LINEAR PROGRAMMING

INTRODUCTION:
Programming problems in general are governed
with the use or allocation of scarce resources
Labour,
Materials,
Machines and
Capital

in the “best” possible manner so that costs are

minimised or profits maximised. In using the term
“best” it is implied that some choice or a set of
alternative courses of actions is available for
making the decision. In general, the best decision
is found by solving a mathematical problem.
The term linear programming merely defines a particular class
of programming problems that meet the following
conditions:
1. The decision variables involved in the problem are non-
negative (i.e. positive or zero).
2. The criterion for selecting the “best” values of the
decision variables can be described by a linear function of
these variables, that is, a mathematical function involving
only the first powers of the variables with no cross
products. The criterion function is normally referred to as
the Objective function.
3. The operating rules governing the process (e.g., scarcity of
resources) can be expressed as a set of linear equations or
linear inequalities. This set is referred to as the Constraint
set.

The last two conditions are the reasons for the use of the
term linear programming.
FORMULATION OF LINEAR PROGRAMMING PROBLEMS
(L.P.P.)
The three basic steps to constructing a linear
programming model are as follows:

• STEP 1: Identify the unknown variables to be

determined (decision variables) and represent them
in terms of algebraic symbols.

• STEP 2: Identify all the restrictions or constraints in

the problem and express them as linear equations or
inequalities which are linear functions of the unknown
variables.

• STEP 3: Identify the objective or criterion and

represent it as a linear function of the decision
variables, which is to be maximised or minimised.
 IMPORTANT NOTE
MODEL BUILDING IS NOT A
SCIENCE BUT PRIMARILY AN
ART AND COMES MAINLY BY
PRACTICE.
HENCE, WORK OUT AS MANY
PROBLEMS ON FORMULATION
AS POSSIBLE.
Economic interpretation of Dual
Max=
Z 40 x1 + 35 x2 Profit
S .T .
2 x1 + 3 x2 ≤ 60 Raw materials constraint
4 x1 + 3 x2 ≤ 96 Labour hours constraint
x1 , x2 ≥ 0

Min=
Z 60 y1 + 96 y2
S .T .
DUAL 2 y1 + 4 y2 ≤ 40
3 y1 + 3 y2 ≤ 35
y1 , y2 ≥ 0
 Cj 40 35 0 0
P
R Basis Qty p/u X1 X2 s1 s2

I X2 8 35 0 1 2/3 -1/3

M X1 18 40 1 0 -1/2 1/2

A Zj 40 35 10/3 25/3
L Z=1000 Cj - Zj 0 0 -10/3 -25/3

Cj 60 96 0 0
D Basis Qty p/u y1 y2 s1 s2

U y2 25/3 96 0 1 -1/2 1/3

A y1 10/3 60 1 0 1/2 -2/3

Zj 60 96 -18 -8
L
Z=1000 Cj - Zj 0 0 18 8
P Basis Qty p/u X1 X2 s1 s2
R X2 8 35 0 1 2/3 -1/3
I
X1 18 40 1 0 -1/2 1/2
M
Cj - Zj 0 0 -10/3 -25/3
A
L
Z=1000

Basis Qty p/u y1 y2 s1 s2

D
U y2 25/3 96 0 1 -1/2 1/3
A y1 10/3 60 1 0 1/2 -2/3
L
Cj - Zj 0 0 18 8
Z=1000
Comparison of Primal and Dual

1. Z= 1000, same for both Primal and Dual

2. Numerical values of variables are same. i.e. x1=18, x2=8 is same as
s1=18, s2=8 in dual. (and vice-versa)
 1
Min Z= x1 + x2
2
S .T . 6 x1 + 2 x2 ≥ 24; 3 x1 + 2 x2 ≥ 18; x1 + 3 x2 ≥ 18; x1 , x2 ≥ 0
Cj 1 1/2 0 0 0 M M M

Basis Qty p/u X1 X2 s1 S2 S3 A1 A2 A3

x1 2 1 1 0 -1/3 1/3 0 1/3 -1/3 0

S3 8 0 0 0 7/6 -8/3 1 -7/6 8/3 -1

x2 6 1/2 0 1 1/2 -1 0 -1/2 1 0

Zj 1 1/2 -1/12 -1/6 0 1/12 1/6 0

Z=1000 Cj - Zj 0 0 1/12 1/6 0 M- M- M
1/12 1/6

Write the values of the dual variables and the slack

variables.
 Max =
Z 5 x1 + 10 x2 +8x3
S .T .
3x1 + 5 x2 +2x3 ≤ 60 Fabrication hours
4 x1 + 4 x2 +4x3 ≤ 72 Finishing hours
2 x1 + 4 x2 +5x3 ≤ 100 Packaging hours
x1 , x2 , x3 ≥ 0
Cj 5 10 8 0 0 0

Basis Qty p/u X1 X2 x3 s1 S2 S3

X2 8 10 1/3 1 0 1/3 -1/6 0

X3 10 8 2/3 0 1 -1/3 5/12 0

s3 18 0 -8/3 0 0 1/3 -17/12 1

Zj 26/3 10 8 2/3 5/3 0

Z= Cj - Zj -11/3 0 0 -2/3 -5/3 0
 Cj 5 10 8 0 0 0

Basis Qty p/u X1 X2 x3 s1 S2 S3

X2 8 10 1/3 1 0 1/3 -1/6 0

X3 10 8 2/3 0 1 -1/3 5/12 0

s3 18 0 -8/3 0 0 1/3 -17/12 1

Zj 26/3 10 8 2/3 5/3 0

Z=160 Cj - Zj -11/3 0 0 -2/3 -5/3 0
Simplex method and Sensitivity Analysis
1)Changes in Objective Function Coefficients. (Cj)
Variables included in the solution
Variables not included in the solution
2)Right Hand Side Ranging (RHS Ranging)
3)Changes in the Technological Coefficients.
 Changes in Objective Function Coefficients. (Cj)
Variables included in the solution
x2

Cj - Zj -11/3 0 0 -2/3 -5/3 0

Coeff in 1/3 1 0 1/3 -1/6 0

x2 row
quotient -11 0 ---- -2 10 ----

Least Least
negative Positive

Least negative means :- least negative value closest to zero on the number line.
Least positive means:- least positive value closest to the zero on the number line.
The least positive quotient (10) provides the
answer to how much the profit could increase
without changing the solution.
The least negative quotient (-2) is the
maximum decrease in the profit that would
not cause a change in the solution.

Accordingly, the range of profit is Rs 8 to Rs

20.
In case the profit per unit falls below Rs.8 or
becomes greater that Rs.20, the optimal
solution would be different, otherwise not.
Changes in Objective Function Coefficients. (Cj)
Variables not included in the solution
Right Hand Side Ranging (RHS Ranging)
Prepare a table as given below
Values in the Values in the Quotient
Quantity column appropriate slack 1 col/2col
variable column

Quantity S1 Qty/S1
8 1/3 24 Least Positive

10 -1/3 - 30 Least Negative

18 1/3 54
The least positive quotient is the answer to the
question as to how many hours can be decreased
in the fabrication department.
The least negative quotient is the answer to the
question as to how many hours in the department
can be increased.
Therefore:
Range: (60 – 24) + (60 + 30 ) = 90 hours. [36 ≤
60 ≤ 90]
Similarly for the finishing department: [xx ≤ 72 ≤
xx]
In packaging department, 18 hours are unutilised.
(because s3 qty is 18)
MaxZ = 70 x1 + 65 x2 + 80 x3 + 75 x4
S .T .
4 x1 + 4 x2 + 3x3 + 7 x4 ≤ 90
6 x1 + 3x2 + 5 x3 + 4 x4 ≤ 120
5 x1 + 2 x2 + 3x3 + 3x4 ≤ 60
6 x1 + 5 x2 + x3 + 2 x4 ≤ 100
x1 , x2 , x3 , x4 ≥ 0
 Cj 70 65 80 75 0 0 0

Basis Qty p/u X1 X2 x3 x4 s1 S2 s3 S4

X2 15 65 -1/2 1 0 2 1/2 0 -1/2 0

S2 25 0 -5/2 0 0 -1/3 1/6 1 -11/6 0

X3 10 80 2 0 1 -1/3 -1/3 0 2/3 0

s4 15 0 13/2 0 0 23/3 -13/6 0 11/6 1

Zj
Z=160 Cj - Zj -115/2 0 0 -85/3 -35/6 0 -125/6 0
 SPECIAL CASES
Multiple Optimal Solutions

Max =
Z 8 x1 + 16 x2
S .T .
x1 + x2 ≤ 200
x2 ≤ 125
3 x1 + 6 x2 ≤ 900
x1 , x2 ≥ 0
 Cj 8 16 0 0 0

Basis Qty p/u X1 X2 s1 S2 S3 RR

S1 25 0 0 0 1 1 -1/3 25
X2 125 16 0 1 0 1 0 125
X1 50 8 1 0 0 -2 1/3 50
Z = 2400 Zj 8 16 0 0 8/3
Cj - Zj 0 0 0 0 -8/3

Cj 8 16 0 0 0

Basis Qty p/u X1 X2 s1 S2 S3 RR

S2 25 0 0 0 1 1 -1/3 25
X2 100 16 0 1 -10 0 1/3 100
X1 100 8 1 0 2 0 -1/3 100
Z = 2400 Zj 8 16 0 0 8/3
Cj - Zj 0 0 0 0 -8/3
 INFEASIBILITY

Max=
Z 20 x1 + 30 x2
S .T .
2 x1 + x2 ≤ 40
4 x1 − x2 ≤ 20
x1 ≥ 30
x1 , x2 ≥ 0
UNBOUNDEDNESS:
When the solution is non-optimal (when
looking at the Cj-Zj row and there are negative
values in the RR column.

DEGENRACY:
When there is a tie in the RR column.
This problem has a solution! Methods to break
the tie :
1. Break it arbitrarily.
2. Choose the column that breaks the tie first.