LINEAR PROGRAMMING PROBLEM

2
Learning Objectives
After studying this chapter, you should be able to
define the linear prOgramming problem.
-FORMULATION
know the assumptions of linear programming.
identify the characteristics of a inear
familiarize the advantages and programming
limitations of linear
problem.
formulation programming
understand the steps followed in the
of a linear problems.
programming problem.
An important or rather essential aspect of
selecting aparticular plan of action from any business or industrial
rse. in any such
activity, there s
amongst several alternatives activity
that
is
'planning. It means
chity. For example, a always an objective to be may be available. Of
tee made available to production manager of an achievedthe purpose of that
manager is to use thesehimself limited resources inindustrial concern has 'profite' as
men, materials and money. Thehis objective. He
Afarmer, a resources in such a way as to have the problem of the
business-man, a
ind or the other. In order to transporter, etc. each has an greatest profit, that may be possible.
plan of action, out of the achieve the objective, he must planobjective, and limited resources of one
several alternatives available to him. properly, that is, to decide upon a
Programming is just another word for 'planning'. But, it
scientific way. The objective
the relevant factors) and (goal) must be identified as a function means something
all the limitations on more-planning
of its variables (the in a
quantities of
inequalities involving those
The variables. resources must be stated as equations and/ or
word linear
indicates that the objective function is linear in its variables, and that the
equations and/or inequalities which represent limitations on resources are also linear.
TERMINOLOGY OF LINEAR
The terms
1.
that are
commornl y PROGRAMMING
used in
connection with a linear programming problem are as under:
the Decision Variables. Decision variables are the unknowns to be determined from the data of
the problem.
product For
example,
to be decision variables in a problem on product mix represent quantities of
a produced, in a diet-mix problem they represent the quantity of diterent foods, in
medium selection media, ec
problem they represent advertising units of different addvertising
 OR-2-2

The essential requirements of these variables are :

operations Resea
() Variables must be inter-related in terms of consumption of
(ü) Relationship among the variables should be linear. resources.
2. Constraints. In business, to attain aprofit from sales of a particular
selling price must be larger than the cost price. This type of relationship item, it is essential
quantities is known as an inequality. Such inequalities occur whenever there isamong two
dunequa
alimit on
supply, sales, etc. These inequalities represent limitations and as such are known as constraint
For example, let a company manufacture x pens of type A and ypens of type B. Each emand,
requires respectively u' and 'o' minutes of the machine shop capacity of which only ' minute
available. Then, the constraint on the production of pens may be expressed as
u.X + V.yS t.
3. Objective Function. An objective function represents an expression of the
system in terms of decision variables. major goal of %.
For example, if we assume that each pen of type A gives a profit of Rs. p and each pen of h
B gives a profit of Rs. q, then the objective of the manufacturer, say profit maximization,
expressed as may be
X.p ty.q = P
and called objective function of the producer.
The objective function is always non-negative. The coefficients associated with the variable in
the objective function are real constants and may represent per unit profits or costs.
4. Linear Relationships. Linear programming deals with
problems in which the objective
function, as well as constraints, are linear functions. Linear relationship,
properties : in general, have tw
() Proportionality which means that the relationships are directly
availability increases by some percentage, then the output shall also proportional, i.e., if resoure
percentage. For example, if one unit of variable consumes "* hours of capacity, increase by the same
consume "5x hours of capacity. five units shal
(ii) Divisibility which means that the
their fractional values are permitted in thevariables are not restricted to integer values, and as such
solution.
5. Non-negativity Restrictions. Since the
negative production (the state of dismantlingnegative values of the decision variables would mean
or state of destruction), these variables must assume
non-negative values. That is,
shall be expressed as under: if x and yare the decision variables, the non-negativity restrictions
x 0
6. Feasible Solution, A set of values of the and y 20.
the problem and non-negativity decision variables that satisfies all the constraints 0
7. restrictions well, is called afeasible solution of the problem.
as
Optimum Feasible Solution, Optimum feasible solution
provides the best feasible choice o
values-which yields the highest (in case if maximization) or lowest
of the objective function. (in case of minimization) vau
ASSUMPTIONS OF LINEAR PROGRAMMING
Following assumptions are kept in mind while solving a
(a) The problem involves only one linear programming problem
(b) All external factors are objective (e.g., profit maximization, cost minimization, etc)
stationary and they remain
price, variable cost and re[ource so over the period. That is, the selling
the entire range of the output. requirement per unit of the product Constant ove
(c) The available quantities of the remain
manufactured from these resources and the
nit of the producs
resources are
known with consumption
certainty.
per
Linear Programming Problem-Formulation

OR-2-3
units of a constrained resource Used equals sum of
The total
d) requirements of the products in a given product mix. the individual resource
(e) The products c¡n be produced in fractions and resources required in their manufacture too
fraction of a unit.
can be employed in
ADVANTAGES OF LINEAR PROGRAMMING
programming helps the management to make
production, i.e., raw material, manpower, money, machines, efficient
(a) Linear etc. use of the
factors of
programming provides information base from which
Linear can be taken.
(6) resources decisions on allocation of scarCe
Linear programming provides flexibility in analysing a
(c) problems. variety of multi-dimensional
improves the quality of
(d Linear programming
or intuition by objectiveanalysis. decision-making
by replacing rules of
thumb,
programming helps inimproving the skill and
providing an insight and prospective into the problemknowledge
(e) Linear situation. of the young executives by
Linear programming is a very good tool for adjusting to meet changing
() sales, deimand, etc. conditions, e.g.,
i Iinear programming helps in detecting bottlenecks in the production process. For example
ne machines may be over utilised and others may
have idle capacity. LP can strike an
optimal balance between the two situations.
THREE STAGES OF AN LPP
Each linear programming problem involves three stages, namely, identification, formulation and
solution :
() Identification involves identification of the available
alternatives of the problem,
establishing linear relationships between variables and specification of the constraints (e.g.,
space, material, money, etc.)
() Formulation of the problem involves construction of a
data. It requires identification of the decision variables, mathenatical mnodel from the given
specifying objective and setting
the
up linear equalities/inequalities for the constraints.
(m) Solution of the problem involves that particular set of values of the
lead to achieve the decision variables which
objective-maximum profit, least cost, etc.
FORMULATION OF AN LPP
Formulation of alinear programming problem means constructing a mathematical model from the
given data. This may be
done in the following six steps
Step 1. Study the
Step 2, to situation find the key decision to be made.
Step 3. Designate the various activities in terms of variables (known as decision variables).
Express the feasible alternatives in terms of decision variables.
theirStepunit4. cost
Tdentor ifyprofitthe objective and express it as alinear function of decision variables involving
Step 5. contributions.
in the Express
form of linear in words the influerncing factors or restrictionslconstraints, Put these restrictions
StepsSte1pto5)
6.
Express
in
equations)
the /inequalities involving the decision variables.
objective function, constraints, and non-negative restrictions
(identified in
linear programming format.
 OR-2-4

Example 1. Afirm is engaged in producing two products, A and B. Each unit Operalions Pea
requires two kg. of raw material and four labour hours for processing, whereas each of
Brequires three kg. of raw material and three hours of labour, of the same type. unit of produr
firm has an availability of 60 kg. of raw material and 96 labour hours. One unit of
yields Rs. 40 and one unit of product Bsold gives Rs. 35 as profit.
Everproduct
week, y
The market for the two products has been surveyed recently which suggests that a
20 units of product A and 10 units of product B can be sold per
Formulate this problem as a linear programming problem to determine as to how maximu
each of the products should be produced per week so that the firm can earn the many unit
maximum
Assume that there is no marketing constraint so that all that is produced can be sold.
Solution. The information given in the problem may be summarised as below :
pt%
Market demand
Product Raw material Processing time
(per week) Proftit per unit
(kg.) (Hrs.)
20
(Rs.)
A 2 4 40
B 3 3 10 35
Availability per week 60 96

We formulate the problem as an LPP in the following six steps :

Step 1. The key decision to be made is to determine the number of units of products Aam
to be produced per week.
Step 2. Let ,designate the number of units of product Aand x, the number of units of prod,
Bwhich the firm decides to produce per week.
Step 3. Feasible alternatives are the sets of values of x and x, such that x > 0 and z,2
since it is not possible to produce negative quantities.
Step4. The objective is fo maximize the total profit from the production of two types
products. Thus objective function is
Z = 40x + 35x,.
Step 5. Each unit of product Arequires 2 kg. of raw material while each unit of product Bnet
3kg. The total consumption, therefore, would be 2r + 3*2, which cannot exceed the to
availability of 60 kg per week. We can express this constraint as 2 + 3x, S 60. Similarly, it i
given thata unit of Arequires 4 labour hours for its production and one unit of Brequires 3hous
With an availability of 96 hours a week, we have 4x,+ 3x, S 96 as the labour hours constraint
Maximum possible sale of product Ais 20 unit and of product Bis 10 units. Therefore, we mus
have
x S 20 and x, s 10.
Step 6. The linear programming problem can be put in the following format :
Determine x, and x, so as to maximize z =40x, + 35x,;
subject to the constraints :
2x, + 3x, S 60 (Raw material constraint)
4x + 3x, S 96 (Labour hours constraint)
X, S20 and X, S 10 (Market demand constrainti
with non-negative restrictions x, X, > 0.
 1 20
legattve restric
EXafnple 6. The Bangalore Florist Company is planning to make up floral
arrangemer
the upcoming festival. The company has available the following supply of flowers at the
shown as under :
Type Number available Cost per flower (Re.)
Red roses 800 0.20
Gardenias 456 0.25
Carnations 4,000 0.15
White roses 920 0.20
Yellow roses 422 0.22
These flowers can be used in any of the four popular arrangements whose make-up and:
prices are as follows :
Arrangement Requirement Selling price
Economy 4red roses Rs. 6
2 gardenias
8carnations
Maytime 8white roses Rs. 8
5 gardenias
10 carnations
4yellow roses
Spring colour 9 red roses Rs. 10
10carnations
9 white roses
6 yellow roses
Deluxe rose 12 red roses Rs. 12
12 white roses
12 yellow roses
 Linear Programming Problem-Formulation
OR-2-9
programming problem which allows the florist
Formulate alinear
arrangement should be made company to determine how
up in order to maximize
manyunits of each profits assuming all
be sold
arrangementscan
formulated LP problem is not required to be solved.
The be made is to determine the number of unit
Solution. Step 1. The key decision to arrangements
colour and Deluxe rose.
Economy, Maytime, Spring denote the number of
oftype 2. Let Xy Ig 1g and y units of arrangement of Economy type,
Step colour type and Deluxe rose type respectively.
Maytimetype, Spring alternatives are x; > 0 j= 1, 2, 3, 4).
Stey3.
Feasible
arrangement type is :
4. The profitfor each 6.00 4(0.20) 2(0.25) - 8(0.15) = Rs. 3.50
Step =

Economy 5(0.25) - 10(0.15) - 4(0.22) = Rs. 2.77

=8.00 8(0.20)
Maytime =10.00 - 9(0.20) - 10(0.15) - 9(0.20) 6(0.22) = Rs. 3.58
Springcolour
= 12.00 12(0.20) - 12(0.20) - 12(0.22) = Rs. 4.56
Deluxe rose
objective function is :
Thus the Maximize z = 3.5x, + 2.77x, + 3.58x, + 4.56x4.
ia 1
to the florist company
E The maximnum quantity of the various type of flowers available yellow
din 2) carnations, 920 white roses and 422 roses. Thus, the
g ped roses, 456 gardenias, 4,000
in 3)
constraints are :
trictios + 9rz t 12x, S 800, 2x, + 5x, S 456, 8x, + 10x, + 10x, S 4,000,
nents 8x, + 9x, + 12x4 S 920, and 4x, + 6x3 + 12x, S 422.
the c Step 6. The given problem can thus be written as an LPP:
Maximize z = 3.5x, + 2.77x, + 3.58x, + 4.56x4;
e) subject to the constraints
4x1 + 9x3 + 12x, S 800 (Red roses constraint)
2x1 + 5x2 S 456 (Gardenias constraint)
8x + 10x, + 10x3 < 4,000 (Carnations constraint)
8x2 + 9x3 + 12x4A S 920 (White roses constraint)
nd
selt
4x, + 6x, + 12x, < 422 (Yellow roses constraint)
(Non-negative restrictions)
Example 7. Production
 System of Inequalities
suno one or moreof the linear inequalities have a common solution, then the
ofinegualities or simultaneous inequalities. The solution set of a inequalities are called
system of inequalities can
Asily be found by drawing the graphs of all the given inequalities. The procedure is shown below
with the help of few examples.
Example 1. Draw the graphs of the following system of inequalities and indicate the solution set :
h 2x + 3ss6 and X+ 4sS4.
Solution. Temporarily consider the inequalities as strict equations. So, we have
2x + 3s =6 and x + 4y = 4
In intercept form, these are re-written as
3 +;=1
2 and +
44 1
= 1.

Clearly, the graph of first equation meets axis of X at (3, 0) and axis of Yat (0, 2). Further, the
graph of second equation meets axis of Xat (4, 0) and axis of Yat (0, 1), The graphs of the two
equations are drawn in Fig. 3.4.
All points lying on the graph of 2x + 3y = 6 and to the left of it satisfy the inequlity
2r + 3y s 6,since 2 x0 +3x0 <6is true, . . the point (0. 0) satisfies first inequality. Further, all
points lying on the graph of x + 4y =4 and to the left of it satisfy the inequality x+4y S*, Smee
0+ 4x0s 4 is true, ie.,the point (0, 0) satisfies second inequality.
 OR-3-4

YA
Operations
3

2x +3y = 6
II
x+4y = 4

4
Feasible
region

Fig. 3.4,
region including
Hence, the commnon solution set of the givern inequalities is the shaded
lyingExample
on the graph of two
2. Draw the equations.
graphs of the following simultaneous inequalities and indicate
solution set: 2 0.
5x + 3ys 15, 3x + 4y S 12, and x >0, y
have
Solution. Temporarily consider the inequalities as strict equations. So, we x =0, y= 0.
5x + 3y = 15, 3x + 4y = 12 and
First two equations can be written as
+ = 1 and += 1.
3 5
Clearly, the graph of first equation meets X-axis at (3, 0) and Y-axis at (0, 5). Further, the gran
of second equation meets X-axis at (4, 0) and Y-axis at (0, 3). The graphs of these equations
drawn in Fig. 3.5.
Y

5A 5x + 3y = 15

2 3x + 4y = 12
Feasible P
1

D
1 3\B 4 5
X

Fig. 3.5.
All points lying on AB and below it satisfy the inequality 5x + 3y s 15. Also, all points lying
on below it satisfy 3x + 4y S 12. (since 5x0 + 3x0 s 15 and 3x0 + 4x0 s
Eurthér, all points lying on the Y-axis (i.e., x = 0) and to the right of it satisfy12).x 2 0 and
noints JyiDg on te A-àxis (.e., y = 0) and above it satisfy y 0.
ade, the common solution set is the shaded regio OBPC 35
 Example 1.(Maximization Case) Acompany produces two types of pens, say Aand B.
Profits on pens A and B are Rs.
1S of superior quality and pen Bis of lower quality. 5 and Rs. Pen
pen respectively. Raw material required for each pen A is twice as that of pen B. The supply of
material is sufficient only for 1,000 pens of type B per day. Pen A requires a special clip and
400 such clips are available per day.For pen B, only 700 clips are available per day. Find graphi
graphica
the products mix so that the company can make a maximum profit.
Solution. Let x, y be the number of type A and type B pens respectively which the
compat
decides to produce per day. The given problem, then, has the following linear programi
formulation :
Maximize z = 5x + 3y; subject to the constraints :
2x + ys 1,000 (for raw material)
x S 400, y S 700 (for clips)
x>0 and y >0. (Non-negativity)
 Lnear
P

We iirrst graph the constraints.

Stey 1. +y s 1,000) represents the OXY Consider
region plane, OR-3-9
represents the region blow' the line x of= the plane 'on i.e., first
Cearly,(2r
400) 70O.
left of the
400 and (y s 700) quadrant of
line 2x + the
y plane.
s
ofthe
line y =
The solution space is the
intersectionfeasible
of all the
represents the region =1,000,
on left
region regions
Step 2. the first quadrant. Thus, we get the
constraintsin OABCD represented
shown shadedbyintheFig.given
3.13
below:
X= 400
D
2x+y= 1000
Feasible B
region

-y=700

Fig. 3.13.
. Since, the solution space is the region OABCD bounded by the lines in the plane, its
points) are :
extreme points (corner A = (0, 700), B= (150, 700), C= (400, 200) and D= (400 o
.0= (0, 0),
Ctoy 4. Values of the objective function at the extreme points are now evaluated :
Extreme point X-coordinate y-coordinate Profit : (5x + 3y)
A 0 700 2,100
B 150 700 2,850
C 400 200 2,600
D 400 2,000
Step 5. Since, the value of objective function is largest at the extreme point B = (150, 700), the
opimal solution to the given LPP is :
x= 150 and y = 700 with maximum z =2,850.
maximumSthe company must produce 150 pens of type A and 700 pens of type B to achieve a
profit of Rs. 2,850.
uires(Minimization
Tequires a minimum of 80 units Case) A company
of chemicalA thatunits
and 60 produces soft drinksBtohasgo ainto
of the chemical contract that
each bottle
of the
Khas adrink. The chemicals are available in a prepared mix from two different suppliers. Supplier
of Aand mix of 4
1 unit units of A and 2 units of B that costs Rs. 10, and supplier X, has a mix of 1 unit
the of B that costs Rs. 4.
How many mixes from company X, and company X, should
cSolompautionn.y
purchase to honour contract requirement and yet minimize cost?
The given data may be summarized as follows :
Supplier Chemical A (units) Cost per mix (Rs.)
Chemical B (units)
4 10
2
Units required 1
1 4
80 60
OR-3"10

Let x be the number of mixes to be purchased from supplier X and y be

Operations Resa
Supplier Xy. Then, since the objective is to minimize the total cost, the given
tollowing linear programming formulation : problemthose fho
Minimize z = 10x + 4y; subject to the constraints :
4x + y 80
2x + y2 60 (Chemical A)
(Chemical B)
x 0 and y 2 0.
(Non-negativity)
Step 1. We first graph the constraints. Consider an OXY plane. Clearly, {4x + y
{2r + y 60) represent the regions 'on the right of the lines 4x + y = 80 and
respectively.
Stey 2. The solution space is the intersection of all the regions represented by L
constraints in the first quadrant. This is shaded in Fig. 3.14 which appears to be unbound
X

09=h+7
Feasibe
region

4x +y= 80

Fig, 3.14.
Now, we try to determine the additional hidden conditions in the problem for which t
solution space (feasible region) becomes bounded.
Each bottle of the drink contains 80 units of chemical Aand 60 units of chemical B. If the drn
are made from the mixes of supplier X, only, each bottle would require 20 mixes to meet th
chemical Arequirement and 30 mixes to meet the chemical B requirement. Therefore, 30 mixe
would meet all minimum requirements. Any lesser amount of mixes would not meet the minimur
requirement for chemical B.
Similarly, if the drinks are made from the mixes of supplier X, only, each bottle would requir
80 mixes to meet both the chemical requirements.
Since, the company that manufactures drinks wants to make a blending of the mixes purchae
from both the supplier X, and X, it is possible that the company would need less than
maximum amount of both supplies to meet the guaranteed minimum.
Thus, we add the two additional hidden conditions x s 30, v s 80. With these additon
restricions, the solution space becomes bounded as shown shaded in Fig. 3.15.
Step 3. As the feasible region ABCD is bounded by the lines 4x + y = 80, 2x + y =
and y= 80, its extreme points are:
A = (0, 80), B = (10, 40), C= (30, 0) and D = (30,
Step 4. Values of the objective function at the extreme points are now 80).
evaluated :
Extreme point X-coordinate y-coordinate Cost : (10x + 4y)
0 320
80
B 10
40 260
30
300
D 30
80 620
 OR-3-11
Problem--Graphical Method
Programming P
Linear
t=30
X |D

Feasible

2r+y= 60
4x +y 80 B
region

-y= 80

Fig, 3.15.
solution is :
Stm 5 Since, objective function is least at the extreme point (10, 40), the optimal
x= 10 and y =40 with minimum z = Rs. 260.
Thus, to honour the contract and yet to minimize cost, the comparny should purchase 10 mixes
rom Xand 40 mixes from X.
Remarks. Itmay be observed that it was not essential to identify hidden constraints to make the
feasible region bounded. In fact unbounded solution space does not pose any problemn in the case of
Minimizatiom problems.