1 + 5y24 y1 ,y20 Only the RHS of the second constraint of the dual problem is changing so we should look to that constraint. Our optimal solution was y1*=1/5 , y2*=3/5 And if we write them in to the second constraints equation: 3*1/5 + 4*3/5 =3 . So it proves the new inequality 3y1 + 4y23 Optimality has not changed. The current basic solution stays optimal.

d) Change the coefficient of x3 in the objective function to c3 = 2


x3 is a basic variable. If there is a change in the objective function coefficient optimality may change. PRIMAL Max Z = 3x1 + x2 + 2x3 St 6x1 + 3x2 + 5x3 25 3x1 + 4x2 + 5x3 20 x1 ,x2 ,x3 0 DUAL Min W = 25y1 + 20y2 St 6y1 + 3y23 3y1 + 4y21 5y1 + 5y22 y1 ,y20 Updated reduced cost on the optimal simplex tableu under x3 will be calculated : z3 c3=y*a3-c3 y*= (1/5 3/5) a3 = 5 5

c3=2 y*a3-c3=4-2=2 Value of Z will change but it will stay optimal since the reduced cost is still bigger than zero

e) Change the coefficient of x2 in constraint 2 to a22=2


x2 is a non basic variable. If there is a change in the coefficient of a nonbasic variable this will show itself in the dual. PRIMAL Max Z = 3x1 + x2 + 4x3 St 6x1 + 3x2 + 5x3 25 3x1 + 2x2 + 5x3 20 x1 ,x2 ,x3 0 DUAL Min W = 25y1 + 20y2 St 6y1 + 3y23 3y1 + 2y21

Only the third constraint of the dual problem is changing so we should look to that constraint. Our optimal solution was y1*=1/5 , y2*=3/5 And if we write them in to the thirdconstraints equation: 3*1/5 + 2*3/5 =9/5 . So it proves the new inequality 3y1 + 2y21 Optimality has not changed. The current basic solution stays optimal.

f) Change the coefficient of x1 in constraint 1 to a11=8


x1 is a basic variable. If there is a change in the constraints coefficient feasibilitiy may change. PRIMAL Max Z = 3x1 + x2 + 4x3 St 8x1 + 3x2 + 5x3 25 3x1 + 4x2 + 5x3 20 x1 ,x2 ,x3 0 DUAL Min W = 25y1 + 20y2 St 8y1 + 3y23 3y1 + 4y21 5y1 + 5y24 y1 ,y20 4

Updated reduced cost on the optimal simplex tableu under x3 will be calculated : z1 c1=y*a1-c1 y*= (1/5 3/5) a1 = 8 3

c1=3 y*a1-c1=17/5 3 = 2/5 Value of Z will change but it will stay feasible and optimal since the reduced cost is still bigger than zero

You might also like