Linear

programming

DR. SHAILESH MATHUR

Associate Professor
Department of Accountancy and Business Statistics
S. S. Jain Subodh P. G. (Aoutonomous) College. Jaipur
Introduction of Linear Programming
The technique of linear programming was formulated by a
Russian mathematician L.V. Kantorovich. But the present version
of simplex method was developed by Geoge B. Dentzig in 1947.
Linear programming consists of two words ‘Linear’ and
‘Programming’. The world linear stand for indicating the
relationships between different variables of degree one, whereas
another word programming means planning and refers to the
proces of selecting best course of action from various
alternatives. Thus, linear programming is a mathematical
technique for allocating limited resources is optimum manner.
In the other words Linear programming is a planning technique
that permits some objective function to be minimized or
maximized within the framework of given situational restrictions.
DSM_Linear_Programming 2
Characteristics of Linear Programming
1. Objective function (In business problems the objective is
generally profit maximization or cost minimization)
2. Constraints (limitations) regarding resources in
mathematical form.
3. The value of variables must be zero or positive and not
negative.
4. Linearity: The relationships between variables must be
linear. Linear means proportional relationship between two
‘or more variable, i.e., the degree of variables should be
maximum one.
5. The number of inputs and outputs need to be finite.

DSM_Linear_Programming 3
Assumptions of Linear Programming
1. There are a number of constraints or restrictions
expressible in quantitative terms.
2. The prices of input and output both are constant.
3. The relationship between objective function and
constraints are linear.
4. The objective function is to be optimized i.e.,
profit maximization or cost minimization

DSM_Linear_Programming 4
 Advantages of Linear Programming
1. LP makes logical thinking and provides better
insight into business problems.
2. Manager can select the best solution with the help
of LP by evaluating the cost and profit of various
alternatives.
3. LP provides an information base for optimum alloca­
tion of scarce resources.
4. LP assists in making adjustments according to
changing conditions.
5. LP helps in solving multi-dimensional problems.
DSM_Linear_Programming 5
 Limitations of Linear Programming
1. This technique could not solve the problems in which variables
cannot be stated quantitatively.
2. In some cases, the results of LP give a confusing and misleading
picture.
3. LP technique cannot solve the business problems of non-linear
nature.
4. The factor of uncertainty is not considered in this technique.
5. This technique is highly mathematical and complicated.
6. If the numbers of variables or constrains involved in LP
problems are quite large, then using costly electronic computers
become essential, which can be operated, only by trained
personnel.
7. Under this technique to explain clearly the objective function is
difficult.
DSM_Linear_Programming 6
 Managerial uses and applications of
Linear Programming
1. Optimizing the product mix when the production line
works under certain specification;
2. Securing least cost combination of inputs;
3. Selecting the location of Plant;
4. Deciding the transportation route;
5. Utilizing the storage and distribution centres;
6. Proper production scheduling and inventory control;
7. Solving the blending problems;
8. Minimizing the raw materials waste;
9. Assigning job to specialized personnel.
DSM_Linear_Programming 7
 Graphical method

Steps in graphical solution of linear programming

problem (LPP) are as the following :
1. Formulate LPP by writing the objective function
(generally maximize profit or minimize cost) and
the constraints.
2. Constraints are changed into equalities.
3. Plot the constraints on the graph.
4. Identify the feasible region and ascertain their
coordinates.
5. Test which point is most profitable.
DSM_Linear_Programming 8
Example 1:
A company manufactures two types of boxes, A and B. the
boxes undergo two major processes: cutting and painting
operation. The profits per unit are Rs. 6 and Rs. 4
respectively. Each A box requires 2 minutes for cutting and
3 minutes for painting operation, whereas each B box
requires 2 minutes for cutting and 1 minute for painting.
The available operation time is 120 minutes and 60
minutes for cutting and painting machine respectively.
Determine the optimum quantities of the two boxes to
maximize the profits.

DSM_Linear_Programming 9
Solution:
Step I: Formulate LPP:
Let x1 be the number of boxes A to be manufactured and x2 be
the number of boxes B to be manufactured.
Objective Function: The objective is to maximize the profits.
Given profits on A box and B box are Rs. 6 and Rs. 4
respectively.
Zmax = 6x1 + 4x2
Constraints: The available machine hours for each machine
and the time consumed by each product are given. Therefore,
the constrains are:
2x1 + 3x2 ≤ 120 …. (i)
2x1 + x2 ≤ 60 …. (ii)
Where x1, x2 ≥ 0
DSM_Linear_Programming 10
Step II: Constraints are changed into equalities:
The inequality constraints are removed by replacing equal to sing to give the
following equations:
2x1 + 3x2 = 120 …. (i)
2x1 + x2 = 60 …. (ii)
Find the co-ordinates of the lines by substituting x1 = 0 and x2 = 0 in each
equation.
In equation (i), put x1 = 0 to get x2:
2 (0) + 3x2 = 120; x2 = 40
now put x2 = 0 to get x1:
2x1 + 3 (0) = 120;x1 = 60
Similarly in equation (ii), put x1 = 0 to get x2:
2 (0) + x2 = 60; x2 = 60
now put x2 = 0 to get x1:
2x1 + (0) = 60; x1 = 30
The lines are drawn on a graph with horizontal and vertical
axis representing boxesDSM_Linear_Programming
x1 and x2 respectively. 11
 Step III: Plot the constraints on the graph:
The line 2x1 + 3x2 = 120 passes through co-ordinates (0, 40) and (60, 0)
The line 2x1 + x2 = 60 passes through co-ordinates (0, 60) and (30, 0)
Now, the objective is to maximize the profit. The
point that lies at the furthermost point of the
feasible area will give the maximum profit. To
locate the point, we need to plot the objective
70

60 C (2x1 + x2 = 60)
function (profit) line.
50
At point ‘A’ (0, 40) Zmax = 6 (0) + 4 (40) =160
A
40
E
At point ‘D’ (30, 0) Zmax = 6 (30) + 4 (0) = 180
30 (2x1 + 3x2 = 120 )
20 At point ‘E’ (15, 30) Zmax = 6 (15) + 4 (30) = 210
10
D B
Therefore, we conclude that to maximize profit,
0
10 20 30 40 50 60 70
15 numbers of corrugated boxes and 30 numbers
of carton boxes should be produced to get a
maximum profit. Substituting X1 = 15 and X1 =
30 in objective function, we get maximum profit:
DSM_Linear_Programming 12
Rs. 210.00
Example 2:
A company produces two products A and B. These products are produce
with the help of two material X and Y. At least 16 kgs of X and 24 kgs
of Y is used in a day. One unit of product A required 2 kgs of X and 4
kgs of Y. One unit of product B required 4 kgs of X and 3 Kgs of Y.
Cost per unit of product A and B are Rs. 6 and Rs 3 respectively.
Find the optimum solution with the help graphical method of LLP
model.
Solution 2:
Decision variable:
X1 = one unit of A and X2 = one unit of B
Objective Function:
Minimize Z = 6 X1 + 3 X2
Constraints:
2 X1 + 4X2 ≥ 16 …. (i)
4 X1 + 3X2 ≥ 24 …. (ii)
Where X1, X2 ≥ 0
DSM_Linear_Programming 13
Constraints are changed into equalities:
The inequality constraints are removed by replacing equal to sing to
give the following equations:
2x1 + 4x2 = 16 …. (i)
4x1 + 3x2 = 24 …. (ii)

Find the co-ordinates of the lines by substituting x 1 = 0 and x2 = 0 in

each equation.

In equation (i), put x1 = 0 to get x2 and put x2 = 0 to get x1:

2 (0) + 4x2 = 16; x2 = 4
2x1 + 4 (0) = 16; x1 = 8

Similarly in equation (ii), put x 1 = 0 to get x2 and put x2 = 0 to get x1:

4 (0) + 3x2 = 24; x2 = 8
4x1 + 3(0) = 24; x1 = 6
DSM_Linear_Programming 14
 Plot the constraints on the graph:
The line 2x1 + 4x2 = 16 passes through co-ordinates (0, 4) and (8, 0)
The line 4x1 + 3x2 = 24 passes through co-ordinates (0, 8) and (6, 0)

Now, the objective is to minimize the cost.

The point that lies at the furthermost point of
12 the feasible area will give the maximum
10 profit. To locate the point, we need to plot the
Product B

8 C
objective function (profit) line.
(2x1 + 4x2 = 16)
At point ‘A’ (0, 8) Z min = 6 (0) + 3 (8) = 24
6

4
At point ‘D’ (8, 0) Z min = 6 (8) + 3 (0) = 48
A
(4x1 + 3x2 = 24 )

2
At point ‘E’ (5, 2) Z min = 6 (5) + 3 (2) = 36
E

0 D B Therefore, we conclude that to minimize cost

2 4 6 8 10 12
Product A Rs. 24 at the point A , where Zero numbers of
product A and 8 numbers of product will be
produced.
DSM_Linear_Programming 15
Question 3:
A Television Company has three major departments for producing its two models A and
B. The relevant information is given below:

Time for any Unit

Department Capacity
A B
I 4 2 1600
II 2.5 1 1200
III 4.5 1.5 1600
Per unit profit 400 100
Find the graphically that for maximum profit how many units of each model must be
produce.
Solution:
Decision variable:
X1 = one unit of A and X2 = one unit of B
Objective Function:
Maximize Z = 400 X1 + 100 X2
Constraints:
For department I 4 X1 + 2X2 ≤ 1600 …. (i)
For department II 2.5 X1 + X2 ≤ 1200 …. (ii)
For department III 4.5 X1 + 1.5X2 ≤ 1600 …. (ii)
Where X1, X2 DSM_Linear_Programming
≥ 0 16
Constraints are changed into equalities:
The inequality constraints are removed by replacing equal to sing to
give the following equations:
For department I 4 x1 + 2x2 = 1600 …. (i)
For department II 2.5x1 + x2 = 1200 …. (ii)
For department III 4.5x1 + 1.5x2 = 1600 …. (ii)

Find the co-ordinates of the lines by substituting x 1 = 0 and x2 = 0 in

each equation.
In equation (i), put x1 = 0 to get x2 and put x2 = 0 to get x1:
4(0) + 2x2 = 1600; x2 = 1600/2 = 800
4x1 + 2(0) = 1600; x1 = 1600/4 = 400
In equation (ii), put x1 = 0 to get x2 and put x2 = 0 to get x1:
2.5(0) + x2 = 1200; x2 = 1200
2.5x1 + (0) = 1200; x1 = 1200/ 2.5 = 480
In equation (iii), put x1 = 0 to get x2 and put x2 = 0 to get x1:
4.5(0) + 1.5x2 = 1600; x2 = 1600/ 1.5 =1067
DSM_Linear_Programming 17
4.5x + 1.5(0) = 1600; x = 1600/4.5 = 356
 4 x1 + 2x2 = 1600; x1 = 400 & x2 = 800
1400 2.5x1 + x2 = 1200; x1 = 480 & x2 = 1200
4.5x1 + 1.5x2 = 1600 ; x1 = 357 & x2 = 1067
1200 C
E
1000 Objective is maximize the profit.
At point ‘A’ (0, 800) Z max
A = 400 (0) + 100(800) = 80,000
800
Product B

At point ‘F’ (357, 0) Z max

600 = 400 (357) + 100(0) = 1,42,800
At point ‘G’ (200, 400) Z max
400 G
= 400(200) + 100(400) = 1,20,000
200

BD
F
200 400 600 800 1000 1200 1400
Product A
DSM_Linear_Programming 18
 Simplex method
Simplex method of Linear Programming was
developed by George B. Dantzig in 1947. This method
provides the Technique of solving linear programming
problem of any magnitude where there are two or more
decision variables.
The simplex method is based on the property that “the
optimal solution of a linear programming problem
(if exists) will always be given by one of the corner
points (or extreme points representing basic feasible
solution) of the feasible region.”

DSM_Linear_Programming 19
 Basic Terms Involved in Simplex Method
1. Slack Variable: It is a variable that is added to the left hand side of
a less than or equal to (≤) constraint to convert into an equality. For
example:
If a constraint is 2x1 + 3x2 ≤ 120, then by added slack variable (S) it
convert into equality equation and written as:
2x1 + 3x2 + S1 = 120

2. Surplus Variable: It is a variable that is subtracted to the left hand

side of a more than or equal to (≥) constraint to convert into an
equality. It is also known as negative slack variable. For example:
If a constraint is 4x1 + 2x2 ≥ 1600, then by subtracting surplus
variable (S) it convert into equality equation and written as:
4x1 + 2x2 – S2 = 1600

DSM_Linear_Programming 20
Cont…
3. Artificial Variable: It is a non -negative variable introduce to
facilitate the computation of an initial basic feasible solution.
In other words, it is a variable that is added to the left hand
side of a more than or equal to (≥) constraint to convert into
an equality. For example:
If a constraint is 4 X1 + 2X2 ≥ 1600, then by subtracting
surplus variable (S) and added artificial variable (A) it
convert into equality equation and written as:
4x1 + 2x2 – S2 + A1 = 1600

DSM_Linear_Programming 21
Use of Slack Variable, Surplus Variable and
Artificial Variable in mixed type constrains

Value (Cost or Profit) to be

Slack, Surplus and added in objective equation
Sign Artificial Variable use
Maximum Minimi.
in constrain equation
Problem Problem
≤ +S +0S +0S
≥ -S +A +0S -MA +0S +MA
= +A -MA +MA

DSM_Linear_Programming 22
Maximization Linear Programming Problem
Problem:
A factory uses three different resources for the manufacture of two different products,
20 units of the resource A, 12 units of B and 16 unit of C being available. 1 unit of the
first product requires 2, 2 and 4 units of the respective resources and 1 unit of the
second product requires 4, 2 and 0 units of the respective resources. It is know that the
first product gives a profit of Rs. 2 per unit and the second product gives a profit of
Rs. 3 per unit. Formulate the LPP. How many units of each should be manufactured
for maximizing the profit.

Resources Product I Product II Available Units

A 2 4 20
B 2 2 12
C 4 0 16
Profit per unit 2 3

DSM_Linear_Programming 23
Formulation of Problem
Decision variable:
One unit of Product I = X1 and
One unit of Product II = X2
Objective Equation:
Maximize Z = 2X1 + 3X2
Constrain Equation
For resource A 2X1 + 4X2 ≤ 20 …. (i)
For resource B 2X1 + 2X2 ≤ 12 …. (ii)
For resource C 4X1 ≤ 16 …. (iii)
Where X1, X2 ≥ 0
Changing unequal equations into equal equation
Maximize Z = 2X1 + 3X2 + 0S1+ 0S2 + 0S3
Subject to 2X1 + 4X2 + S1 = 20
2X1 + 2X2 + S2 = 12
4X1 + S3 = 16
Where X1, X2, S1, S2, S3 ≥ 0
DSM_Linear_Programming 24
 Designing the initial Simplex Table
Z = 2X1 + 3X2 + 0S1+ 0S2 + 0S3

Contri- 2 3 0 0 0
Variable Quantity /
bution Real Variable Column Ratio
Column Column Value
(Cj) Column X1 X2 S1 S2 S3

0 S1 20 2 4 1 0 0
0 S2 12 2 2 0 1 0

0 S3 16 4 0 0 0 1
Zj

Net Evaliation Raw (Cj - Zj )

2X1 + 4X2 + S1 = 20 …(i) 2X1 + 2X2 + S2 = 12 …(ii)

4X1 + S3 = 16 …(iii)

DSM_Linear_Programming 25
Calculation of Zj Raw Value = [∑(Entry in contribution column x
Corresponding entries in Variable column)]
Calculation of Cj – Zj Raw Value = Cj value of column – Zj value of column

2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3

0 S1 20 2 4 1 0 0
0 S2 12 20 02 00 01 0
0 S3 16 24 30 00 00 01
Zj Zj Value Cj – Zj Value
X1 -Column
Net Evaliation Raw (C (0 x 2) + (0 x 2) + (0 x 4) = 0 2-0=2
j Zj )
X2 Column (0 x 4) + (0 x 2) + (0 x 0) = 0 3-0=3
S1 Column (0 x 1) + (0 x 0) + (0 x 0) = 0 0-0=0
S2 Column (0 x 0) + (0 x 1) + (0 x 0) = 0 0-0=0
S3 Column (0 x 0) + (0 x 0) + (0 x 1) = 0 0-0=0
DSM_Linear_Programming 26
 2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3

0 S1 20 2 4 1 0 0
0 S2 12 2 2 0 1 0
0 S3 16 4 0 0 0 1
3
Zj 0 0 0 0 0
Identification
Net Evaliation of
RawKey
(Cj Column:
- Zj ) 2 3 0 0 0
In net evaluation raw, the column showing the maximum positive
value is known as KEY COLUMN.

DSM_Linear_Programming 27
 Outgoing Variable: Variable of Key Raw Incoming Variable: Variable of Key Column

2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3
5
6
0 S1 20 2 4 1 0 0
∞
0 S2 12 2 2 0 1 0
0 S3 16 4 0 0 0 1
Zj 0 0 0 0 0
Identification of Key Raw:
Net Evaliation Raw (Cj - Zj ) 2 3 0 0 0
Select that raw as key raw which gives least positive ratio.
Value of Quantity or Value column Ratio for S1 Raw = 20 / 4 = 5
Ratio = Ratio for S2 Raw = 12 / 2 = 6
Corresponding value of Key column
Ratio for S3 Raw = 16 / 0 = ∞
Key Number: The entry which lies at the intersection of key raw and
key column is called KEY NUMBER.
DSM_Linear_Programming 28
 Development of II Simplex Table
In the basic variable column of II simplex table, outgoing variable (S1)
replace by incoming variable (X2) and the other basic variable and their
position will be same.
Transformation of Key Raw
Old entry of Key Raw
New values of Key
=
Raw Key Number
Key Raw Volume Col X1 X2 S1 S2 S3
Old Entry 20 2 4 1 0 0
Key Number 4 4 4 4 4 4
New Entry 20/4 = 5 2/4 = 1/2 4/4 = 1 1/4 0/4 =0 0/4 =0

Transformation of Non-Key Raw

Corresponding Entry of Key Raw x
Corresponding entry of Key Column
New values of Old
= -
Non-Key Raw Entry Key Number
DSM_Linear_Programming 29
 2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3

0 S1 20 2 4 1 0 0 5
0 S2 12 2 2 0 1 0 6
0 S3 16 4 0 0 0 1 ∞
Zj 0 0 0 0 0
Transformation of Non-Key
Net Evaliation Raw (CjRaw
- Zj ) S2 : 2 3 0 0 0
Vol Col X1 X2 S1 S2 S3

Old Entry 12 2 2 0 1 0

Corresponding
20 2 4 1 0 0
Entry of Key Raw
Corresponding
2 2 2 2 2 2
entry of Key Col.
Key Number 4 4 4 4 4 4

12- [(20 x 2- [(2 x 2) / 2- [(4 x 0- [(1 x 1- [(0 x 0- [(0 x

New Entry 4] = 1
2) / 4] = 2 2) / 4] = 0 2) / 4] = - 2) / 4] = 1 2) / 4] = 0
1/2
DSM_Linear_Programming 30
 2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3

0 S1 20 2 4 1 0 0 5
0 S2 12 2 2 0 1 0 6
0 S3 16 4 0 0 0 1 ∞
Zj 0 0 0 0 0
Transformation of Non-Key
Net Evaliation Raw (CjRaw
- Zj ) S3 : 2 3 0 0 0
Vol Col X1 X2 S1 S2 S3

Old Entry 16 4 0 0 0 1

Corresponding
20 2 4 1 0 0
Entry of Key Raw
Corresponding
0 0 0 0 0 0
entry of Key Col.
Key Number 4 4 4 4 4 4

16- [(20 x 4- [(2 x 0) / 0- [(4 x 0- [(1 x 0- [(0 x 1- [(0 x

New Entry 4] = 4
0) / 4] = 16 0) / 4] = 0 0) / 4] = 0 0) / 4] = 0 0) / 4] = 1
DSM_Linear_Programming 31
 Key Column II Simplex Table Key Raw

2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
Key Number

(Cj) X1 X2 S1 S2 S3 5/(1/2) = 10
2/1 = 2
3 X2 5 1/2 1 1/4 0 0
16/4 = 4
0 S2 2 1
Zj 3/2 30 -1/2
3/4 01 00
S3 Raw (C 16
Net0 Evaluation 4
1/2 00 0
-3/4 00 01
j - Zj )

Zj Raw Value = [∑(Entry in Cj column x Net Evaluation Raw = Cj - Zj

Corresponding entries in Variable column)] X1 Column = 2 – 3/2 = 1/2
X1 Column = (3 x 1/2) + (0 x 1) + (0 x 4) = 3/2 X2 Column = 3 - 3 = 0
X2 Column = (3 x 1) + (0 x 0) + (0 x 0) = 3 S1 Column = 0 – 3/4 = -3/4
S1 Column = (3 x 1/4) + (0 x -1/2) + (0 x 0) = 3/4 S2 Column = 0 - 0 = 0
S2 Column = (3 x 0) + (0 x 1) + (0 x 0) = 0 S3 Column = 0 - 0 = 0
S3 Column = (3 x 0) + (0 x 0) + (0 x 1) = 0 Value of Quantity or Value column
Ratio =
Corresponding value of Key column
DSM_Linear_Programming 32
 Development of III Simplex Table
In the basic variable column of III simplex table, outgoing variable (S2)
replace by incoming variable (X1) and the other basic variable and their
position will be same.
Transformation of Key Raw
Old entry of Key Raw
New values of Key
=
Raw Key Number
Volume X1 X2 S1 S2 S3
Key Raw
Col
Old Entry 2 1 0 -1/2 1 0
Key Number 1 1 1 1 1 1
(-1/2/1) =
New Entry 2/1 = 2 1/1 = 1 0/1 = 0 1/1 =1 0/1 = 0
-1/2
Transformation of Non-Key Raw
Corresponding Entry of Key Raw x
Corresponding entry of Key Column
New values of Old
= -
Non-Key Raw Entry Key Number
DSM_Linear_Programming 33
 2 3 0 0 0
Quantity / Real Variable Column
Contri-
bution Variable Value
Ratio
Column (Cj) Column Column
X1 X2 S1 S2 S3

3 X2 5 1/2 1 1/4 0 0 10
0 S2 2 1 0 -1/2 1 0 2
0 S3 16 4 0 0 0 1 4
Zj 3/2 3 3/4 0 0
Transformation of Non-Key Raw X2:
Net Evaliation Raw (Cj - Zj ) 1/2 0 -3/4 0 0
Vol Col X1 X2 S1 S2 S3

Old Entry 5 1/2 1 1/4 0 0

Corresponding
2 1 0 -1/2 1 0
Entry of Key Raw
Corresponding
1/2 1/2 1/2 1/2 1/2 1/2
entry of Key Col.

Key Number 1 1 1 1 1 1

1/4-[( - 0- [(1 x 0- [(0 x

New Entry 5- [(2 x 1/2) / 1/2- [(1 x 1/ 1-(0x1/2) / 1/2x1/2) / 1/2) / 1] = 1/2) / 1] =
1] = 4 2) / 1] = 0 1] = 1 1] = 1/2 0
DSM_Linear_Programming -1/2 34
 2 3 0 0 0
Quantity / Real Variable Column
Contri-
bution Variable Value
Ratio
Column (Cj) Column Column
X1 X2 S1 S2 S3

3 X2 5 1/2 1 1/4 0 0 10
0 S2 2 1 0 -1/2 1 0 2
0 S3 16 4 0 0 0 1 4
Zj 3/2 3 3/4 0 0
Transformation of Non-Key Raw S3:
Net Evaliation Raw (Cj - Zj ) 1/2 0 -3/4 0 0
Vol Col X1 X2 S1 S2 S3

Old Entry 16 4 0 0 0 1

Corresponding
2 1 0 -1/2 1 0
Entry of Key Raw
Corresponding
4 4 4 4 4 4
entry of Key Col.

Key Number 1 1 1 1 1 1

0 - [( -1/2
New Entry 16- [(2 x 4) / 4- [(1 x 4/ 2) 0 - (0 x x 4) / 1] = 0- [(1 x 1- [(0 x
1] = 8 / 1] = 0 4) / 1] = 0 2 4) / 1] = -4 4) / 1] = 1
DSM_Linear_Programming 35
 III Simplex Table
2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3

3 X2 4 0 1 1/2 -1/2 0
2 X1 2 12
Zj 30 -1/2
1/2 1
1/2 00
S3 Raw (C -8Z )
Net0 Evaluation 00 00 2
-1/2 -4
-1/2 01
j j

Zj Raw Value = [∑(Entry in Cj column x Net Evaluation Raw = Cj - Zj

Corresponding entries in Variable column)] X1 Column = 2 –2 = 0
X1 Column = (3 x 0) + (2 x 1) + (0 x 0) = 2 X2 Column = 3 - 3 = 0
X2 Column = (3 x 1) + (2x 0) + (0 x 0) = 3 S1 Column = 0 – 1/2 = -1/2
S1 Column = (3 x 1/2) + (2 x -1/2) + (0 x 2) = 1/2 S2 Column = 0 – 1/2 = -1/2
S2 Column = (3 x -1/2) + (2 x 1) + (0 x -4) = 1/2 S3 Column = 0 - 0 = 0
S3 Column = (3 x 0) + (2 x 0) + (0 x 1) = 0

DSM_Linear_Programming 36
 III Simplex Table
2 3 0 0 0
Contri- Quantity Real Variable Column
bution Variable / Value
Column Ratio
Column Column
(Cj) X1 X2 S1 S2 S3

3 X2 4 0 1 1/2 -1/2 0
2 X1 2 1 0 -1/2 1 0
0 S3 8 0 0 2 -4 1
Zj 2 3 1/2 1/4 0
Since all the entries of row in table III are negative or Zero, this shows
Net Evaluation Raw (Cj - Zj ) 0 0 -1/2 -1/4 0
that current programme cannot be improved further and hence it is the
optimal programme.
The optimal programme is:
X1 = 2 and X2= 4
Maximize Profit: Maximize Z = 2X1 + 3X2
Z = 2 (2) + 3 (4) = 16
DSM_Linear_Programming 37
Problem:
A farmer has 1,000 acres of land on which he can grow corn, wheat or soyabean,
Each acre of corn costs Rs. 100 for preparation, requires 7 man days of work an
yield a profit of Rs. 30. An acre of wheat costs Rs. 120 to prepare, requires 10
man days work and yields a profit Rs. 40. An acre of soyabeans costs 70 to
prepare, requires 8 man days and yields a profit of Rs. 20. If the farmer has
Rs.1,00,000 for preparation and can count on 8,000 man days of work how many
acres should be allocated to each crop to maximize profit. Formulate the above
problem as lines programming problem.

Available
Resources Corn Wheat Soyabean
Units
Preparation
10 12 7 10,000
Cost
Land 1 1 1 1000
Man Days 7 10 8 8,000
Profit per unit 30 40 20

DSM_Linear_Programming 38
Formulation of Problem
Decision variable:
One unit of Corn = X1; One unit of Wheat = X2 and
One unit of Soyabean = X3
Objective Equation:
Maximize Z = 30X1 + 40X2 + 20X3
Constrain Equation
For Preparation 10X1 + 12X2 + 7X3 ≤ 10,000 …. (i)
For Land X1 + X2 + X3 ≤ 1,000 …. (ii)
For Man Days 7X1 + 10X2 + 8X3 ≤ 8,000 …. (iii)
Where X1, X2, X3 ≥ 0
Changing unequal equations into equal equation
Maximize Z = 30X1 + 40X2 + 20X3 + 0S1+ 0S2 + 0S3
Subject to 10X1 + 12X2 + 7X3 + S1 = 10,000
X1 + X2 + X3 + S2 = 1,000
7X1 + 10X2 + 8X3 + S3 = 8,000
Where X1, X2, X3, S1, S2, S3 ≥ 0
DSM_Linear_Programming 39
 Designing the initial Simplex Table I
Z = 30X1 + 40X2 + 20X3 + 0S1+ 0S2 + 0S3 X1 + X2 + X3 + S2 = 1,000
10X1 + 12X2 + 7X3 + S1 = 10,000 7X1 + 10X2 + 8X3 + S3 = 8,000
30 40 20 0 0 0
(Cj) Quantity /
Column Var.
Value Real Variable Column Ratio
Col.
Column X1 X2 X3 S1 S2 S3

0 S1 10,000 10 12 7 1 0 0 10,000/12=833
Key Number

0 S2 1,000 1 1 1 0 1 0 1,000/1=1,000

0 S3 8,000 7 10 8 0 0 1 8,000/10=800

Zj 0 0 0 0 0 0
Net Evaliation Raw (Cj - Zj ) 30 40 20 0 0 0

Zj Value X1 Column (0 x 10) + (0 x 1) + (0 x 7) = 0 30 - 0 = 30

Cj – Zj
Key Column X2 Column (0 x 12) + (0 x 1) + (0 x 10) = 0 40 - 0 = 40
Value
X3 Column (0 x 7) + (0 x 1) + (0 x 8) = 0 20 - 0 = 20
Key Raw S1 Column (0 x 1) + (0 x 0) + (0 x 0) = 0
0-0=0
S2 Column (0 x 0) + (0 x 1) + (0 x 0) = 0
0-0=0
S3 Column (0 x 0) + (0 x 0) + (0 x 1) = 0
DSM_Linear_Programming 0 - 0 = 040
0-0=0
 Development of II Simplex Table
For II Simplex Table outgoing variable (S3) replace by incoming variable (X2)
and the other basic variable and their position will be same.
Old entry of Key Raw
Transformation of Key Raw: New values of Key
=
Raw Key Number
Volume X1 X2 X3 S1 S2 S3
Key Raw
Col
Old Entry 8,000 7 10 8 0 0 1
Key Number 10 10 10 10 10 10 10
New Entry 800 7/10 1 4/5 0 0 1/10
Transformation of Non-Key Raw
Corresponding Entry of Key Raw x
Corresponding entry of Key Column
New values of Old
= -
Non-Key Raw Entry Key Number

DSM_Linear_Programming 41
Transformation of Non-Key Raw S1:

Vol Col X1 X2 X3 S1 S2 S3
Old Entry 10,000 10 12 7 1 0 0
Corresponding
8,000 7 10 8 0 0 1
Entry of Key
Corresponding
entry of Key 12 12 12 12 12 12 12
Col.
Key Number 10 10 10 10 10 10 10

New Entry
400 8/5 0 -13/5 1 0 -6/5

Volume Column = 10,000 - [(8,000 x 12) / 10] = 400

X1 Column = 10 - [(7 x 12) / 10] = 8/5
X2 Column = 12 - [(10 x 12) / 10] = 0
X3 Column = 7 - [(8 x 12) / 10] = -13/5
S1 Column = 1 - [(0 x 12) / 10] = 1
S2 Column = 0 - [(0 x 12) / 10] = 0
S3 Column = 0 - [(1 x 12) / 10] = -6/5
DSM_Linear_Programming 42
Transformation of Non-Key Raw S2:

Vol Col X1 X2 X3 S1 S2 S3
Old Entry 1,000 1 1 1 0 1 0
Corresponding
8,000 7 10 8 0 0 1
Entry of Key
Corresponding
entry of Key 1 1 1 1 1 1 1
Col.
Key Number 10 10 10 10 10 10 10

New Entry
200 3/10 0 1/5 0 1 -1/10

Volume Column = 1,000 - [(8,000 x 1) / 10] = 200

X1 Column = 1 - [(7 x 1) / 10] = 3/10
X2 Column = 1 - [(10 x 1) / 10] = 0
X3 Column = 1 - [(8 x 1) / 10] = 1/5
S1 Column = 0 - [(0 x 1) / 10] = 0
S2 Column = 1 - [(0 x 1) / 10] = 1
S3 Column = 0 - [(1 x 1) / 10] = -1/10
DSM_Linear_Programming 43
 Key Column Key Raw Simplex Table II
30 40 20 0 0 0
(Cj) Quantity Real Variable Column
Var. Ratio
Col. / Value
Col.
Column
X1 X2 X3 S1 S2 S3
Key Number

0 S1 400 8/5 0 -13/5 1 0 -6/5 400/(8/5)=250

0 S2 200 3/10 0 1/5 0 1 -1/10 200/(3/10)=667

40 X2 800 7/10 1 4/5 0 0 1/10 800/(7/10)=1143

28 40 32 0 0 4

Zj 2 0 -12 0 0 -4

ZNet Evaliation Raw (Cj - Zinj )Cj column x

j Raw Value = [∑(Entry Net Evaluation Raw = Cj - Zj
Corresponding entries in Variable column)]
X1 Col. = 30 - 28 = 2
X1 Column (0 x 8/5) + (0 x 3/5) + (40 x 7/10) = 28
X2 Col. = 40 - 40 = 0
X2 Column (0 x 0) + (0 x 0) + (40 x 1) = 40
X3 Col. = 20 - 32 = -12
X3 Column (0 x -13/5) + (0 x 1/5) + (40 x 4/5) = 32
S1 Column (0 x 1) + (0 x 0) + (40 x 0) = 0 S1 Col. = 0 - 0 = 0
S2 Column (0 x 0) + (0 x 1) + (40 x 0) = 0 S2 Col. = 0 - 0 = 0
S3 Column (0 x -6/5) + (0 x -1/10) + (40 x 1/10) = 4 S3 Col. = 0 - 4 = -4
DSM_Linear_Programming 44
 Development of III Simplex Table
For III Simplex Table outgoing variable (S1) replace by incoming variable
(X1) and the other basic variable and their position will be same.
Old entry of Key Raw
Transformation of Key Raw: New values of Key
=
Raw Key Number
Volume X1 X2 X3 S1 S2 S3
Key Raw
Col
Old Entry 400 8/5 0 -13/5 1 0 -6/5
Key Number 8/5 8/5 8/5 8/5 8/5 8/5 8/5
400/(8/5) (-13/5)/ 1/ (8/5) (-6/5)/
New Entry 1 0 0
=250 (8/5)= -13/8 =5/8 (8/5)= -3/4

Transformation of Non-Key Raw

Corresponding Entry of Key Raw x
Corresponding entry of Key Column
New values of Old
= -
Non-Key Raw Entry Key Number
DSM_Linear_Programming 45
Transformation of Non-Key Raw S2:

Vol Col X1 X2 X3 S1 S2 S3
Old Entry 200 3/10 0 1/5 0 1 -1/10
Corresponding
Entry of Key Raw
400 8/5 0 -13/5 1 0 -6/5
Corresponding
entry of Key Col.
3/10 3/10 3/10 3/10 3/10 3/10 3/10
Key Number 8/5 8/5 8/5 8/5 8/5 8/5 8/5

New Entry 125 0 0 11/16 -3/16 1 1/8

Calculation of New Entry:

Volume Column = 200 - [(400 x 3/10) / (8/5)] = 125
X1 Column = 3/10 - [(8/5 x 3/10) / (8/5)] = 0
X2 Column = 0 - [(0 x 3/10) / (8/5)] = 0
X3 Column = 1/5 - [(-13/5 x 3/10) / (8/5)] = 11/16
S1 Column = 0 - [(1 x 3/10) / (8/5)] = -3/16
S2 Column = 1 - [(0 x 3/10) / (8/5)] = 1
S3 Column = -1/10 - [(-6/5 x 3/10) / (8/5)] = 1/8
DSM_Linear_Programming 46
Transformation of Non-Key Raw X2:

Vol Col X1 X2 X3 S1 S2 S3
Old Entry 800 7/10 1 4/5 0 0 1/10
Corresponding
Entry of Key Raw
400 8/5 0 -13/5 1 0 -6/5
Corresponding
entry of Key Col.
7/10 7/10 7/10 7/10 7/10 7/10 7/10
Key Number 8/5 8/5 8/5 8/5 8/5 8/5 8/5

New Entry 625 0 1 31/16 -7/16 0 5/8

Calculation of New Entry:

Volume Column = 800 - [(400 x 7/10) / (8/5)] = 625
X1 Column = 7/10 - [(8/5 x 7/10) / (8/5)] = 0
X2 Column = 1 - [(0 x 7/10) / (8/5)] = 1
X3 Column = 4/5 - [(-13/5 x 7/10) / (8/5)] = 31/16
S1 Column = 0 - [(1 x 7/10) / (8/5)] = -7/16
S2 Column = 0 - [(0 x 3/10) / (8/5)] = 0
S3 Column = 1/10 - [(-6/5 x 7/10) / (8/5)] = 5/8
DSM_Linear_Programming 47
 Simplex Table III
30 40 20 0 0 0
(Cj) Quantity Real Variable Column
Var. Ratio
Col. / Value
Col.
Column
X1 X2 X3 S1 S2 S3
30 X1 250 1 0 -13/8 5/8 0 -3/4
0 S2 124 0 0 11/16 -3/16 1 1/8

40 X2 625 0 1 31/16 -7/16 0 5/8

30 40 115/4 5/4 0 5/2

Zj 0 0 -35/4 -5/4 0 -5/2

ZNet Evaliation Raw (Cj - Zinj )Cj column x Net Evaluation Raw = Cj - Zj
j Raw Value = [∑(Entry
Corresponding entries in Variable column)] 30 - 3 = 0
X1 Column (30 x 1) + (0 x 0) + (40 x 0) = 30 40 - 40 = 0
X2 Column (30 x 0) + (0 x 0) + (40 x 1) = 40 20 – 115/4 = -35/4
X3 Column (30 x -13/8) + (0 x 11/16) + (40 x 31/16) = 115/4 0 – 5/4 = – 5/4
S1 Column (30 x 5/8) + (0 x -3/16) + (40 x -7/16) = 5/4 0-0=0
S2 Column (30 x 0) + (0 x 1) + (40 x 0) = 0 0 – 5/2 = – 5/2
S3 Column (30 x –3/4) + (0 x 1/8) + (40 x 5/8) = 5/2
DSM_Linear_Programming 48
 III Simplex Table
30 40 20 0 0 0
(Cj) Quantity Real Variable Column
Var. Ratio
Col. / Value
Col.
Column
X1 X2 X3 S1 S2 S3

30 X1 250 1 0 -13/8 5/8 0 -3/4

0 S2 124 0 0 11/16 -3/16 1 1/8
40 X2 625 0 1 31/16 -7/16 0 5/8
Zj 30 40 115/4 5/4 0 5/2
Since all theRaw
Net Evaliation entries
(Cj - Zof
j ) row0 in table
0 III are negative
-35/4 -5/4 0 or Zero,
-5/2 this shows
that current programme cannot be improved further and hence it is the
optimal programme.
The optimal programme is:
X1 = 250 X2= 625 and X3= 0
Maximize Profit: Z = 30 (250) + 40 (625) + 20 (0) = Rs. 32,500

DSM_Linear_Programming 49
 Simplex Method
(When all or some of constrains are in the form of
grater than or equal)
1. Big M Method 2. Two Phase Method

1. Big M Method
Value (Cost or Profit) to be
Slack, Surplus and added in objective equation
Sign Artificial Variable use
Maximum Minimi.
in constrain equation
Problem Problem
≥ -S +A +0S -MA +0S +MA
= +A -MA +MA

DSM_Linear_Programming 50
Problem:
Vadilal Company is experimenting with three types of ice-creams P, Q and R in
which three kinds of dry fruits viz, A, B and C will be used. Taking various
factors into consideration it has been decided to use at most 600 kg. of A, at least
480 kg. of B and exactly 540 kg. of C.
Ice-cream P requires 3, 2 and 2 kg. of A, B and C respectively, ice-cream Q
requires 1, 4 and 3 kg. of A, B and C respectively and ice-cream R requires 6, 2
and 3 kg. of A, B and C respectively. The profits contributions from P, Q and R
respectively are Rs. 2, Rs. 3 and Rs. 4 per kg. Formulate the above as linear
programming problem and find the optimum product-mix.

Ice-cream
Dry-Fruits Uses (kg.)
P Q R
A 3 1 6 At most 600
B 2 4 2 At least 480
C 2 3 3 Exactly 540
Profit per unit 2 3 4

DSM_Linear_Programming 51
Formulation of Problem
Decision variable:
One unit of Ice-Cream P = X1; One unit of Ice-Cream Q = X2 and
One unit of Ice-Cream R = X3
Objective Equation:
Maximize Z = 2X1 + 3X2 + 4X3
Constrain Equation
For dry-fruits A 3X1 + X2 + 6X3 ≤ 600 …. (i)
For dry-fruits B 2X1 + 4X2 + 2X3 ≥ 480 …. (ii)
For dry-fruits C 2X1 + 3X2 + 3X3 = 540 …. (iii)
Where X1, X2, X3 ≥ 0
Changing unequal equations into equal equation
Maximize Z = 2X1 + 3X2 + 4X3 + 0S1+ 0S2 – MA1 – MA2
Subject to 3X1 + X2 + 6X3 + S1 = 600
2X1 + 4X2 + 2X3 - S2 + A1 = 480
2X1 + 3X2 + 3X3 + A2 = 540
Where X1, X2, X3, S1, S2, A1, A2 ≥ 0
DSM_Linear_Programming 52
 Key Column Designing the initial Simplex Table I Key Raw
Z = 2X1 + 3X2 + 4X3 + 0S1+ 0S2 – MA1 – MA 2X1 + 4X2 + 2X3 - S2 + A1 = 480
3X1 + X2 + 6X3 + S1 = 600 2X1 + 3X2 + 3X3 + A2 = 540
2 3 4 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1A
Col. X1 X2 X3 S1 S2
2
0 S1 600 3 1 6 1 0 0 0 600 / 1 = 600

-M A1 480 2 4 2 0 -1 1 0 480 / 4 = 120

Key Number

-M A2 540 2 3 3 0 0 0 1 540 / 3 = 180

-4M -7M -5M 0 M -M -M
Zj
2+4M 3+7M 4+5M 0 -M 0 0
Net Evaliation Raw
Zj Raw Value
(Cj - Zj ) Net Evaluation Raw = Cj - Zj Ratio:
X1 Col. = (0 x 3) + (-M x 2) + (-M x 2) = -4M Value of Quantity or Value
X1 Col. = 2 – (-4M) = 2 + 4M
X2 Col. = (0 x 1) + (-M x 4) + (-M x 3) = -7M column / Corresponding value
X2 Col. = 3 – (-7M) = 3 + 7M
X3 Col. = (0 x 6) + (-M x 2) + (-M x 3) = -5M of Key column
X3 Col. = 4 – (-5M) = 4 + 5M
S1 Col. = (0 x 1) + (-M x 0) + (-M x 0) = 0
S1 Col. = 0 – 0 = 0
S2 Col. = (0 x 0) + (-M x -1) + (-M x 0) = M
S2 Col. = 0 –M = -M
A1 Col. = (0 x 0) + (-M x 1) + (-M x 0) = -M
A1 Col. = -M – (-M) = 0
A2 Col. = (0 x 0) + (-M x 0) + (-M x 1) = -M
A2 Col. = -M DSM_Linear_Programming
– (-M) = 0 53
 Development of II Simplex Table
For II Simplex Table outgoing variable (A1) replace by incoming variable
(X2) and the other basic variable and their position will be same.
Transformation of Key Raw (A1): New Transformation of Non-Key Raw (S1):
Value = Old Entry/Key Number New Value = Old Entry – [(Corresponding
Qt. / Vol. Col. = 480 / 4 = 120 Entry of Key Raw x Corresponding entry of
X1 Col. = 2 / 4 = 1/2 Key Column) / Key Number]
X2 Col. = 4 / 4 = 1 Qt./Vol. Col. = 600 – (480 x 1 / 4) =480
X3 Col. = 2 / 4 = 1/2 X1 Col. = 3 – (2 x 1 / 4) = 5/2
S1 Col. = 0 / 4 = 0 X2 Col. = 1 – (4 x 1 / 4) = 0
S2 Col. = -1 / 4 = -1/4 X3 Col. = 6 – (2 x 1 / 4) = 11/2
A2 Col. = 0 / 4 = 0 S1 Col. = 1 – (0 x 1 / 4) = 1
S2 Col. = 0 – (-1 x 1 / 4) = 1/4
A2 Col. = 0 – (0 x 1 / 4) = 0
Transformation of Non-Key Raw (A2): Qt./Vol. Col. = 580 – (480 x 3 / 4) =180
New Value = Old Entry – X1 Col. = 2 – (2 x 3 / 4) = 1/2
[(Corresponding Entry of Key Raw x X2 Col. = 3 – (4 x 3 / 4) = 0
Corresponding entry of Key Column) / X3 Col. = 3 – (2 x 3 / 4) = 3/2
Key Number]
S1 Col. = 0 – (0 x 3 / 4) = 0
S2 Col. = 0 – (-1 x 3 / 4) = 3/4
DSM_Linear_Programming 54
A2 Col. = 1 – (0 x 3 / 4) = 1
 Key Column Designing the initial Simplex Table II Key Raw
2 3 4 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1A
Col. X1 X2 X3 S1 S2
2
480/(11/2)=
0 S1 480 5/2 0 11/2 1 1/4 -- 0
960/11
Key Number

3 X2 120 1/2 1 1/2 0 -1/4 -- 0 120/(1/2)=240

-M A2 180 1/2 0 3/2 0 ¾ -- 1 180/(3/2)=120

(3– (3- (-3 -
3 3M) /2 0 3M) /4 -- -M
M) /2
Zj (1+M) / (5+3M) 3+3M/
0 0 -- 0
2 /2 4
Net Evaliation Raw
Zj Raw Value Net Evaluation Raw = Cj - Zj
(Cj - Zj )
X1 Col. = (0 x 5/2) + (3 x 1/2) + (-M x 1/2) = (3–M)/2 X1 Col. = 2 – [(3–M)/2] = (1 + M)/2
X2 Col. = (0 x 0) + (3 x 1) + (-M x 0) = 3 X2 Col. = 3 – 3 = 0
X3 Col. = (0 x 11/2) + (3 x 1/2) + (-M x 3/2) = (3-3M) /2 X3 Col. = 4 – [(3-3M) /2] = (5+3M)/2
S1 Col. = (0 x 1) + (3 x 0) + (-M x 0) = 0 S1 Col. = 0 – 0 = 0
S2 Col. = (0 x 1/4) + (3 x -1/4) + (-M x 3/4) = (-3-3M)/4 S2 Col. = 0 –[(-3-3M)/4] = 3+3M/4
A2 Col. = (0 x 0) + (3 x 0) + (-M x 1) = -M A2 Col. = -M – (-M) = 0
Ratio = Value of Quantity or Value column / Corresponding value of Key column
DSM_Linear_Programming 55
 Development of III Simplex Table
For III Simplex Table outgoing variable (S1) replace by incoming variable
(X3) and the other basic variable and their position will be same.
Transformation of Key Raw (S1): New Transformation of Non-Key Raw (X2): New
Value = Old Entry/Key Number Value = Old Entry – [(Corresponding Entry of
Qt. / Vol. Col. = 480 / (11/2) = 960/11
Key Raw x Corresponding entry of Key
X1 Col. = (5/2) / (11/2) = 5/11 Column) / Key Number]
X2 Col. = 0 / (11/2) = 0 Qt./Vol. Col. = 120 – [(480 x
X3 Col. = (11/2) / (11/2) = 1 1/2)/(11/2) ]=840/11
S1 Col. = 1 / (11/2) = 2/11 X1 Col. = 1/2 – [(5/2 x 1/2) / (11/2)] = 3/11
S2 Col. = (1/4) / (11/2) = 1/22 X2 Col. = 1 – [(0 x ½) / (11/2)] = 1
A2 Col. = 0 / (11/2) = 0 X3 Col. = 1/2 – [(11/2 x ½) / 11/2)] = 0
S1 Col. = 0 – [(1 x ½) / (11/2)] = -1/11
S2 Col. = -1/4 – [(1/4 x ½) / (11/2)] = -3/11
A2 Col.Col.
Transformation of Non-Key Raw (A2): Qt./Vol. = 0 –= [(0
180x–½) / (11/2)]
[(480 x 3/2)= /0(11/2)] =540/11
New Value = Old Entry – X1 Col. = 1/2 – [(5/2 x 3/2) / (11/2)] = -2/11
[(Corresponding Entry of Key Raw x X2 Col. = 0 – [(0 x 3/2) / (11/2)] = 0
Corresponding entry of Key Column) / X Col. = 3/2 – [(11/2 x 3/2) / (11/2)] = 0
3
Key Number] S1 Col. = 0 – [(1 x 3/2) / (11/2)] = -3/11
S2 Col. = 3/4 – [(1/4 x 3/2) / (11/2)] = 15/22
A2 Col. = 1DSM_Linear_Programming
– [(0 x 3/2) / (11/2)] = 1 56
 Key Column Designing the initial Simplex Table III Key Raw
2 3 4 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1
Col. X1 X2 X3 S1 S2
A2
(960/11)/
4 X1 960/11 5/11 0 1 2/11 1/22 -- 0
(1/22)=1920
Key Number

X2 (840/11)/(-3/11)
3 840/11 3/11 1 0 -1/11 -3/11 -- 0
=-280

A2 (540/11)/(15/22)
-M 540/11 -2/11 0 0 -3/11 15/22 -- 1
72
=72
(29+2M (-14 -
3 4 5+3M/11 15M) /22 -- -M
)/11
Zj (14+15M)/
(-7- 0 0 (-5-3M)/11 (14+15M) -- 0
2M) / 11 22
/22
Net Evaliation Raw
(Cj - Zj )
Zj Raw Value Net Evaluation Raw = Cj - Zj
X1 Col. = (4 x 5/11) + (3 x 3/11) + (-M x -2/11) = (29+2M)/11 X1 Col. = 2 – [(29–2M)/11] = (-7+2M)/11
X2 Col. = (4 x 0) + (3 x 1) + (-M x 0) = 3 X2 Col. = 3 – 3 = 0
X3 Col. = (4 x 1) + (3 x 0) + (-M x 0) = 4 X3 Col. = 4 – 4 = 0
S1 Col. = (4 x 2/11) + (3 x -1/11) + (-M x -3/11) = 5+3M/11 S1 Col. = 0 – 5+3M/11 = (-5-3M)/11
S2 Col. = (4 x 1/22) + (3 x -3/11) + (-M x 15/22) = (-14-15M)/22 S2 Col. = 0 –(-14-15M)/22 = (14+15M)/22
A2 Col. = (4 x 0) + (3 x 0) + (-M x 1) = -M A2 Col. = -M – (-M) = 0value of Key column57
DSM_Linear_Programming
Ratio = Value of Quantity or Value column / Corresponding
 Development of IV Simplex Table
For III Simplex Table outgoing variable (A2) replace by incoming variable
(S2) and the other basic variable and their position will be same.
Transformation of Key Raw (A2): New Transformation of Non-Key Raw (X3): New Value = Old
Value = Old Entry/Key Number Entry – [(Corresponding Entry of Key Raw x Corresponding
Qt. / Vol. Col. = (540/11)/(15/22) = 72 entry of Key Column) / Key Number]
X1 Col. = (-2/11) /(15/22) = -4/15 Qt./Vol. Col. = 960/11–[(540/11 x 1/22)/(15/22) ]= 84
X2 Col. = 0 /(15/22) = 0 X1 Col. = 5/11 – [(-2/11 x 1/22)/(15/22) ] = 7/15
X3 Col. = 0 /(15/22) = 0 X2 Col. = 0 – [(0 x 1/22)/(15/22) ] = 0
S1 Col. = (-3/11) / (15/22) = -2/5 X3 Col. = 1 – [(0 x 1/22)/(15/22) ] = 1
S2 Col. = (15/22) /(15/22) = 1 S1 Col. = 2/11 – [(-3/11 x 1/22)/(15/22) ] = 1/5
S2 Col. = 1/22 – [(15/22 x 1/22)/(15/22) ] = 0
Transformation of Non-Key Raw (X2): Qt./Vol. Col. = 840/11 – [(540/11 x 1/22)/(15/22) ] = 96
New Value = Old Entry – [(Corresponding X1 Col. = 3/11– [(-2/11 x-3/11)/(15/22) ] = 1/5
Entry of Key Raw x Corresponding entry of X2 Col. = 1 – [(0 x-3/11)/(15/22) ] = 1
Key Column) / Key Number] X3 Col. = 0 – [(0 x -3/11)/(15/22) ] = 0
S1 Col. = -1/11 – [(-3/11 x -3/11/(15/22) ] = -1/5
S2 Col. = -3/11 – [(15/22 x-3/11)/(15/22) ] = 0

DSM_Linear_Programming 58
 Designing the initial Simplex Table IV
2 3 4 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1
Col. X1 X2 X3 S1 S2
A2
4 X1 84 7/15 0 1 1/5 0 -- --

3 X2 96 1/5 1 0 -1/5 0 -- --

0 S2 72 -4/5 0 0 -2/5 1 -- --

37/15 3 4 1/5 0 -- --
Zj
-7/15 0 0 -1/5 0 -- --
Net Evaliation Raw
(Cj - Zj )
Zj Raw Value Net Evaluation Raw = Cj - Zj
X1 Col. = (4 x 7/15) + (3 x 1/5) + (0 x -4/5) = 37/15 X1 Col. = 2 – 37/15 = -7/15
X2 Col. = (4 x 0) + (3 x 1) + (0 x 0) = 3 X2 Col. = 3 – 3 = 0
X3 Col. = (4 x 1) + (3 x 0) + (0 x 0) = 4 X3 Col. = 4 – 4 = 0
S1 Col. = (4 x 1/5) + (3 x -1/5) + (0 x -2/5) = 1/5 S1 Col. = 0 –1/5 = -1/5
S2 Col. = (4 x 0) + (3 x 0) + (0 x 1) = 0 S2 Col. = 0 –0 = 0
DSM_Linear_Programming 59
 IV Simplex Table
2 3 4 0 0 -M -M
(Cj) Quantit Real Variable Column
Var. Ratio
Col. y / Value
Col.
Column
X1 X2 X3 S1 S2 A1 A2

4 X3 84 7/15 0 1 1/5 0 -- --
3 X2 96 1/5 1 0 1/5 0 -- --
0 S2 72 -4/15 0 0 -2/5 1 -- --
Zj 37/15 3 4 7/5 0 -- --

Since all the

Net Evaliation entries
Raw row in 0table 0III are
(Cj - Zj )of -7/15 -7/5negative
0 or Zero,
-- -- this shows
that current programme cannot be improved further and hence it is the
optimal programme.
The optimal programme is:
X1 = 0, X2= 96 and X3= 84
Maximize Profit: Z = 2 (0) + 3 (96) + 4 (84) = Rs. 624

DSM_Linear_Programming 60
2. Two Phase Method
In this method solution is obtained in two phases as follows:
A. Phase I:
Step 1: Constructed auxiliary linear programming problem
Change all the requirement constraints to equalities by using slack or
surplus variables and add artificial variable wherever necessary. Assign
(-1) coefficient to each of the artificial variable (A) and (0) coefficient to
all other variables (in place of original coefficients) in the objective
function (after changing it to maximization, if it is not already).
Step 2. Solve the auxiliary problems by using simplex method until either of
the three possibilities do arise.
a) Max Z = 0 and at least one artificial variable appears in the basic
variable column at positive level. In this case the given problem has no
feasible solution.
b) Max Z =0 and at least one artificial variable appear in basic variable
column at zero level. In this case processed to phase II.
c) Max Z = 0 and no artificial variable appear in basis. In this case proceed
to phase II.
DSM_Linear_Programming 61
 B. Phase II:
Assign actual coefficients to the variable in the objective equation apply
simplex method to the modified simplex table obtained in phase I, until
an optimal basic feasible solution (if exists) has been obtained.
Problem:
Solve the following LPP:
Objective Equation:
Maximize Z = 2X1 + 3X2 + 4X3
Constrain Equation
For dry-fruits A 3X1 + X2 + 6X3 ≤ 600
For dry-fruits B 2X1 + 4X2 + 2X3 ≥ 480
For dry-fruits C 2X1 + 3X2 + 3X3 = 540
Where X1, X2, X3 ≥ 0

Solution:
Phase I:
Step 1: Constructed auxiliary linear programming problem
Maximize Z = 0X1 + 0X2 + 0X3 + 0S1+ 0S2 – 1A1 – 1A2
Subject to 3X1 + X2 + 6X3 + S1 = 600
2X1 + 4X2 + 2X3 - S2 + A1 = 480
2X1 + 3X2 + 3X3 + A2 = 540 DSM_Linear_Programming 62
 Key Column Designing the initial Simplex Table I Key Raw
Z = 0X1 + 0X2 + 0X3 + 0S1+ 0S2 – 1A1 – 1A2 2X1 + 4X2 + 2X3 - S2 + A1 = 480
3X1 + X2 + 6X3 + S1 = 600 2X1 + 3X2 + 3X3 + A2 = 540
0 0 0 0 0 -1 -1
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1A
Col. X1 X2 X3 S1 S2
2
0 S1 600 3 1 6 1 0 0 0 600 / 1 = 600

-1 A1 480 2 4 2 0 -1 1 0 480 / 4 = 120

Key Number

-1 A2 540 2 3 3 0 0 0 1 540 / 3 = 180

-4 -7 -5 0 1 -1 -1
Zj
4 7 5 0 -1 0 0
Net Evaliation Raw
Zj Raw Value
(Cj - Zj ) Net Evaluation Raw = Cj - Zj Ratio:
X1 Col. = (0 x 3) + (-1 x 2) + (-1 x 2) = -4 Value of Quantity or Value
X1 Col. = 0 – (-4) = 4
X2 Col. = (0 x 1) + (-1 x 4) + (-1 x 3) = -7 column / Corresponding value
X2 Col. = 0 – (-7) = 7
X3 Col. = (0 x 6) + (-1 x 2) + (-1 x 3) = -5 of Key column
X3 Col. = 0 – (-5) = 5
S1 Col. = (0 x 1) + (-1 x 0) + (-1 x 0) = 0
S1 Col. = 0 – 0 = 0
S2 Col. = (0 x 0) + (-1 x -1) + (-1 x 0) = 1
S2 Col. = 0 –1 = -1
A1 Col. = (0 x 0) + (-1 x 1) + (-1 x 0) = -1
A1 Col. = -1 – (-1) = 0
A2 Col. = (0 x 0) + (-1 x 0) + (-1 x 1) = -1
A2 Col. = -1 –DSM_Linear_Programming
(-1) = 0 63
 Development of II Simplex Table
For II Simplex Table outgoing variable (A1) replace by incoming variable
(X2) and the other basic variable and their position will be same.
Transformation of Key Raw (A1): New Transformation of Non-Key Raw (S1):
Value = Old Entry/Key Number New Value = Old Entry – [(Corresponding
Qt. / Vol. Col. = 480 / 4 = 120 Entry of Key Raw x Corresponding entry of
X1 Col. = 2 / 4 = 1/2 Key Column) / Key Number]
X2 Col. = 4 / 4 = 1 Qt./Vol. Col. = 600 – (480 x 1 / 4) =480
X3 Col. = 2 / 4 = 1/2 X1 Col. = 3 – (2 x 1 / 4) = 5/2
S1 Col. = 0 / 4 = 0 X2 Col. = 1 – (4 x 1 / 4) = 0
S2 Col. = -1 / 4 = -1/4 X3 Col. = 6 – (2 x 1 / 4) = 11/2
A2 Col. = 0 / 4 = 0 S1 Col. = 1 – (0 x 1 / 4) = 1
S2 Col. = 0 – (-1 x 1 / 4) = 1/4
A2 Col. = 0 – (0 x 1 / 4) = 0
Transformation of Non-Key Raw (A2): Qt./Vol. Col. = 580 – (480 x 3 / 4) =180
New Value = Old Entry – X1 Col. = 2 – (2 x 3 / 4) = 1/2
[(Corresponding Entry of Key Raw x X2 Col. = 3 – (4 x 3 / 4) = 0
Corresponding entry of Key Column) / X3 Col. = 3 – (2 x 3 / 4) = 3/2
Key Number]
S1 Col. = 0 – (0 x 3 / 4) = 0
S2 Col. = 0 – (-1 x 3 / 4) = 3/4
DSM_Linear_Programming 64
A2 Col. = 1 – (0 x 3 / 4) = 1
 Key Column Designing the initial Simplex Table II Key Raw
0 0 0 0 0 -1 -1
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1A
Col. X1 X2 X3 S1 S2
2
480/(11/2)=
0 S1 480 5/2 0 11/2 1 1/4 -- 0
960/11
Key Number

0 X2 120 1/2 1 1/2 0 -1/4 -- 0 120/(1/2)=240

-1 A2 180 1/2 0 3/2 0 3/4 -- 1 180/(3/2)=120

-1 /2 0 -3 /2 0 -3 /4 -- -1

Zj
1/2 0 3/2 0 3/4 -- 0
Net Evaliation Raw
Zj Raw Value Net Evaluation Raw = Cj - Zj
(Cj - Zj )
X1 Col. = (0 x 5/2) + (0 x 1/2) + (-1 x 1/2) = –1/2 X1 Col. = 0 – (-1/2) = 1 / 2
X2 Col. = (0 x 0) + (0 x 1) + (-1 x 0) = 0 X2 Col. = 0 – 0 = 0
X3 Col. = (0 x 11/2) + (0 x 1/2) + (-1 x 3/2) = -3 /2 X3 Col. = 0 – (-3 /2) = 3 / 2
S1 Col. = (0 x 1) + (0 x 0) + (-1 x 0) = 0 S1 Col. = 0 – 0 = 0
S2 Col. = (0 x 1/4) + (0 x -1/4) + (-1 x 3/4) = -3 /4 S2 Col. = 0 –(-3/4) = 3/4
A2 Col. = (0 x 0) + (0 x 0) + (-1 x 1) = -1 A2 Col. = -1 – (-1) = 0
Ratio = Value of Quantity or Value column / Corresponding value of Key column
DSM_Linear_Programming 65
 Development of III Simplex Table
For III Simplex Table outgoing variable (S1) replace by incoming variable
(X3) and the other basic variable and their position will be same.
Transformation of Key Raw (S1): New Transformation of Non-Key Raw (X2): New
Value = Old Entry/Key Number Value = Old Entry – [(Corresponding Entry of
Qt. / Vol. Col. = 480 / (11/2) = 960/11
Key Raw x Corresponding entry of Key
X1 Col. = (5/2) / (11/2) = 5/11 Column) / Key Number]
X2 Col. = 0 / (11/2) = 0 Qt./Vol. Col. = 120 – [(480 x
X3 Col. = (11/2) / (11/2) = 1 1/2)/(11/2) ]=840/11
S1 Col. = 1 / (11/2) = 2/11 X1 Col. = 1/2 – [(5/2 x 1/2) / (11/2)] = 3/11
S2 Col. = (1/4) / (11/2) = 1/22 X2 Col. = 1 – [(0 x ½) / (11/2)] = 1
A2 Col. = 0 / (11/2) = 0 X3 Col. = 1/2 – [(11/2 x ½) / 11/2)] = 0
S1 Col. = 0 – [(1 x ½) / (11/2)] = -1/11
S2 Col. = -1/4 – [(1/4 x ½) / (11/2)] = -3/11
A2 Col.Col.
Transformation of Non-Key Raw (A2): Qt./Vol. = 0 –= [(0
180x–½) / (11/2)]
[(480 x 3/2)= /0(11/2)] =540/11
New Value = Old Entry – X1 Col. = 1/2 – [(5/2 x 3/2) / (11/2)] = -2/11
[(Corresponding Entry of Key Raw x X2 Col. = 0 – [(0 x 3/2) / (11/2)] = 0
Corresponding entry of Key Column) / X Col. = 3/2 – [(11/2 x 3/2) / (11/2)] = 0
3
Key Number] S1 Col. = 0 – [(1 x 3/2) / (11/2)] = -3/11
S2 Col. = 3/4 – [(1/4 x 3/2) / (11/2)] = 15/22
A2 Col. = 1DSM_Linear_Programming
– [(0 x 3/2) / (11/2)] = 1 66
 Key Column Designing the initial Simplex Table III Key Raw
0 0 0 0 0 -1 -1
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1
Col. X1 X2 X3 S1 S2
A2
(960/11)/
0 X1 960/11 5/11 0 1 2/11 1/22 -- 0
(1/22)=1920
Key Number

X2 (840/11)/(-3/11)
0 840/11 3/11 1 0 -1/11 -3/11 -- 0
= -280

A2 (540/11)/(15/22)
-1 540/11 -2/11 0 0 -3/11 15/22 -- 1
72
=72
2 / 11 0 0 3 / 11 -15 /22 -- -1
Zj
-2 / 11 0 0 -3 / 11 15
15//22
22 -- 0
Net Evaliation Raw
(Cj - Zj )
Zj Raw Value Net Evaluation Raw = Cj - Zj
X1 Col. = (0 x 5/11) + (0 x 3/11) + (-1 x -2/11) = 2 / 11 X1 Col. = 0 – (2/11) = -2 / 11
X2 Col. = (0 x 0) + (0 x 1) + (-1 x 0) = 0 X2 Col. = 0 – 0 = 0
X3 Col. = (0 x 1) + (0 x 0) + (-1 x 0) = 0 X3 Col. = 0 – 0 = 0
S1 Col. = (0 x 2/11) + (0 x -1/11) + (-1 x -3/11) = 3 / 11 S1 Col. = 0 – (3 / 11) = -3 / 11
S2 Col. = (0 x 1/22) + (0 x -3/11) + (-1 x 15/22) = 15 / 22 S2 Col. = 0 –(-15 / 22) = 15 / 22
A2 Col. = (0 x 0) + (0 x 0) + (-1 x 1) = -1 A2 Col. = -1 – (-1) = 0 value of Key column67
DSM_Linear_Programming
Ratio = Value of Quantity or Value column / Corresponding
 Development of IV Simplex Table
For III Simplex Table outgoing variable (A2) replace by incoming variable
(S2) and the other basic variable and their position will be same.
Transformation of Key Raw (A2): New Transformation of Non-Key Raw (X3): New Value = Old
Value = Old Entry/Key Number Entry – [(Corresponding Entry of Key Raw x Corresponding
Qt. / Vol. Col. = (540/11)/(15/22) = 72 entry of Key Column) / Key Number]
X1 Col. = (-2/11) /(15/22) = -4/15 Qt./Vol. Col. = 960/11–[(540/11 x 1/22)/(15/22) ]= 84
X2 Col. = 0 /(15/22) = 0 X1 Col. = 5/11 – [(-2/11 x 1/22)/(15/22) ] = 7/15
X3 Col. = 0 /(15/22) = 0 X2 Col. = 0 – [(0 x 1/22)/(15/22) ] = 0
S1 Col. = (-3/11) / (15/22) = -2/5 X3 Col. = 1 – [(0 x 1/22)/(15/22) ] = 1
S2 Col. = (15/22) /(15/22) = 1 S1 Col. = 2/11 – [(-3/11 x 1/22)/(15/22) ] = 1/5
S2 Col. = 1/22 – [(15/22 x 1/22)/(15/22) ] = 0
Transformation of Non-Key Raw (X2): Qt./Vol. Col. = 840/11 – [(540/11 x 1/22)/(15/22) ] = 96
New Value = Old Entry – [(Corresponding X1 Col. = 3/11– [(-2/11 x-3/11)/(15/22) ] = 1/5
Entry of Key Raw x Corresponding entry of X2 Col. = 1 – [(0 x-3/11)/(15/22) ] = 1
Key Column) / Key Number] X3 Col. = 0 – [(0 x -3/11)/(15/22) ] = 0
S1 Col. = -1/11 – [(-3/11 x -3/11/(15/22) ] = 1/5
S2 Col. = -3/11 – [(15/22 x-3/11)/(15/22) ] = 0

DSM_Linear_Programming 68
 Designing the initial Simplex Table IV
0 0 0 0 0 -1 -1
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. A1
Col. X1 X2 X3 S1 S2
A2
0 X1 84 7/15 0 1 1/5 0 -- --

0 X2 96 1/5 1 0 -1/5 0 -- --

0 S2 72 -4/5 0 0 -2/5 1 -- --

0 0 0 0 0 -- --
Zj
0 0 0 0 0 -- --
Net Evaliation Raw
Phase(C
II:j - Zj )
Assign actual coefficients to the variable in the objective equation apply simplex
method to the modified simplex table obtained in phase I, until an optimal basic
feasible solution (if exists) has been obtained.

DSM_Linear_Programming 69
 Simplex Table IV
2 3 4 0 0 -M -M
(Cj) Quantit Real Variable Column
Var. Ratio
Col. y / Value
Col.
Column
X1 X2 X3 S1 S2 A1 A2

4 X1 84 7/15 0 1 1/5 0 -- --
3 X2 96 1/5 1 0 -1/5 0 -- --
0 S2 72 -4/5
37/15 30 40 -2/5
1/5 10 -- --
--
Zj -7/15 0 0 -1/5 0 -- --

ZNet Evaliation Raw (Cj - Zj )

j Raw Value Net Evaluation Raw = Cj - Zj
X1 Col. = (4 x 7/15) + (3 x 1/5) + (0 x -4/5) = 37/15
X1 Col. = 2 – 37/15 = -7/15
X2 Col. = (4 x 0) + (3 x 1) + (0 x 0) = 3
X2 Col. = 3 – 3 = 0
X3 Col. = (4 x 1) + (3 x 0) + (0 x 0) = 4
X3 Col. = 4 – 4 = 0
S1 Col. = (4 x 1/5) + (3 x -1/5) + (0 x -2/5) = 1/5
S1 Col. = 0 –1/5 = -1/5
S2 Col. = (4 x 0) + (3 x 0) + (0 x 1) = 0
S2 Col. = 0 –0 = 0

DSM_Linear_Programming 70
 IV Simplex Table
2 3 4 0 0 -M -M
(Cj) Quantit Real Variable Column
Var. Ratio
Col. y / Value
Col.
Column
X1 X2 X3 S1 S2 A1 A2

4 X3 84 7/15 0 1 1/5 0 -- --
3 X2 96 1/5 1 0 1/5 0 -- --
0 S2 72 -4/15 0 0 -2/5 1 -- --
Zj 37/15 3 4 7/5 0 -- --

Since all the

Net Evaluation Rawentries
j(C - Z ) of -7/15
j 0 0 Raw
Net Evaluation -7/5 in 0table--IV are
-- negative or
Zero, this shows that current programme cannot be improved further and
hence it is the optimal programme.
The optimal programme is:
X1 = 0, X2= 96 and X3= 84
Maximize Profit: Z = 2 (0) + 3 (96) + 4 (84) = Rs. 624

DSM_Linear_Programming 71
 Minimize Linear Programming Problem
Method I:
Minimization problems are solved same as maximization problem
and Slack Variable, Surplus Variable and Artificial Variable in
mixed type constrains are used as follows:
Slack, Surplus and
Value (Cost or Profit) to be
Sign Artificial Variable use
added in objective equation
in constrain equation
≤ +S +0S
≥ -S +A +0S +MA
= +A +MA
Method II:
Minimization problem convert into maximization problem by taking
the objective function Z* = - Z.
For example: Min. Z = X1 + X2
Convert into Maximization: Max Z* = -Z = - X1 - X2
DSM_Linear_Programming 72
Question 13:
Solve the following LPP:
Minimize Z = 5X1 + 2X2
Subject to: 3X1 + X2 = 4
2X1 + X2 ≥ 3
X1 + 2X2 ≤ 3
Where X1, X2 ≥ 0

Solution:
Minimize Z = 5X1 + 2X2 + 0S1+ 0S2 +MA1 +MA2
Subject to 3X1 + X2 + A1 = 4
2X1 + X2 – S1 + A2 = 3
X1 + 2X2 + S2 = 3
Where X1, X2, S1, S2, A1, A2 ≥ 0
DSM_Linear_Programming 73
 Key Column Simplex Table I Key Raw
Z = 5X1 + 2X2 + 0S1+ 0S2 +MA1 +MA2 2X1 + X2 – S1 + A2 = 3
3X1 + X2 + A1 = 4 X1 + 2X2 + S2 = 3

5 2 0 0 +M +M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. X1 X2 S1 S2 A1 A2
Col.
+M A1 4 3 1 0 0 1 0 4/3

+M A2 3 2 1 -1 0 0 1 3/2
Key Number

0 S2 3 1 2 0 1 0 0 3 / 1= 3
5M 2M -M 0 M M

Zj 5-5M
5-5M 2-2M M 0 0 0
Zj Net
Raw Evaliation
Value Raw
Net Evaluation Raw = Cj - Zj Ratio:
X1 Col. =(C j - Z
(+M x j3)) + (+M x 2) + (0 x 1) = 5M Value of Quantity or Value
X1 Col. = 5 – 5M = 5 – 5M
X2 Col. = (+M x 1) + (+M x 1) + (0 x 2) = 2M column / Corresponding value
X2 Col. = 2 – 2M = 2 – 2M
S1 Col. = (+M x 0) + (+M x -1) + (0 x 0) = -M of Key column
S1 Col. = 0 – (-M) = M
S2 Col. = (+M x 0) + (+M x 0) + (0 x 1) = 0
S2 Col. = 0 – 0 = 0
A1 Col. = (+M x 1) + (+M x 0) + (0 x 0) = M
A1 Col. = M – M = 0
A2 Col. = (+M x 0) + (+M x 1) + (0 x 0) = M
A2 Col. = M – M = 0
DSM_Linear_Programming 74
 5 2 0 0 +M +M
(Cj Real Variable Column
Qt./
)
Var. Valu
Col
Col. e
.
Col. X1 X2 S1 S2 A1 A2

+M A1 4 3 1 0 0 1 0

+M A2 3 2 1 -1 0 0 1

0 S2 3 Simplex
1 2 Table
0 1II 0 0

5 2 0 0 +M +M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col.
Col.
X1 X2 S1 S2 A1 A2
5 X1 4/3 1 1/3 0 0 -- 0 4

+M A2 1/3 0 1/3 -1 0 -- 1 1

0 S2 5/3 0 5/3 0 1 -- 0 1

5 5+M/3 -M 0 -- M
Zj
0 1-M/3 M 0 -- 0
Net Evaluation DSM_Linear_Programming 75
Raw (C - Z )
 5 2 0 0 +M +M
(Cj Real Variable Column
Qt./
)
Col
Var. Valu Since all the entries of NER in table
Col. e
.
Col. X1 X2 S1 S2 A1 A2
III are positive or Zero, this shows
that current programme cannot be
improved further and hence it is the
5 X1 4/3 1 1/3 0 0 -- 0
optimal programme.
+M A2 1/3 0 1/3 -1 0 -- 1 The optimal programme is:
0 S2 5/3 0 Simplex
5/3 0Table
1 III
-- 0 X1 = 1 and X2= 1

5 2 0 0 +M +M Minimize Z
(Cj) Qt./ Real Variable Column = 5X1 + 2X2
Var.
Col. Value Ratio = 5(1) + 2(1)
Col.
Col. =7
X1 X2 S1 S2 A1 A2
5 X11 1 1 0 1 0 -- --

2 X22 1 0 1 -3 0 -- --

0 S2 5/3 0 0 5 1 -- --

5 2 -1 0 -- --
Zj
0 0 1 0 -- --
Net Evaluation DSM_Linear_Programming 76
Raw (C - Z )
Question 21:
Solve the following LPP:
Minimize Z = 50X1 + 30X2
Subject to: X1 + X2 = 5
X1 + 2X2 ≤ 6
5X1 + 2X2 ≥ 10
Where X1, X2 ≥ 0
Solution:
Minimization problem convert into maximization problem by
taking the objective function Z* = - Z
Convert into Maximization: Max Z* = - Z = - 50X1 - 30X2
Maximize Z* = - 50X1 - 30X2 + 0S1+ 0S2 -MA1 -MA2
Subject to X1 + X2 + A1 = 5
X1 + 2X2 + S1 = 6
5X1 + 2X2 – S2 + A2 = 10
DSM_Linear_Programming 77
 Key Column Simplex Table I Key Raw
Z* = - 50X1 - 30X2 + 0S1+ 0S2 -MA1 -MA2 X1 + 2X2 + S1 = 6
X1 + X2 + A1 = 5 5X1 + 2X2 – S2 + A2 = 10

-50 -30 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col. X2 S2 A2
Col. X1 S1 A1
-M A1 5 1 1 0 0 1 0 5

0 S1 6 1 2 1 0 0 0 6
Key Number

-M A2 10 5 2 0 -1 0 1 2
-6M -3M 0 M -M -M

Zj -50+6M -30+3M 0 -M 0 0

Zj Net
Raw Evaliation
Value Raw
Net Evaluation Raw = Cj - Zj Ratio:
X1 Col. =(C j -x
(-M Z1)
j ) + (0 x 1) + (-M x 5) = -6M
Value of Quantity or
X1 Col. = -50 – (-6M) = -50 + 6M
X2 Col. = (-M x 1) + (0 x 2) + (-M x 2) = -3M Value column /
X2 Col. = -30 –(-3M) = -30 + 3M
S1 Col. = (-M x 0) + (0 x 1) + (0-Mx 0) = 0 Corresponding value of
S1 Col. = 0 – 0 = 0
S2 Col. = (-M x 0) + (0 x 0) + (-M x -1) = M Key column
S2 Col. = 0 – M = -M
A1 Col. = (-M x 1) + (0 x 0) + (-M x 0) = -M
A1 Col. = -M – (-M) = 0
A2 Col. = (-M x 0) + (0 x 0) + (-M x 1) = -M
A2 Col. = -M – (-M) = 0
DSM_Linear_Programming 78
 -50 -30 0 0 -M -M
(Cj Real Variable Column
Qt./
)
Var. Valu
Col
Col. e
.
Col. X1 X2 S1 S2 A1 A2

-M A1 5 1 1 0 0 1 0

0 S1 6 1 2 1 0 0 0

-M A2 10 Simplex
5 2 Table
0 II
-1 0 1

-50 -30 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col.
Col.
X1 X2 S1 S2 A1 A2
-M A1 3 0 3/5 0 1/5 1 -- 5

0 S1 4 0 8/5 1 1/5 0 -- 5/2

-50 X1 2 1 2/5 0 -1/5 0 -- 5

-50 (-3M- (-M+ -M
100)/5 0 50)5 --
Zj
0 (- 0 (M - 0 --
50+3M)
50+3M)/5
/ 50)5
Net Evaluation
5 DSM_Linear_Programming 79
Raw (C - Z )
 -50 -30 0 0 -M -M
(Cj Real Variable Column
Qt./
)
Var. Valu
Col
Col. e
.
Col. X1 X2 S1 S2 A1 A2

-M A1 3 0 3/5 0 1/5 1 --

0 S1 4 0 8/5 1 1/5 0 --

-50 X1 2 Simplex
1 2/5 Table
0 IiI
-1/5 0 --

-50 -30 0 0 -M -M
(Cj) Qt./ Real Variable Column
Var. Ratio
Col. Value
Col.
Col.
X1 X2 S1 S2 A1 A2
-M A1 3/2 0 0 -3/8 1/8 1 -- 12

-30 X2 5/2 0 1 5/8 1/8 0 -- 5/2

-50 X1 1 1 0 -1/4 -1/4 0 -- -4

-50 -30 (3M- (-M+ -M
50) /8 70)8 --
Zj (M
(M--
0 0 (-3M+ 0 --
50) /8 70)/8
Net Evaluation DSM_Linear_Programming 80
Raw (C - Z )
 -50 -30 0 0 -M -M
(Cj Real Variable Column
Qt./
)
Var. Valu Since all the entries of NER in table
Col
.
Col. e IV are negative or Zero, this shows
Col. X1 X2 S1 S2 A1 A2
that current programme cannot be
improved further and hence it is the
-M A1 3/2 0 0 -3/8 1/8 1 -- optimal programme.
-30 X2 5/2 0 1 5/8 1/8 0 -- The optimal programme is:
-50 X1 1 1 Simplex
0 -1/4 Table
-1/4 0IV --
X1 = 4 and X2= 1

-50 -30 0 0 -M -M Maximize Z*

(Cj) Qt./ Real Variable Column = -50X1 - 30X2
Var.
Col. Value Ratio = -50(4) - 30(1)
Col.
Col.
X1 X2 S1 S2 A1 A2 = -230
OR
0 S2 12 0 0 -3 1 -- --
Minimize Z
-30 X2 1 0 1 1 0 -- -- = 230

-50 X1 4 1 0 -1 0 -- --

-50 -30 20 0 -- --
Zj
0 0 -20 0 -- --
Net Evaluation
Raw (Cj - Zj ) DSM_Linear_Programming 81
Some Special Classes of Linear Programming
1. If Net Evaluation Raw (Cj - Zj) indicate that programme
cannot be improved any more (it means all NER values are
positive or zero in the case of minimize problem and negative
or zero in the case of maximize problem) but in the basic
variable column their exists some artificial variable (A) at
positive level, then the problem has no feasible solution.
2. While solving the problem by simplex method if NER indicate
that programme can be further, but all the entries of key
column are negative, i.e. we can not select outgoing variable,
then the given problem has unbounded solution.
3. If all the entries of NER are either 0 or Negative / Positive, but
entry corresponding to some non-basis (S) is Zero in NER
then the poblem has Multiple Optimal Solution.
In this case non basic variable (S) replaced by basic variable
and find out multiple optimal solution.

DSM_Linear_Programming 82
Cont…

4. Degeneracy: A condition sometimes encountered in a linear

programming problem where there is tie at the rime of selection of
Key row
a) If there is a tie between two or more rows, then select that row
as key row which is nearer to the bottom.
In case there exist an artificial variable in the basic
corresponding to some tied row, then that row should be given
preference, so that outgoing variable is the artificial variable.
b) If there is a tie between two or more columns:
i. If there is a tie between the columns corresponding to
decision varible selection can be made arbitrarily.
ii. If there is a tie between the columns corresponding to
decision variable and slack variable or surplus varible,
select the column corresponding to decision variable.
iii. If artificial variable exists in the basis then select that
column as key column such that outgoing variable is the
artificial variable.
DSM_Linear_Programming 83
 DUALITY
Association with every linear programming problem there exists
another linear programming problem known as Dual Problem of
the given problem (called primal), which can be derived from the
data of the given problem. If the given problem is a maximizing
problem its dual will be a minimizing problem and vice versa.

Formulation of Dual
Step I:
First put the primal problem in such a way so that all the required
constrain are in the of right directional in equalities i.e.
In ≤ type inequality if primal is a maximizing problem; and
In ≥ type inequality if primal is a minimizing problem

DSM_Linear_Programming 84
 Procedure for converting all constrains
into right direction
Type of Procedure (For maximizing Procedure (For minimizing
constraint Problem) Problem)
Multiply both side of the inequality
Less Than No Change by (-1) to change it into more than
(≤) & equal to type in equality.

Multiply both side of the inequality

More
by (-1) to change it into less than & No Change
Than (≥) equal to type in equality.

(a) Break the equality into two (a) Break the equality into two
inequality, one of less than type inequality, one of less than type
Equal To and other of more than type, and other of more than type,
(=) (b) Multiply both side of the more (b) Multiply both side of the less
than inequality by (-1) to than inequality by (-1) to
change it into less than & change it into more than &
equal to type in equality. equal to type in equality.

DSM_Linear_Programming 85
Example: Find the dual of the following LPP:
Max Z = 4X1 + 5X2
Subject to: 2X1 + X2 + 6X3 ≤ 6
X1 X3 ≥ 2
X1 - X2 + 3X3 = 4
Step I: In maximizing problem all the constrains are required in the less
than (≤) directional in equalities:
First constraint is in right direction:
2X1 + X2 + 6X3 ≤ 6
Second constraint is in wrong direction (≥ ), for changing in less than
direction multiply both side by (-1):
-X1 - X3 ≤ -2
Third constraint is in equality form, so it can be written in to inequalities
as follows:
X1 - X2 + 3X3 ≤ 4 ….(i)
X1 - X2 + 3X3 ≥ 4 ….(ii)
Equation (ii) is in wrong direction (≥ ), for changing in less than direction
multiply both side by (-1):
-X + X - 3X ≤ -4 DSM_Linear_Programming 86
Now: Max Z = 4X1 + 5X2
Subject to: 2X1 + X2 + 6X3 ≤ 6
-X1 - X3 ≤ -2
X1 - X2 + 3X3 ≤ 4
-X1 + X2 - 3X3 ≤ -4

Step II (of Formulation of Dual):

(a) Convert the given variables X1, X2, X3 … into Y1, Y2, Y3 …
respectively
(b) Convert the coefficients of the objective function into the
contents of the constraint function respectively. Similarly
convert the contents of constraints into the coefficient of the
objective function respectively.
(c) Convert the columns of coefficient of the constraints in the
primal problem into the rows of coefficient of the constraints
in the dual problem respectively
DSM_Linear_Programming 87
Max Z = 4X1 + 5X2 + 0X3
Subject to: 2X1 + X2 + 6X3 ≤ 6
-X1 - X3 ≤ -2
X1 - X2 + 3X3 ≤ 4 The coefficient matrix of
-X1 + X2 - 3X3 ≤ -4 dual problem will be:
The coefficient matrix of Y1 Y2 Y3 Y4
primal problem will be:
6 -2 4 -4
X1 X2 X3 Sol.

2 1 6 6 2 -1 1 -1
-1 0 -1 -2
1 -1 3 4 1 0 -1 1
-1 1 -3 -4
6 -1 3 -3

Taking one dual variable from each column (say Y1, Y2, Y3 & Y4)
Min Z = 6Y1 - 2Y2 + 4Y3 - 4Y4
Subject to: 2Y1 - Y2 + Y3 - 1Y4 ≥ 4
Y1 - Y3 + Y4 ≥ 5
DSM_Linear_Programming 88
 Dual Problem
Objective Equation
Min Z = 6Y1 - 2Y2 + 4Y3 - 4Y4
Constrain Equation
Subject to: 2Y1 - Y2 + Y3 - 1Y4 ≥ 4
Y1 - Y 3 + Y 4 ≥ 5
6Y1 - Y2 + 3Y3 - 3Y4 ≥ 0

DSM_Linear_Programming 89