The Simplex Method: MAXIMIZATION

1. Convert the LP model into its standard form by adding slack variable for <= or subtracting surplus
variable to >= constraints.
2. Set up the initial simplex tableau. (sample tableau)

BV X1 X2 X3 S1 S2 S3 b
S1
S2
S3
Z
3. Determine the entering variable. The entering variable is the most negative coefficient in the
objective row.
4. Determine the leaving or outgoing variable. The leaving variable is the smallest non-negative ratio.
5. Identify the pivot element. The pivot element is the intersection between the pivot row and pivot
column.
6. Set up another table for new iterations.
Determine the new pivot row
NPR = old pivot row / pivot element
Determine also the coefficients of all other rows including the objective row.
New Rows = old rows – (its pivot column coefficient x NPR)
7. Repeat steps 3 to 6 until reaching the optimum solution. Optimum solution is reach when the
coefficients of the objective row are all positive or zero

Slack variable (s)= amount of unused resources

Surplus variable (-s) = amount by which a minimum requirement is exceeded

Example 1. Maximize: z=2 x 1−x 2+2 x 3

Subj to: 2 x1 + x 2 ≤ 10
x 1+ 2 x 2−x 3 ≤ 20
x 2+ 2 x 3 ≤5
x1 , x2 , x3 ≥ 0

Solution: using the basic feasible solution

x 1 , x 2 , x 3 , s1 , s 2 , s3=(0,0,0,10,20,5). The initial simplex tableau and solutions is as follows.

BV X1 X2 X3 S1 S2 S3 b
S1 2 1 0 1 0 0 30
S2 1 2 -2 0 1 0 20
S3 0 1 2 0 0 1 5
Z -2 1 -2 0 0 0 0
BV X1 X2 X3 S1 S2 S3 b
S1 2 1 0 1 0 0 30 PR
S2 1 2 -2 0 1 0 20
S3 0 1 2 0 0 1 5
Z -2 1 -2 0 0 0 0
PC

BV X1 X2 X3 S1 S2 S3 b
X1 1 ½ 0 ½ 0 0 5
S2 0 3/2 -2 -1/2 1 0 15
S3 0 1 2 0 0 1 5 PR
Z 0 3/2 -2 1 0 0 10
PC

BV X1 X2 X3 S1 S2 S3 b
X1 1 ½ 0 ½ 0 0 5
S2 0 5/2 0 -1/2 1 1 20
x3 0 ½ 1 0 0 ½ 5/2
Z 0 3 0 1 0 1 15

It implies that the optimal solution is

x 1 , x 2 , x 3 , s1 , s 2 , s3=(5,0,5/2,0,20,0) and the maximum value of z = 15.

The Simplex Method: MINIMIZATION

Illustration:

Minimize: z=3 x 1+ 2 x 2

Constraints: 2 x1 +3 x 2 ≥ 6

x 1+ 4 x 2 ≥ 7

x1 , x2 ≥ 0

1. Transform the given into augmented matrix corresponding to this minimization problem

2 3 6

[ ]
1st constraints
2nd constraints
1 4 7 Objective row
3 2 0

2. Form the transpose of this matrix, that is the rows will become the columns.
2 1 3

[ ]
3 4 2
6 7 0
1st constraints
2nd constraints
Objective row
3. This implies that the dual maximization is as follows
Maximize: z=6 x 1 +7 x2
Constraints: 2 x1 + x 2 ≤ 3
3 x 1+ 4 x 2 ≤ 2
x1 , x2 ≥ 0
4. To find the optimal solution, follow the steps in maximization.

Example 1: MINIMIZE: 2 x1 +10 x 2+ 8 x 3

Constraints: x 1+ x2 + x 3 ≥ 6
x 2+ 2 x 3 ≥8
- x 1+ 2 x 2 +2 x 3 ≥ 4
x1, x2 , x3≥ 0
Solution:
1. Transform the given into augmented matrix,
1 1 16

2. Form the transpose matrix,

[ ]
0
−1
2
1 28
2 24
10 80
 1 0 −1 2

[ 1 1 2 10
1 2 2 8
6840
]
3. The dual maximization is
Maximize: 6 x 1+ 8 x2 + 4 x 3
Constraints: x 1−x 3 ≤ 2
x 1+ x2 +2 x 3 ≤ 10
x 1+ 2 x 2 +2 x 3 ≤8
x1 , x2 , x3 ≥ 0

4. Set up initial tableau. Follow steps in maximization.

  x1 x2 x3 s1 s2 s3 b
s1 1 0 -1 1 0 0 2 #DIV/0!
s2 1 1 2 0 1 0 10 10
s3 1 2 2 0 0 1 8 4
z -6 -8 -4 0 0 0 0

  x1 x2 x3 s1 s2 s3 b
s1 1 0 -1 1 0 0 2 2
s2 0.5 0 1 0 1 -0.5 6 12
x2 0.5 1 1 0 0 0.5 4 8
z -2 0 4 0 0 4 32

  x1 x2 x3 s1 s2 s3 b
x1 1 0 -1 1 0 0 2
s2 0 0 1.5 -0.5 1 -0.5 5
x2 0 1 1.5 -0.5 0 0.5 3
z 0 0 2 2 0 4 36

Note: The x - values corresponding to this optimal solution are obtained from the entries in the bottom row
corresponding to slack variable columns. Hence, the optimal solution occurs when x1 = 2, x2 = 0, x3 = 4.
And the optimal value is z = 36.

Do the following: Find the optimal solution.

1. Maximize: z=x 1 +2 x 2 + x 3−x 4

Constraints: x 1+ x2 +3 x 3+ 4 x 4 ≤ 60
 2 x2 +2 x 3+5 x 4 ≤ 50
2 x1 +3 x 2+ 6 x 4 ≤ 72
x 1 , x 2 , x 3 , x 4 ≥0

2. Minimize: z=8 x 1 +4 x 2 +6 x 3
Constraints: 3 x 1+2 x 2+ x3 ≥ 6
4 x1 + x 2 +3 x3 ≥ 7
2 x1 + x 2 +4 x 3 ≥ 8
x1, x2 , x3≥ 0