Linear

Programming
 DISCLAIMER
“The content provided herein are created and owned by various authors and licensed
to Sorting Hat Technologies Private Limited (“Company”). The Company disclaims all
rights and liabilities in relation to the content. The author of the content shall be solely
responsible towards, without limitation, any claims, liabilities, damages or suits which
may arise with respect to the same.”
 Linear Programming

INTRODUCTION:
The mathematical definition of linear programming is optimizing a linear
function with some constraints in linear inequalities form.
The term “linear programming” consists of two words a linear and
programming. The term “linear” describes the relationship between
multiple variables with first degree. The term “programming” refers to
the process of choosing the best solution for a variety of alternatives
Linear Programming model are used in resource development. It can be
used in manufacturing, calculating how to allocate staff and equipment
to reduce operating costs.
It can be used in high-quality business activities, to determine what
products should be sold and at what prices to maximize the profits.
It can also be used in transportation, to determine how resources are
used to get work done in a limited amount of time.
In this chapter, we will solve problems of linear programming by
graphical method only, although there are many other ways to solve
those problems.

MATHEMATICAL MODELING:
Let understand the mathematical formulation by a simple example of
the manufacturing industry.
The industry produces chairs and tables. Cost and time require to
manufactured a chair is ₹ 2 and 3 hours and for table ₹ 4 and 2 hours
respectively. The factory can operate 150 hours maximum with budget
of ₹ 220 to produce these products per week. If each chair sells for ₹ 6
and each table sells for ₹ 7, then how many individual products should
be made this week to increase profits?
Let x be the number of chairs produced, and let y be the number of the
table produced. Then the first constrain in form of inequality is
2x + 4y ≤ 220 …(1)
Time constraints
Linear Programming

3x + 2y ≤ 150 …(2)
Further, two very common constraints are written below as It’s not
possible to produce less than 0 of any components, so constraints are

1.
 x ≥ 0
 …(3)
y ≥ 0
These are called non-negative constraints.
All 3 equations of the above form a system of constraints.

2x + 4y ≤ 220
 3x + 2y ≤ 150

 …(4)
 x ≥0
 y ≥0

Each chair costs ₹ 2 to make and sells for ₹ 6. This gives a profit of
₹ 4 per chair. Each table costs ₹ 4 to make and sells for ₹ 7. This gives a
profit of ₹ 3 per table. The profit function can be defined as
profit Z = 4x + 3y …(5)

IMPORTANT TERMINOLOGIES USED IN LINEAR PROGRAMMING:

With the help of the above example, we will understand key definitions
used in Linear Programming.
(a) Decision Variables:
The total number of chair and table units specified in x & y respectively
are my decision variables. Decision variables are under our control.
(b) Objective Function:
The function which we need to maximize or minimize to get the optimal
value. Z is our objective function(equation-5) as the company wants to
maximize the profit.
(c) Constraints:
The linear inequalities on the decision variables of LPP are called
constraints. In the above example, equation (4) contains all constraints.
(d) Non-Negative constraints:
In linear programming decision variables x and y are taken as non-
negative integers (equation-3).
Linear Programming

2.
GRAPHICAL SOLUTION OF LINEAR INEQUALITIES
IN TWO VARIABLES:
A line divides the XY plane into two parts and
each part is called half-plane as shown in the
figure.

A point within the XY- plane can either lie on

a line or can be either of the half-planes I or
II.

Steps to find solution region :

(i) Solution region in the xy plane is a collection
of points that satisfies the inequality
(ii) In order to find the solution region, it is
advised to take any point (generally (0, 0)) and
check whether it satisfies the inequality or
not.
(iii) For inequality Ax + By + C ≥ 0 or ≤ 0, solid line
AX + BY + C = 0 are also the part of solution
region.
(iv) For Ax + By +C > 0 or < 0, dotted line
AX + BY + C = 0 are not part of solution region.

Q. Solve 4x + 2y > 8 graphically

Sol. Step-1: Draw the dotted line 4x + 2y = 8 in the XY plane.

Step-2: Choose a point generally (0,0)
which does not lie on a line and put in
given inequality
0 > 8 which is false
So, the region (I), where (0,0) lies, does
not represent the solution region.
Hence, region (II) excluding the points
Linear Programming

on the line represents the solution

region.

3.
 Q. Solve 40x – 20y ≤ 120 graphically

Sol. Step-1: Draw the Solid line

40x – 20y = 120 in the XY plane.
Step-2: Choose a point generally (0,0)
which does not lie on a line and put in
given inequality
0 ≤ 120 which is right
So, the region (I), where (0,0) lies,
represents the solution region.

SOLUTION OF SYSTEM OF LINEAR INEQUALITIES IN TWO VARIABLES

Q. Represent solution region of linear inequalities on XY plane

x + y ≥ 6; x–y≤4

Sol. x + y ≥ 6 ...(1)
x – y ≤ 4 ...(2)
Two equations divide the XY plane
into four regions.
Step-1: First draw the solution region
of equation (1) and equation (2) as
shown in the figure.
Step-2: The solution region of inquality
(1) and inquality (2) is represented by
region (I), (II) and (II), (III) respectively.
Common region II is a required solution
region of a given system of inequality.
Linear Programming

4.
 Q. Represent solution region of linear inequalities on XY plane
x + 2y ≤ 6 ; 2x + y ≤ 6 ; x ≥ 0 ; y≥0

Sol. x + 2y ≤ 6 ...(1)
2x + y ≤ 6 ...(2)
x ≥ 0 ...(3)
y ≥ 0 ...(4)

Since x ≥ 0, y ≥ 0, the solution

region will lie in the first quadrant
only.
The solution region of inequality (1)
and (2), represented by region (I), (II)
and (II), (III) respectively, lies below the
lines.
Solution region is represented by
region (II).
ANALYSIS OF SOLUTION REGION BY THE
GRAPHICAL METHOD IN THE LPP MODEL
The collection of points that satisfies all of the
constraints is known as a feasible solution
and that region in XY plane is a feasible region
and region in which feasible solution does not
lie is known as infeasible region.

Fundamental Theorem
Theorem-1: Let S and Z = Ax + By be the
feasible region and the objective function
respectively for an LPP (Linear Programming
Problem). When objective function Z has
an optimal solution, then this optimal value
always lies on corner points.
Theorem-2: Let S and Z = Ax + By be the
feasible region and the objective function
Linear Programming

respectively for an LPP (Linear Programming

Problem). For bounded region S, maximum
and minimum values of Z generally exist at
corner points of the feasible region.

5.
 STEPS TO FOLLOW FOR THE SOLUTION OF THE LPP MODEL :
Steps Bounded Region Unbounded Region

1 Determine vertex Determine vertex points of the

points of the feasible feasible region
region
2 Evaluate objective Evaluate objective function Z at
function Z at corner corner points
points
3 Let M is maximum Let M is maximum and m is the
and m is the minimum minimum value of Z
value of Z
4 Max. of Z is M and min. M is the maximum value, If Ax +
of Z is m By > M does not pass through the
feasible region. Otherwise, no
maximum value exists
5 m is the minimum value, if Ax +
By < m does not pass through
the feasible region. Otherwise, no
minimum value exists.

Note: This method is also known as Corner Point Method.

Q. Solve the following linear programming problem graphically.

maximize z = x + 2y
subject to constraints
3x + 9y ≥ 18
y–x≥0
x ≥ 0 and y ≥ 0

Sol. The solution region of constraints are

shaded as shown in the figure.
Now evaluate the value of Z at vertex
points (0,2) and (1.5,1.5)
Linear Programming

Corner points z = x + 2y
(0,2) 4
(1.5,1.5) 4.5 = max

6.
 As solution region is un-bounded we
need to draw the x + 2y > 4.5 and
feasible region have common points
with solution region of In-equality x +
2y > 4.5
Hence no max value of z exist.

Q. Minimize z = x + 1.5y
subject to constraints
2x + 3y ≥ 12 ; 4x + 3y ≥ 12
find the optimal solution of objective function.

Sol. The solution region determined by the

given constraints is un-bounded and
corner-points are found by solving
linear-equations.
Now evaluate the value of z at corner points.
Corner points z = x + 1.5y
(6,0) 6 = min
(0,4) 6 = min
The value of z has the same value at
both corner-points.
Hence, all points lying on the line
segment joining (0, 4) and (6, 0) are
optimal solution.

Q. Solve the linear programming problem by graphical method :

maximize z = 4x + 5y, If possible subject to constraints
x – y ≤ –2 ; – x + y ≤ 0 ; x ≥ 0 ; y≥0

Sol. Draw the solution region of x – y ≤ –2

and – x + y ≤ 0 in first quadrant.
Solution region of x – y ≤ – 2 is
represented by green colour and of –x
+ y ≤ 0 by light Blue colour.
Linear Programming

As observed by graph, there is no

common solution region which can
form feasible region.
Hence maximum z = 4x + 5y does not
exists.

7.
 DIET PROBLEM

Q. A Dietician recommends having 6 units of vitamin E and 8 units of vitamin K in

their daily diet.
For this Ramakant consume two foods A and B. Food A contains 3 units / Kg
of vitamin E and 2 units / Kg of vitamin K. Food B contains 2 units / Kg of
vitamin E and 4 units / Kg of vitamin K.
If costs to Ramakant 40 ₹ per Kg to purchase Food-A and 60 ₹ per Kg to
purchase Food-B.
Formulate a mathematical model of LPP to minimize the expenditure of
Ramakant

Sol. Let Ramakant purchase x kg and y kg of food A and food B respectively.

Food Vitamin E Vitamin K cost
A 3 2 40
B 2 4 60
Quantity recommends by a 6 8
dietician
Constraints
3x + 2y ≥ 6
2x + 4y ≥8
x ≥ 0, y ≥ 0
Objective Function: Z = (40x + 60y)
The feasible region determined by
the given constrain is unbounded and
corner points are found by solving
linear equations.
Now evaluate the value of Z at corner
points
Vertex point Z = 40x + 60y
(0, 3) 180 = max
(1, 1.5) 130 = min
(4, 0) 160
Linear Programming

Now draw the graph of 40x + 60y < 130 as the region is unbounded and check
whether the half-plane has any point common with the feasible solution. It
is observed that there is no point common. So, 130 is the minimum value of
objective function Z.

8.
MANUFACTURING PROBLEMS :

Q. Elon musk has 3 machines in the Space-X manufacturing factory, named M-1,
M-2 and M-3. To produce rocket MARS-1 and MOON-2 all 3 machines are used
M-1 and M-2 have the capacity to run at most 12 hrs. and M-3 must run for
at least five hours a day. Time taken by each machine to produce one unit of
each item is given in the table
No. of hours required
Item
M-1 M-2 M-3
MARS-1 1 2 1
MOON-2 2 1 1.25
Elon can earn a profit of ₹ 600 and ₹ 400 on items MARS-1 and MOON-2
respectively. Find the optimal quantity of each rocket to maximize the profit
and maximum profit?

Sol. Let x units of MARS-1 and y units of MOON-2 are produced.

Objective function Z = 600x + 400y
Constraints
x + 2y ≤ 12
2x + y ≤ 12
x + 1.25y ≥ 5
x ≥ 0, y ≥ 0
the feasible region is bounded as
shown in the figure and corner points
are determined by solving constrained
equations.
Now evaluate the value of Z at corner
points
Corner Points Z = 600X + 400Y
(0, 6) 2400
(6, 0) 3600
(5, 0) 3000
Linear Programming

(4, 4) 4000 = max

(0, 4) 1600 = min
Max profit to Elon is ₹ 4000 and the optimal number of each rocket is 4.

9.
 TRANSPORTATION PROBLEM:

Q. Government of INDIA decides to distribute free ration for which the government
stored food in godowns A and B of the capacity of 1000 quintals and 500
quintals respectively. Government supply ration to three ration shop L, M,
and N whose requirements are 600,500 and 400 respectively. Transportation
costs from godowns to ration shops are given in the following table.
Transportation cost per quintal (in Rs)
From/To A B
L 60 40
M 30 20
N 25 30
Indian government wants to minimize the cost of transportation to continuously
supply free ration. What is the minimum cost?

Sol. Let Godown A supply x quintals and y quintals to L and M respectively.

For better understanding, we will draw a flow chart.

Constraints
x≥0 600 – x ≥ 0 500 – y ≥ 0 x + y – 600 ≥ 0 1000– (x + y) ≥ 0
y≥0 ⇒ x ≤ 600 ⇒ y ≤ 500 ⇒ x + y ≥ 600 ⇒ x + y ≤ 1000
Objective Function:
Total expenditure on transportation
Z = 60x + 30y + [1000 – (x + y)]25 + (600 – x)40 + (500 – y)20 + [x + y –600]30
Z = 25x + 15y + 41000
Linear Programming

10.
Corner points:
Corner points are determined by solving constraints equations
A(100, 500), B(500, 500), C(600, 400) and D(600, 0)

Now evaluate the value of Z at corner points

Vertex Point Objective Function Z = 25x + 15y + 41000
(100, 500) 51000 = min
(500, 500) 61000
(600, 400) 62000 = max
(600, 0) 56000
Hence minimum expenditure on transportation is ₹ 51000
And optimal quantity supply is given in the flow diagram.

Linear Programming

11.
 SUMMARY

• A linear Programming problem used to find the optimal value of objective

function under the constrain of linear in-equalities.

• LPP model used for

(i) Diet Problems
(ii) manufacturing Problems
(iii) Transportation Problems

• The solution region is common region represented by all linear

constraints.

• Optimal solution is a point or set of points in solution region which

maximize or minimize the objective function.

• For un-bounded region, maximum or minimum may not exist. However,

if it exists, it must occur at vertex point.

• If two corner points of the solution region both give same optimal
solutions. then every point on one segment joining these points is also
an optimal solution.
Linear Programming

12.
13.
14.