Linear Programming

(An Optimization Technique)

General Optimization Problem Involves :
1. Decision Variables
2. Objective Criterion (Max/Min)
3. Constraints: (i) Equality (ii) Inequality

Linear Programming(LP) :
• It is one of the important “Optimization
Techniques.”
• Most Versatile, powerful & useful technique.
• Developed in 1947 by George B. Dantzig during
World War II to solve military problems, while
working with ‘US Air Force’.
• The word Linear in LP indicates -
relationship among variables both in objective
function and constraints.
• while Programming means - Systematic
Mathematical technique.
• General Linear Programming Problem (LPP)
with n variables and m constraints can be
stated as follows:
n
Optimize (Max.or Min.)Z = ∑ c j x j (1)
j =1
n

∑ a x (≤, =, ≥)b ; i = 1,2...m

j =1
ij j i (2)

x j ≥ 0; j = 1,2...n (3)
 Types of Variables
• Decision Variables : x are decision variables
j

• Slack Variables
n

∑a x ≤ b
j= 1
ij j i →

∑a ij x j + wi = bi ; i = 1,2...m
j= 1

wi are slack variables.

• Surplus Variables
n
∑ aij x j ≥ bi →
j= 1

n
∑a x
j= 1
ij j − si = bi ; i = 1,2...m

si are surplus variables

• Artificial Variables :
Required for solution methodology
• Components of LPP :
The three important components of LPP
are as follows:

(a) Decision Variables

(b) Objective Function - Max. or Min.
(c) Constraints-limitations.
 Assumptions of LP

1. Certainty:cj, aij , bi are known & constants.

2. Divisibility: Values of decision variable
can be integer or fractional.
3. Additivity: Total contribution = Sum of
contribution of all variables
4. Linearity: Relationship among variables
both in objective function & constraints.
• Formulation of LPP :
Steps to be followed.

1. To decide (define) decision variables.

2. To formulate objective function.

3. To formulate constraints.
CASE - 1 Data for LP Formulation.
Factory: Two products P1 & P2
Profit/piece of P1 = Rs.3/-
Profit/piece of P2 = Rs.5/-
Time Constraint:
Time required/piece of P1 = 3 hrs.
Time required/piece of P2 = 2 hrs.
Max. time available = 18 hrs.
Material Constraint:
Max. Material available for P1 = 4 pieces.
Max. Material available for P2 = 6 pieces.
CASE - 1 Formulation of LPP :
Let decision variable x1 = no of pieces of P1
x2 = no of pieces of P2
Max. Z = 3x1 + 5x2 Objective Equation

3x1 + 2 x 2 ≤ 18 − ( I ) Time Constraint

x1 ≤ 4 − ( II ) MaterialConstraint
x2 ≤ 6 − ( III ) MaterialConstraint
x1, x 2 ≥ 0 Non − negativityConditions
CASE - 2
If Time Constraint is modified to:
Time available is not less than 18 hours.
Max. Z = 3x1 + 5x2 Objective Equation

3x1 + 2 x 2 ≥ 18 − ( I ) Time Constraint

x1 ≤ 4 − ( II ) MaterialConstraint
x2 ≤ 6 − ( III ) MaterialConstraint
x1, x 2 ≥ 0 Non − negativityConditions
CASE - 3
If Time Constraint is modified to:
Time available is exactly 18 hours.
Max. Z = 3x1 + 5x2 Objective Equation

3x1 + 2 x 2 = 18 − ( I ) Time Constraint

x1 ≤ 4 − ( II ) MaterialConstraint
x2 ≤ 6 − ( III ) MaterialConstraint
x1, x 2 ≥ 0 Non − negativityConditions
 Types of Solution
1. Feasible Solution: Solution with values of
decision variables(xj) which satisfy all
Constrints & Non-negativity condition.
2. Basic Solution: For a set of m equations
in n variables (n>m), basic solution is a
solution obtained by setting (n-m)
variables equal to zero and solving for
remaining m equations in m variables.
No. of possible Basic solutions = nCm
= n!/(m!.(n-m)!)
(i) Basic Variables : Having values > 0.
(ii) Non-basic Variables :Having values = 0.
3. Basic Feasible Solution : It is a Basic
Solution which satisfies (3).
4. Non-degenerate/degenerate B.F.S
5. Optimal Basic Feasible Solution : It is a
Basic Feasible Solution which optimizes
objective function.
6. Unbounded Solution :
7. Infeasible Solution :
8. Unique/Alternative Optimal Solutions :
• Methods to solve LPP :
1. Graphical Method - for only two variables.
2. Simplex Method - Universal method.
3. Assignment Method - Special method.
4. Transportation Method - Special method.

Note : Methods (2), (3) and (4) are iterative

methods.
(1) Graphical Method
( for only two variables)
 Graphical Method Max Z = 3x1 + 5 x2
9 Case 1. s / t 3 x1 + 2 x 2 ≤ 18 − (I )
x1 ≤ 4 − ( II )
II x2 ≤ 6 − ( III )
Opt. Pt. (2,6) x1, x 2 ≥ 0
6
III

Zopt = 36.

3
F.R.
Z=15.
I
Z=0.
2 4 6 x1
 Max Z = 3 x1 + 5 x 2
Graphical Method
s / t 3 x1 + 2 x 2 ≥ 18 − ( I )
9 Case 2. x1 ≤ 4 − ( II )
x2 ≤ 6 − ( III )
II x1, x 2 ≥ 0
Opt. Pt.
6
(4,6) III
F.R.
Zopt = 42

Z=15.
I

2 4 6 x1
 Graphical Method Max Z = 3 x1 + 5 x 2
9 s / t 3 x1 + 2 x 2 = 18 − ( I )
Case 3.
x1 ≤ 4 − ( II )
II x2 ≤ 6 − ( III )
x1, x 2 ≥ 0
Opt. Pt.
6 (2,6) III

F.L.
Zopt = 36
3

Z=15.
I

2 4 6 x1
 Graphical Method
9 Special Case.
II Alternative Optimal
Solutions
6
III

3
F.R.
I

2 4 6 x1
 Graphical
9
Method
Special Case. Opt. Sol. (Max)
II
(Unbounded)
6 Opt. Pt.
(Min.) F.R.
(Unbounded)

III
3
I
Z line

2 4 6 x1
 Graphical
9
Method
Special Case.

Infeasible
Solution
3 F.R. Not
possible

2 4 6 x1
 x2
5
Opt. Pt. (0.5,
4 4.5)
Zopt = 12.5
2
F.R.
Z=6 III
-3 -2 -1 0 2 4 5 x1
-1
II -2 I Max Z = −2 x1 + 3 x 2
-3 s / t 2 x1 − x 2 ≤ 4 − (I )
− 3x1 + x 2 ≤ 3 − ( II )
Graphical -4
Method x1 + x 2 ≤ 5 − ( III )
x1, x 2 ≥ 0
Special Case.
• Graphical Method – to
discuss typical constraints
to identify regions of
constraints.
Typical Case How to plot → 3x1 + 2 x 2 ≥ −6 ?
x2 3x1 + 2 x 2 = −6
When x 32 = 0 → x1 = −2 ∴ Point( − 2 ,0 )
When x1 = 0 → x 2 = −3 ∴ Point( 0,−3 )
2

-3 -2 -1 0 1 2 3 4 5 x1

-1

-2

-3
Typical Case How to plot → 3x1 − 2 x 2 ≤ 6 ?
x2 3 x1 − 2 x 2 = 6
When 3 x 2 = 0 → x1 = 2 ∴ Point( 2,0 )
When x1 = 0 → x 2 = −3 ∴ Point( 0,−3 )
2

-3 -2 -1 0 1 2 3 4 5 x1

-1

-2

-3
Typical Case How to plot → x1 + x 2 ≥ 0 ?
x2 x1 + x 2 = 0
3 x1 = − x 2
When x1 = 1→ x 2 = −1 ∴ Point(1,−1 )
2
For arrow
1 x 2 = 0 → x1 ≥ 0

-3 -2 -1 0 1 2 3 4 5 x1

-1

-2

-3
 Recapitulate
• What is LPP ?
• Types of variables.
• Components of LPP.
• Assumptions of LP.
• Formulation of LPP.
• Types of solutions.
• Graphical Method.
• Unique & Alternative Optimal Solution,
Unbounded FR, Infeasible solution.
(2) Simplex Method
(Universal method)
• Simplex Method to solve
• Max. Problem with
• All ≤ constraints.
Max Z = 3x1 + 5 x 2 Simplex Method
s / t 3 x1 + 2 x 2 ≤ 18 − ( I ) Case 1.
x1 ≤ 4 − ( II )
x2 ≤ 6 − ( III )
x1, x 2 ≥ 0
Standard Form:
Max Z = 3x1+5x2+0w1+0w2+0w3
3x1+2x2+w1+0w2+0w3 = 18
x1+0x2+0w1+w2+0w3 = 4
0x1+x2+0w1+0w2+w3 = 6
To prepare initial Tableau:
Tableau - I
cj 3 5 0 0 0
ci xi bi x1 x2 w1 w2 w3
0 w1 18 3 2 1 0 0
0 w2 4 1 0 0 1 0
0 w3 6 0 1 0 0 1
Ij Z=0 3 5 0 0 0
•Ij = (cj -Ej) = Cj- (Σ aij .ci)
Interpretation of Tableau
Tableau - I
cj 3 5 0 0 0
ci xi bi x1 x2 w1 w2 w3 Ratio
0 w1 18 3 2 1 0 0 18/2 = 9
0 w2 4 1 0 0 1 0 4/0 = α
0 w3 6 0 1 0 0 1 6/1 = 6
Ij Z=0 3 5 0 0 0
• Key Column → Max +ve Ij
• Key Row → Min positive ratio.
How to get next tableau ?
• Leaving variable : w3
• Entering variable : x2
• If the key element is 1, then key row remain same in
the new simplex tableau
•If the key element is other than 1, then divide each
element in the key row (including b value) by the key
element to find new values of that row
•Make the other elements in the key column equal to
‘0’ by performning elementary row operations with
new key row obtained as above.
 Tableau - I
cj 3 5 0 0 0
ci xi bi x1 x2 w1 w2 w3 Ratio
0 w1 18 3 2 1 0 0 18/2 = 9
0 w2 4 1 0 0 1 0 4/0 = α
0 w3 6 0 1 0 0 1 6/1 = 6
Ij Z=0 -3 -5 0 0 0
R1 – 2R3 gives
2 2-2(1) = 0 , 3 3- 2(0) = 3 and 1 1 – 2(0) = 1
• 18 →18 - (6*2)/1 = 6
Tableau - II
cj 3 5 0 0 0
ci xi bi x1 x2 w1 w2 w3 Ratio
0 w1 6 3 0 1 0 -2 6/3 = 2
0 w2 4 1 0 0 1 0 4/1 = 4
5 x2 6 0 1 0 0 1 6/0 = α
Ij Z = 30 3 0 0 0 -5

• Key Column → Max +ve Ij

• Key Row → Min positive ratio.
Tableau - III
cj 3 5 0 0 0
ci xi bi x1 x2 w1 w2 w3
3 x1 2 1 0 1/3 0 -2/3
0 w2 2 0 0 -1/3 1 2/3
5 x2 6 0 1 0 0 1
Ij Z = 36 0 0 -1 0 -3
• This is the final Tableau.

The Optimal Solution is x1 = 2, x2 = 6

giving Z = 36
• Interpretation of
Simplex Method through
Graphical method.
 Interpretation of Simplex Method
9 Through Graphical Method

II
T2 Opt. Pt. (2,6)
6
III
T3 Zopt = 36.
Max Z = 3 x1 + 5 x 2
s / t 3 x1 + 2 x 2 ≤ 18 − ( I )
3
F.R. x1 ≤ 4
x2 ≤ 6
− ( II )
− ( III )
Z=15. x1, x 2 ≥ 0
I
T1
2 4 6 x1
 GATE - 2002
(1) A furniture manufacture produces Chairs & Tables.
The wood working department is capable of producing
200 chairs or 100 tables or any proportionate
combinations of these per week. The weekly demand
for chairs and tables is limited to 150 and 80 units
respectively.The profit from a chair is Rs. 100 and that
from a table is Rs. 300.
(a) Set up the problem as a Linear Program.

(b) Determine optimal product mix and optimal value

of objective function.
(c) If a profit of each table drops to Rs. 200 per unit,
what is the product mix and profit ?
 GATE - 2002
(a) Formulation

MaxZ = 100x1 + 300x2

x x
S/t 1 + 2 ≤1
200 100

x1 ≤ 150

x2 ≤ 80

x1, x2 ≥ 0
(b) Optimal Solution
x2

100 (40, 80) Z = 28000

80

50 Z = 15000

200

150 x1
(c) Multiple Optimal Solutions
x2

100 (40, 80) Z = 20000

80
75 Opt. Line

50 Z = 15000

200

150 x1
 GATE - 2003
(1) A manufacture produces two types of products, 1 and
2, at production levels of x1 and x2 respectively. The
profit is given 2x1 + 5x2. The production constraints
are :
x1 + 3x2 ≤ 40
3x1 + x2 ≤ 24
x1 + x2 ≤ 10
x1 > 0, x2 > 0

The maximum profit which can meet the constraint is

(A) 29 (B) 38 (C) 44 (D) 75
 x2

20 Ans. = (C) 44

Opt. Pt. (0,10) Z = 50

10
Z = 50

10 40
25 x1
 GATE - 2000

Solve : Max Z = 4x1 + 6x2 + x3

S/t 2x1 + x2 + 3x3 ≤ 5
x1, x2, x3 ≥ 0
If x2 ≤ 2 is added then what will be the
solution ?
 GATE - 2000
Solution : Through Simplex Method, in
two iterations the solution of
basic problem is :

x1 = 0, x2 = 5, x3 = 0 giving z = 30

If x2 ≤ 2 is added then the solution through

Simplex Method, in three iterations will be :

x1 = 3/2, x2 = 2, x3 = 0 giving z = 18
 GATE - 2008
Max Z = 4x1 + 6x2
3x1 + 2x2 ≤ 6
2x1 + 3x2 ≤ 6

x1, x2 ≥ 0

Q.1 After introducing slack variables w1 and w2, the initial

feasible solution is represented by the tableau below.
cj 4 6 0 0
ci xi bi x1 x2 w1 w2

0 w1 6 3 2 1 0
0 w2 6 2 3 0 1
Ij 0 -4 -6 0 0
 GATE - 2008
After some Simplex interactions, the following tableau is obtained.

cj 4 6 0 0
ci xi bi x1 x2 w1 w2

0 w1 2 5/3 0 1 -1/3
6 x2 2 2/3 1 0 1/3
Ij 12 0 0 0 2
 GATE - 2008
Q.2 The dual for given LP is :

(A) Min Z = 6u + 6v (B) Max Z = 6u + 6v

S/t 3u + 2v ≥ 4 S/t 3u + 2v ≤ 4
2u + 3v ≥ 6 2u + 3v ≤ 6
u,v≥ 0 u,v≥ 0

(C) Max Z = 4u + 6v (D) Min Z = 4u + 6v

S/t 3u + 2v ≥ 6 S/t 3u + 2v ≤ 6
2u + 3v ≥ 6 2u + 3v ≤ 6
u,v≥ 0 u,v≥ 0