Scribd
Upload
112 views

4 - Revised Simplex

opti simplex

Uploaded by

f20220818
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
112 views

4 - Revised Simplex

opti simplex

Uploaded by

f20220818
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/ 10

REVISED SIMPLEX METHOD

Dr. Anil Kumar


A-408

The Revised Simplex Method offers an Efficient


Computational procedure for solving Linear Programming
Problems.

The iterative steps of the Revised Simplex Method are


exactly the same as in the Simplex Method Tableau.

The main difference is that computation on the Revised


Simplex Method is based on matrix manipulation rather than
on row operations.

Anil Kumar, BITS Goa 2


Advantages of the Revised Simplex Method
• For LPP, which involves a large number of variables and
large numbers of constraints, the amount of computations is
very much reduced in RSM.
• RSM works with a reduced table as it stores only the basic
variables, the basic inverse and the constants.
• Data can be stored in less space.
• There is less accumulation of round-off errors since no
calculations are carried out on a column unless it is ready to
enter the basis.
• It helps understand duality theory, sensitivity analysis, etc.

Anil Kumar, BITS Goa 3

We know that

Basic z x Solution
j
z 1 cB B-1 Aj - c j c B B -1 b

XB 0 B-1 Aj B-1 b

Anil Kumar, BITS Goa 4


Revised Simplex Algorithm
• Consider the following L.P.P
Max (or Min) z = CX
Subject to
AX = b, X ≥ 0, b ≥ 0.

• Construct a starting basic feasible solution, and let and B be its


associated basis and cB objective coefficients.
• Compute the inverse B-1 by using an appropriate inversion
method.

Anil Kumar, BITS Goa 5

Algorithm cont….

• For each nonbasic variable x j , compute


z j - c j = cBB-1 Aj - c j

If z j - c j ³ 0 in Max. ( ≤ 0 in Min.) for all nonbasic x j stop;


then the optimal solution given by

X B = B-1b,
z = cB X B

Anil Kumar, BITS Goa 6


Algorithm cont……

• Else, determine the entering variable as the nonbasic


variable x j with the most negative (positive) z j - c j in the
case of Max. (Min.)

• Compute B-1 Aj , corresponding to the entering variable x j .

• Compute B-1b, and hence determine the leaving


variable xi using the minimum ratio test.

Anil Kumar, BITS Goa 7

Algorithm cont…..

• From the current basis B, form a new basis by replacing


the leaving vector Ai with the entering vector A j . Go to
step 2 to start a new iteration.

Anil Kumar, BITS Goa 8


Example 1: Consider Problem

max Z = x1+ x2

Subject to

3x1 + 2 x2 ≤ 6
x1 + 4x2 ≤ 4
x1, x2 ≥ 0

Anil Kumar, BITS Goa 9

Introduce Slack and Artificial variables wherever


necessary.

Standard form:-
max Z = x1+ x2
Subject to
3x1 + 2x2 + s1 = 6
x1 + 4x2 + s2 = 4
x1, x2, s1, s2 ≥ 0.

Anil Kumar, BITS Goa 10


A = ( A1 A2 A3 A4 )
Where

æ 3ö æ 2ö æ1 ö æ0ö
A1 = ç ÷ , A2 = ç ÷ , A3 = ç ÷ , A4 = ç ÷
è1 ø è 4ø è 0ø è1 ø

æ6ö
b=ç ÷
è 4ø
æ s1 ö
Starting Basic Variables X B0 = ç ÷
è s2 ø

Anil Kumar, BITS Goa 11

Initial Basis Matrix


æ1 0ö
B0 = ç ÷
è0 1ø

cB0 = ( 0 0 )
Starting BFS

-1 æ 1 0 öæ 6 ö æ 6 ö
X B0 = B0 b = ç ÷ç ÷ = ç ÷
è 0 1 øè 4 ø è 4 ø

Anil Kumar, BITS Goa 12


Starting Table

Entering Pivot Min.


Basis B-1 XB
Variable Column Ratio

s1 1 0 6

s2 0 1 4

Anil Kumar, BITS Goa 13

Find z j - c j for nonbasic variables x1 , x2


z1 - c1 = cB0 B0 -1 A1 - c1
æ 1 0 ö æ 3ö
= ( 0 0) ç ÷ ç ÷ -1
è 0 1 ø è1 ø
= -1

z2 - c2 = cB0 B0 -1 A2 - c2
æ 1 0 öæ 2 ö
= (0 0) ç ÷ç ÷ - 1
è 0 1 øè 4 ø
= -1

Anil Kumar, BITS Goa 14


Take x1 , as entering variable

Pivot column

-1 æ 1 0 öæ 3 ö æ 3 ö
B A1 = ç ÷ç ÷ = ç ÷
è 0 1 øè1 ø è1 ø

Anil Kumar, BITS Goa 15

Entering Pivot Min.


Basis B-1 XB Variable Column Ratio.

s1
1 0 6 3 6/3 =2
x1
s2
0 1 4 1 4/1 =4

Anil Kumar, BITS Goa 16


Example-2
• Use revised simplex method to find the solution of the
following LPP
max z = x1 + 2 x2 - x3
s.t. 2 x1 + x2 + x3 £ 14
4 x1 + 2 x2 + 3 x3 £ 28
2 x1 + 5 x2 + 5 x3 £ 30
x1 , x2 , x3 ³ 0

Anil Kumar, BITS Goa 24

Example-3
• Solve the following Problem by the Revised Simplex
method (together with M-Method)

min z = x1 + x2
s.t. x1 + 2 x2 ³ 7
4 x1 + x2 ³ 6
x1 , x2 ³ 0

Anil Kumar, BITS Goa 25


Example-4
• Solve the following Problem by the Revised Simplex
method (together with Two-Phase method)

min z = x1 + x2
s.t. x1 + 2 x2 ³ 7
4 x1 + x2 ³ 6
x1 , x2 ³ 0

Anil Kumar, BITS Goa 26

Example 5: Solve the following Problem by the Revised


Simplex method given the starting basic feasible vector XB0
= (x2, x4, x5)T

Minimize Z = 7x2 +11x3 -10 x4 + 26x6


Subject to
x2 - x3 + x5 + x6 =6
x2 - x3 + x4 + 3x6 = 8
x1 + x2 - 3x3+ x4 + x5 = 12
x1 , x2 , x3 , x4 , x5 , x6 ≥ 0

Anil Kumar, BITS Goa 27

You might also like