Scribd
Upload
49 views25 pages

Dual Simplex

The document provides examples and explanations of different outcomes in linear programming problems: 1) Example 1 demonstrates an optimal solution found through the simplex method with multiple iterations. The optimal solution is x1=2, x2=6, x3=2, x4=0 with an objective value of 36. 2) Example 2 shows an unbounded solution, where the objective function can be increased indefinitely without violating constraints. 3) Example 3 finds that the problem is infeasible, as an artificial variable remains in the basis even when the optimality criterion is satisfied, indicating no feasible solution exists.

Uploaded by

PavanKumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
49 views25 pages

Dual Simplex

The document provides examples and explanations of different outcomes in linear programming problems: 1) Example 1 demonstrates an optimal solution found through the simplex method with multiple iterations. The optimal solution is x1=2, x2=6, x3=2, x4=0 with an objective value of 36. 2) Example 2 shows an unbounded solution, where the objective function can be increased indefinitely without violating constraints. 3) Example 3 finds that the problem is infeasible, as an artificial variable remains in the basis even when the optimality criterion is satisfied, indicating no feasible solution exists.

Uploaded by

PavanKumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 25

INDIANINSTITUTEOFSCIENCE

Water Resources Systems:


Modeling Techniques and Analysis
Lecture - 12
Course Instructor : Prof. P. P. MUJUMDAR
Department of Civil Engg., IISc.

Summary of the previous lecture


Multiple solutions
Non-basic variable in the final tableau, with a
coefficient of zero in the Z-row.
Artificial variables
To ensure an initial basic feasible solution.
Add artificial variable to constraints of the
type > and =.
Penalize artificial variable in the OF.
Modify Row-0 with

RC
EE
P

Example 1
Maximize

Z 3 x1 5 x2
s.t.

x1 4
2 x2 12
3 x1 2 x2 18
x1 0
x2 0
3

Example 1 (Contd.)

Z 3 x1 5 x2 M A1 0
3 x1 2 x2 A1 18

Row-0
Row containing
artificial
variables

x1

x2

x3

x4

bi

-3

-5

18

RC
EE
P
Transformed
Row-0

-3M-3

-2M-5

R = M; P = 1

-18M
4

Example 1 (Contd.)
In case of multiple artificial variables, carryout the
transformation one by one.
Use the transformed Z-row in the initial simplex
table.

Departing variable

Example 1 (Contd.)
Iteration-1

Entering variable

Basis

Row

x1

x2

x3

x4

A1

bi

bi/aij

-3M-3

-2M-5

-18M

x3

x4

12

A1

18

Pivot point
6

Example 1 (Contd.)
Entering variable

Departing variable

Iteration-2

x4

A1

bi

bi/aij

-6M+12

12

-3

Basis

Row

x1

x1

x4

A1

x2

x3

-2M-5 3M+3

Example 1 (Contd.)
Entering variable

Departing variable

Iteration-3
Basis

Row

x1

x2

x3

x4

A1

bi

bi/aij

-9/2

M+5/2

27

x1

x4

-1

x2

-3/2

1/2

Example 1 (Contd.)
Iteration-4
Basis

Row

x1

x2

x3

x4

A1

bi

3/2

M+1

36

x1

-1/3

1/3

x3

1/3

-1/3

x2

1/2

Example 1 (Contd.)
Since all coefficients in the Z-row are non-negative this is
the optimal solution.

Z 36
x1 2
x2 6
x3 2

Note that this is the same


solution as obtained
earlier, with the constraint

3 x1 2 x2 18

x4 0
A1 0

Binding (tight) constraint

10

LP Unbounded Solution
Unbounded solution:
No departing variable can be found at some iteration
Implies that Z can be increased indefinitely without
violating any constraint

11

Example 2
Maximize

Z x1 x2
s.t.

x1 5
x2 10

x1 tends to infinity, therefore


Z tends to infinity;
hence the problem is
unbounded.

x1 0
x2 0
12

Example 2 (Contd.)
x2 = 10

Z x1 x2
x1 5

x1 = 5

x2
Feasible space
unbounded in
the direction in
which Z
increases

x2 10
x1 0
x2 0

x1
13

Example 2 (Contd.)
The problem is converted to standard LP form
Maximize
s.t.

Z x1 x2

x1 5

x1 x3 5

x2 10

x2 x4 10

x1 0

x1 0; x2 0

x2 0

x3 0; x4 0

14

Example 2 (Contd.)
Add artificial variable to constraint 1

Z x1 x2 M A1 0

x1 x3 A1 5
Transformation of coefficients in Row-0
x1

x2

x3

x4

bi

-1

-1

-1

-1-M

-1

-5M

15

Example 2 (Contd.)
Departing variable

Iteration-1

Entering variable

Basis

Row

x1

x2

x3

x4

A1

bi

bi/aij

-1-M

-1

-5M

A1

-1

x4

10

Pivot point
16

Example 2 (Contd.)

Departing variable

Iteration-2

Entering variable

Basis

Row

x1

x2

x3

x4

A1

bi

bi/aij

-1

-1

1+M

x1

-1

x4

10

10

Pivot point
17

Example 2 (Contd.)
Iteration-3

Entering variable

Basis

Row

x1

x2

x3

x4

A1

bi

bi/aij

-1

1+M

-5

x1

-1

x2

10

No departing variable
Hence the problem is unbounded
18

LP Infeasible Solution
Infeasible solution:
One or more artificial variables remain in the basis
even when optimality criterion is satisfied.

19

Example 3
Minimize

Z 3 x1 5 x2
s.t.

x1 x2 4
2 x1 2 x2 32
x1 0
x2 0
20

Example 3 (Contd.)
x2

Z 3 x1 5 x2
x1 x2 4
2 x1 2 x2 32
x1 0
x2 0

No feasible space
can be found

x1
21

Example 3 (Contd.)
The problem is converted to standard LP form
Maximize
s.t.

Z 3x1 5 x2

x1 x2 4

x1 x2 x3 4

2 x1 2 x2 32

2 x1 2 x2 x4 32

x1 0

x1 0; x2 0

x2 0

x3 0; x4 0
22

Example 3 (Contd.)
Add artificial variable to constraint 2

Z 3 x1 5 x2 M A1 0

2 x1 2 x2 x4 A1 32
Transformation of coefficients in Row-0
x1

x2

x3

x4

bi

-3

-5

-1

32

-3-2M

-5-2M

-32M

23

Example 3 (Contd.)
Departing variable

Iteration-1

Entering variable

Basis

Row

x1

x2

x3

x4

A1

bi

bi/aij

-3-2M

-5-2M

-32M

x3

A1

-1

32

16

Pivot point
24

Example 3 (Contd.)
Iteration-2

Basis

Row

x1

x2

x3

x4

A1

bi

5+2M

20-24M

x2

A1

-2

-1

24

All coefficients in the Z-row are non-negative


Artificial variable still remains in the basis; Therefore the
problem is infeasible.

25

You might also like