Linear Programming

Optimization Techniques in Design

Dr. S. B. Matekar,
Associate Professor,
Pimpri Chinchwad College of Engineering
Unit 2: Linear Programming
Simplex algorithm,

Two phases of the simplex method,

Primal-dual simplex method,

Revised Sim[lex Method

2
Linear Programming (LP)
It is one of the optimization techniques for solving engineering
problems. The term ‘linear’ implies that the objective function and
constraints are ‘linear’ functions of ‘nonnegative’ decision
variables.

The conditions of LP problems (LPP):

1. Objective function must be a linear function of decision variables.
2. Constraints should be linear functions of decision variables.
3. All the decision variables must be non-negative.

3
Graphical solution
An organization is manufacturing two products P1 and P2. The profits per
ton of the two products are ₹ 50 and ₹ 60 respectively. Both the products
require processing in three types of machines. The following table indicates
available machine hours per week and the time required on each machine for
one ton of P1 and P2. Find the product mix that will maximize profit.
maximize
Total available
Machine Product 1 Product 2 machine Constraints,

hours per week

1 2 1 300
2 3 4 509 Non negativity Condition,
3 4 7 812

4
Graphical solution

5
Simplex method

Constraints,
maximize

Constraints,

Objective function

Non negativity Condition,

6
Constraints,

Objective function

7
8
Linear Programming
A company assembles three types of toys, using three operations. Available operation
time for the three operations are 430, 460 and 420 minutes respectively. The profit
per toy is ₹. 3, ₹. 2 and ₹. 5 respectively. The assembly time per operation is shown in
table. Determine optimal production for maximum profit to the company.

maximize

Operation
Toy 1 Toy 2 Toy 3
time Constraints,
1 2 1 430
3 0 2 460
1 4 0 420

Non negativity Condition,

9
Maximization Problem

maximize maximize

Constraints, Constraints,

Non negativity Condition, Non negativity Condition,

10
Minimization Problem

minimize

Constraints,

Non negativity Condition,

11
Artificial variable
Slack variables
Hira Gupta Page 162
• Added when the constraint is ≤
• Negated when the constraint is ≥

Artificial variables
• In many linear programming problems the constraints are ≥ of = type in
such cases the artificial variables are added.
• These variables are fictitious and have no physical meaning.
• They assume the role of slack variables only to be replaced in the later
iterations.
• Thus, they serve the purpose to get the starting basic feasible solution so
that simplex algorithm can be applied.
12
Artificial variable

maximize

Constraints,

Non negativity Condition,

13
Two phase method
Phase I
(minimize)
Hira Gupta Page 204

At the end of nth table the artificial variable driven out and minimization is achieved

Phase II
(maximize)

Start with the last optimized table in Phase I

14
maximize

Constraints, Constraints,

Non negativity Condition, Non negativity Condition,

15
16
Constraints, Constraints,

Non negativity Condition, Non negativity Condition,

17
Primal Dual Simplex Method
Hira Gupta Page 507

For every linear programing problem there is a related

unique linear programing problem involving the same
data of the original problem.

• The original LPP is Primal Program

• Inverting this problem results in Dual Program
• Variables in the dual programme are termed as dual
variables.

18
Primal Dual Simplex Method

19
Primal Dual Simplex Method

20
Primal Dual Simplex Method

21
Primal Dual Simplex Method

22
23
Constraints,

24
Phase I OF minimize
25
minimize maximize

Constraints,

26
 Sensitivity analysis
When the optimal solution of LPP has been attained, two situations may arise:

1. During formulation it is assumed that the constants of the parameters are known,

certain and does not change over the time. But, in actual this is a dynamic situation.

Thus, the current optimal solution changes when parameter changes.

2. Wrong value of a parameter has been used.

3. A design variable or constraint has been omitted while formulating the problem.

A company assembles three types of toys, using three operations. Available operation time for the
three operations are 430, 460 and 420 minutes respectively. The profit per toy is ₹. 3, ₹. 2 and ₹. 5
respectively. The assembly time per operation is shown in table. Determine optimal production for
maximum profit to the company.
Constraints,
Operation
Toy 1 Toy 2 Toy 3
time
maximize
1 2 1 430

3 0 2 460
27
1 4 0 420
 Sensitivity analysis
Changes may be discrete or continuous

• Study of effect of discrete changes in parameters on the optimal

solution is called as Sensitivity analysis or post optimality analysis

• Study of effect of continuous changes in parameters on the

optimal solution is called as Parametric programming

28
Sensitivity analysis

29
Sensitivity analysis

30
 Sensitivity analysis: Case 1

Solve the problem:

Maximize

Subject to

31
32
33