Simplex Method

Introduction:
1
In the previous chapter, we discussed about the graphical method for solving linear programming
problems. Although the graphical method is an invaluable aid to understand the properties of linear
programming models, it provides very little help in handling practical problems. In this chapter, we
concentrate on the simplex method for solving linear programming problems with a larger number of
variables.

Many different methods have been proposed to solve li near programming problems, but simplex method
has proved to be the most effective. This method is applicable to any problem that can be formulated in
terms of linear objective function, subject to a set of linear constraints.

Basic Te rminology

Slack var iable

It is a variable that is added to the left-hand side of a less than or equal to type constraint to convert the
constraint into an equality. In economic terms, slack variables represent left -over or unused capacity.

Specifically:
ai1 x1 + ai2 x2 + ai3 x3 + .........+ ainxn bi can be written as
ai1 x1 + ai2 x2 + ai3 x3 + .........+ ainxn + si = bi

Where i = 1, 2, ..., m

Surplus variable

It is a variable subtracted from the left -hand side of a greater than or equal to type constraint to convert
the constraint into equality. It is also known as negative slack variable. In economic terms, surplus
variables represent over fulfillment of the requirement.

Specifically:
ai1 x1 + ai2 x2 + ai3 x3 + .........+ ainxn bi can be written as
ai1 x1 + ai2 x2 + ai3 x3 + .........+ ainxn - si = bi

Where i = 1, 2, ..., m

Artificial variable

It is a non negative variable introduced to facilitate the computation of an initial basic feasible solution. In
other words, a variable added to the left-hand side of a greater than or equal to type constraint to convert
the constraint into an equality is called an artif icial variable.

Basic Variables and Non-Basic Varia bles: The variables attached to the independent column vectors of
the basis matrix are known as basic variables and the remaining (n- m) variables whose values are
assumed to be zero are know n as non-basic variables.

Basic Solution: Given a system of m simultaneous linear equations containing n variables (n>m) and the
set of equations be AX=b, R(A)=m. If any m*n, non singular matr ix be arbitrating selected from A and if
we assume all (n- m) variables zero which are not associated with column matrix, the solution so obtained
is basic solution.

Non-degene rate basic solution: If all components of a solution set corresponding to a basic variable
are zero quantities then the basic solution is known as non-degenerate basic solution.

DJ
 Simplex Method
Dege nerate basic solution:
2
If some components of the solution set corresponding to the basic variables are zero the basic solution is
know n as degenerate solution.

A Basic feasible solution is said to be degenerate solution if one or more than one basic variable are zero.

Non- Degenerate Solution:

A Non- Degenerate Solution feasible solution is the basic feasible solution which has exactly m positive xi
(i=1,2,…,m) i.e. None of the basic variables are zero.

At what condition Simplex method has an unbounded solution?

a) If corresponding to any negative Z j C j , all elements of Xj column are negative or zero (≤0). Then the
solution is unbounded.
b) If all artificial vectors are driven out from the basis but some non basis victors are less than zero.

At what condition Simplex method has an no feasible solution?

If all Z j Cj 0 but some artif icial variables are present at the positive level in the optimal solution.

At what condition Simplex method has an alternat ive solut ion?

If all Z j Cj 0 , the alternative solution exist if any non-basic Z j C j is also zero.

Simplex Method

Consider the general linear programming problem

Maximize z = c 1 x1 + c2 x2 + c3 x3 + .........+ c nxn

Subject to

a11 x1 + a12 x2 + a13 x3 + .........+ a1 nxn b1

a21 x1 + a22 x2 + a23 x3 + .........+ a2 nxn b2
.........................................................
am1 x1 + am2 x2 + am3 x3 + .........+ amnxn bm
x1 , x2 ,....., xn 0

Where:
c j (j = 1, 2, ...., n) in the objective function are called the cost or profit coefficients.
bi (i = 1, 2, ...., m) are called resources.
aij (i = 1, 2, ...., m; j = 1, 2, ...., n) are called technological coef ficients or input-output coefficients.

Converting inequalities to equalities

Introducing slack variables to convert inequalities to equalities

a11 x1 + a12 x2 + a13 x3 + .........+ a1 nxn + s1 = b1

a21 x1 + a22 x2 + a23 x3 + .........+ a2 nxn + s2 = b2
................................................................

DJ
 Simplex Method
am1 x1 + am2 x2 + am3 x3 + .....+ amnxn + sm = bm
x1 , x2 ,....., xn 0 3
s1 , s2,....., sm 0

An initial basic feasible solution is obtained by setting x1 = x2 =........ = xn = 0

s1 = b1
s2 = b2
..............
sm = bm

The initial simplex table is formed by w riting out the coefficients and constraints of a LPP in a systema tic
tabular form. The following table shows the structure of a simplex table.

Example-01: Using simplex method to solve the following LPP:

Max Z=3x 1 + 2x2

Subject to x1 + x2 4

x1 - x2 2

x1 , x2 0.

Solution: Introducing slack variables x3 and x4 in 1st and 2nd constraints respectively.

The reformulated LPP becomes

Max Z=3x1 + 2x2 +0x3 +0x4

Subject to

x1 + x2 + x3 =4

x1 -x2 + x4 =2, x1 , x2 , x3 ,x4 0.

Using simplex method, we have

Table-

Cj 3 2 0 0
CB XB b y1 y2 y3 y4 Min Ratio
0 x3 4 1 1 1 0 4/1=4
0 x4 2 1 -1 0 1 2/1=2*
Zj Cj 0 -3* -2 0 0

0 x3* 2 0 2 1 -1 2
1*
2

DJ
 Simplex Method
3 x1 2 1 -1 0 1 ----

6 0 -5* 0 3
4
Zj Cj
2 x2 1 0 1 1/2 -1/2
3 x1 3 1 0 1/2 1/2
Zj Cj 11 0 0 5/2 1/2

Since all Zj Cj 0 .Hence the solution is optimal. The optimal solution is Max Z=11 at x1 =3, x2 =1.

Example-02: Using simplex method to solve the following LPP:

Max Z=60x1 + 50x2

Subject to
x1 + 2x2 40

3x1 +2x2 60

x1 , x2 0.

Solution: Introducing slack variables x3 and x4 in 1st and 2nd constraints respectively.

The reformulated LPP becomes

Max Z=60x1 + 50x2 +0x3 +0x4

Subject to

x1 + 2x2 + x3 =40

3x1 +2x2 + x4 =60

x1 , x2 , x3 ,x4 0.

Using simplex method, we have

Table-

Cj 60 40 0 0
CB XB b y1 y2 y3 y4 Min Ratio
0 x3 40 1 2 1 0 40/1=40

0 x4 60 3 2 0 1 20→
Zj Cj 0 -60↑ -50 0 0

0 x3* 20 0 4/3 1 -1/3 1/3

30

DJ
 Simplex Method
60 x1 20 1 2/3 0 1/3
5
Zj C j 120 0 -10↑ 0 20

50 x2 15 0 1 4/3 -1/4

60 x1 10 1 0 -1/2 1/2
Zj Cj 130 0 0 30/4 70/4

Since all Zj Cj 0 .Hence the solution is optimal and the optimal solution is Max Z=1350 at x1 =10,
x2 =15.

Exercise

1. Use simplex algorithm to solve the follow ing LPP: WBUT CS-06

Minimize Z=2x+3y

Subject to 3x+5y 0

5x+3y 60;

X y 0

2. Use Big-M simplex method to solve the problem

Maximize Z = 3x1 x2
Subject to 2x1 + x2 2,
x1 + 3x2 2,
x1 4,
x1 , x2 0.

3. Solve by “Penalty method” (Big-M method)

Maximize Z=2X1 +X2 -3X3
Subject to
X1 +X2 +2X3 ≤5
2X 1 + 3X 2 + 4X 3 = 12 M (CS)-402/07

4. Solve the follow ing LPP by Charnes' Big M- method: MM-40 1/2004-05
Maximize z = 5x1 + 3x2

Subject to 3x1 +2x2 ≤ 18

x1 +x2 ≥7
4x1 + x2 ≤ 13
x1 , x2 ≥0

5. Solve the following L.P.P. by simplex method (A-D)

A. Maximize z 4 x1 7 x 2 B Maximize z 4 x1 7 x 2
Subject to 2 x1 x2 1000, Subject to 2 x1 x2 1000,
10 x1 10 x 2 6000,
10 x1 10 x 2 6000,
2 x1 4 x 2 2000,
2 x1 4 x 2 2000,
x1, x 2 0, DJ
x1, x 2 0,
[ Ans. x1 200, x 2 400; z max 3600.]
 Simplex Method

Check the results of Problems 4 and 5 graphically.

C Maximize z = 3x 1 x 2 3x 3 Minimize D. Minimize z = x 1 3x 2 2x 3

Subject to 2x 1 x 2 x 3 2, Subje Subject to 3x 1 x 2 2x 3 7,
x 1 2x 2 3x 3 5, 2x1 4x 2 12,
2x 1 2x 2 x 3 6, 4x 1 3x 2 8x 3 10,
x 1, x 2 , x 3 0. x 1, x 2 , x 3 0.
[Ans. x1 1
5 , x2 0, x 3 8
5 ; z max
27
5.] [Ans. [Ans. x1 4, , x 2 5, x 3 0; z min 11.]

6. Solve the following L.P.P. by any method.

Maximize z = 7x1 3x 2 7. Use Charnes Big - M method to

Subject to x1 +2 x 2 3 ,
Maximize z 3 x1 x 2
x1 x 2 4,
x1 52 , Subject to 2 x1 x2 2,

x 2 32 , x1 3 x 2 3,
x1, x 2 0. x2 4,
[Ans. x 1 5 2, x 2 3 2 ; z max 22. ] x1, x 2 0,
8. Use Charnes Big - M method to 09. Use Charnes Big - M method (method of penalty ) to
[ Ans. x1 3, x 2 0; z max 9.]
Maximize z x1 5 x 2 Maximize z = 5x 2x 1 3x 2 3
Subject to 3x1 4 x 2 6, Subject to 2x 1 x 2 x 3 2,
x1 3x 2 3, 3x1 4x 2 3,
x1, x 2 0, x 2 3x 3 5,
[ Ans. x1 0, x 2 3 2 ; z max 15 2 .] x 1, x 2 , x 3 0.
[Ans. x 1 23 3 , x 2 5, x 3 0; z max 85
3.]

Pre pared by Dr. Dipa k Kumar Ja na

DJ