Qabd Unit 2 PDF

UNIT-2 Linear Programming Method and Transportation
2.1 Problem
Unit
Linear Programming
2 Method and
Transportation Problem
Learning Objectives
After studying this unit, one would be able to understand,
v Concept, Structure, Assumptions and Application areas of LPP.
v Guidelines for Formulation of LPP – Formulation of LPP for different areas.
v Concept of LPP by Graphical Method – Extreme Point Method.
v Concept of Simplex Method and procedure for solving LPP through Simplex Method.
v Concept of Converting Primal LPP to Dual LPP.
v Advantages and Limitations of LPP.
v Mathematical Model for Transportation Problems.
Methods of Solving Initial Feasible Solution using NWCR, LCM and VAM Methods.
v
Optimality Test by Modi Method.
v
v Concept of Unbalanced Supply and Demand and Degeneracy in TP.
Introduction
Linear programming problems deal with optimal allocation of resources which are limited. The main objective
of an LPP is to optimize the given problem i.e., either profit maximization or cost minimization. Firms generally
desire to maximize profits and minimize losses and investment cost.
Transportation problem is a special type of linear programming problem in which goods or products are transferred
from sources to destination for minimizing the total cost of transportation. The main aim of the transportation
problem is to reduce the transportation cost and fulfill the needs of the individuals located at the destinations.
Transportation model includes many activities such as, transporting the product from plants to warehouses,
warehouses to wholesalers, wholesalers to retailers and retailers to customers. The transportation model
is effectively used in scheduling, production, investments, plant location, inventory control, employment
scheduling, product mix problems and so on.
SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

2.2 Quantitative analysis for business decisions
2.1 Linear Programming – Concept, structure, Assumptions and

application areas of LPP
Q1. What is linear programming? What are its requirements? Discuss the structure/components of LPP.
Answer : Model Paper-I, Q4(a)
Linear Programming
Linear programming problem is defined as a method which is adopted to optimize an objective function which is
subjected to set of constraints. It is named as linear programming since objective function and set of constraints are linear
equations.
Requirements of a LPP
To solve a LPP some basic requirements must be fulfilled. They are,
1. Objective function availability
2. Set of constraints
3. Both objective function and constraints must be linear functions.
Structure/Components of LPP
The major components of LPP are depicted in the following figure,
Components of LPP
Objective Constraints Decision

function variables
Demand Supply Equality

constraints constraints constraints
Figure
1. Objective Function
This is the function which is formulated by using contributions (i.e., profits, cost) and this function is to be optimized
i.e., to maximize the profit and minimize the loss respectively.
It is formulated by considering contributions and decision variables. It should be a linear equation.
Example
Profit per doll of type A = 30/- and profit per doll of type B = 40/-. Express the objective function.
If profit is given, objective function is to be maximized and if cost is given it should be minimized.
Given that profit per doll of type A = 30/-
So, if ‘x’ dolls of type A are produced, total profit would be 30x.
Similarly, if ‘y’ dolls of type B are produced, then total profit would be 40y.
Objective function is denoted by ‘Z’. It can be written as,
Z = 30x + 40y
∴ Maximize Z = 30 x + 40 y
This is the expression for objective function.

2. Constraints
An objective function is always limited by some restrictions called as constraints.
Example
An examination is conducted for 100 marks. So, objective of a student is to obtain maximum marks which is limited
to only 100 that is the student can require exactly 100 or less than 100. Therefore the objective is said to be subjected to
some constraints.
UNIT-2 Linear Programming Method and Transportation 2.3 Problem
Constraints are further classified as, Example
(i) Demand constraint If 3x1 + 5x2 ≥ 15 is in equation, then S2 is surplus
(ii) Supply constraint variable, it should be added on R.H.S so,
(iii) Equality constraint. 3x1 + 5x2 = 15 + S2 ⇒ 3x1 + 5x2 – S2 = 15
(i) Demand Constraint Where, x1, x2, S2 ≥ 0.
Demand is analogous to requirement. It has lower (c) Artificial Variables
limit but upper limit is not mentioned.
The variable which should be added in case of
Example minimization Linear Programming Problem (LPP)
At least 50,000/- is required to buy an excellent quality along with surplus variable is called an artificial
guitar. variable.
Therefore, in demand constraint (≥) inequality is used. Example
(ii) Supply Constraint Let, Z = 40x + 120y be minimization function i.e.,
Supply is related to availability. So it has upper limit. Min Z = 40x + 120y
Example Subject to constraints, x – 3y ≥ 8
From given quantity of raw material, not more than – 15x + 40y ≤ – 600
100 products can be produced. Since, a constraint has ≥ inequality we make it into
Therefore, in supply constraint (≤) inequality is used. equality by deducting surplus value (S1) and by adding an
artificial variables (A1).
(iii) Equality Constraint
Here, the availability and requirement should be Such that, x – 3y – S1 + A1 = 8
exactly equal. S1 = Surplus variables
Example A1 = Artificial variable.
Customer who purchase a pair of shoe will not
compromise with the feet size. He wants the exact and – 15x + 40y + S2 = – 600
size of shoe. S2 = Slack variable.
Therefore, equality constraint (=) symbol is used. Q2. State the assumptions of LPP.
3. Decision Variables Answer :
The variables which are required to be determined A Linear Programming Problem (LPP) has the
using different techniques are said to be decision variables. following assumptions,
Example 1. Proportionality
The variables in objective function x, y are nothing This assumption is for both objective function and
but the decision variables. constraints.
(a) Slack Variables The contribution of each activity to the value of the
When a supply constraint is given i.e., ‘≤’ type of objective function Z is proportional to the level of the activity
inequality. Thus, in order to change inequality to the x j
as represented by ‘cj xj’ term in the objective function.
equality a non-negative variable should be added on The contribution of each activity to the left hand side
left side of an equation. of each functional constraint is proportional to the level of
Example the activity xj as represented by ‘aij xj’ term in the constraint.
This is also called as linearity assumption.
If 3x1 + 5x2 ≤ 15 is an equation then (s1) a slack
variable must be added on L.H.S of inequality. So, it 2. Additivity
becomes, Every function in a linear programming model
(whether the objective function or the function on the
3x1 + 5x2 + S1 = 15 left hand side of a functional constraint) is the sum of the
Where, x1, x2, S1 are ≥ 0 individual contributions to the respective activities.
(b) Surplus Variables Example
When a requirement constraint is given i.e., ‘≥’ type The total cost incurred by two productions x and y must
of inequality. Thus, in order to change inequality to be equal to the sum of costs incurred separately by x and y.
equality, variable should be added to inequality on Similarly, the amount of resources consumed by x and y must
right hand side of an equation. be equal to the sum of resources used for x and y individually.
3. Divisibility 6. Military Applications
Decision variables in a linear programming problem LP models can be used for various military operations
are allowed to have any values including non-integer values like wars, terrorists and defense against communal clashes
that satisfy the functional and non-negative constraints. Thus, etc. The problem is to determine the optimum number of
these variables are not restricted to just integer values. Since, weapons to be used that minimizes the cost of operations.
each decision variable represents the level of some activity,
7. Agricultural Applications
it is being assumed that the activities can be undertaken at
fractional levels. If only integer variables are desired as LP can be applied in agricultural planning for
in case of machines or men as decision variables then the allocating the limited resources such as acreage, water,
integer programming method may be applied to get the labour, working capital etc., so as to maximize the net
desired values. income.
4. Certainty (Deterministic) 8. Facilities Location
The value assigned to each parameter of an LP model LP determines the most economic and efficient
is assumed to be known as constant. In real applications, the manner of locating manufacturing plants and distribution
certainty assumption is thus satisfied precisely. centres for physical distribution of goods.
In such cases there may be random variables and they 9. Hydel Power Generation
are solved using parametric programming or stochastic LP
In this problem, variations in storage of dams which
model or by conducting sensitivity analysis.
generate power is determined so as to maximize the energy
5. Multiplicatively obtained from the entire system.
It can be explained with the help of an example if one 10. Portfolio Selection
person takes two hours on one machine, 10 persons will take
20 hours on the same machine. LP gives the optimal solution to the selection
of specific investments from among a wide variety of
Q3. Explain various applications of LPP.
alternatives. This helps the managers of banks, insurance
Answer : companies, investment services etc.
The various applications of linear programming are
as follows,
2.1.1 Guidelines for Formulation
of LPP – Formulation of LPP for
1. Production Management
Different Areas
Linear programming can be applied in production
Q4. How LPP can be formulated? Discuss
management for determining product mix, product
mathematical formulation of LPP.
smoothing, assembly time balancing, production scheduling
and blending problem. Answer :
2. Marketing Management The various steps/guidelines involved in the
formulation of LPP are as follows,
Linear programming can be applied in marketing
area by helping in analysing the effectiveness of advertising (a) Identification and Selection of Variables
campaign, advertising media mix and the shortest route for The first step in the formulation of LPP is identification
travelling salesman. of the variables called decision variables or design variables.
3. Manpower Management After the identification of variables, it is then expressed
LP allows the human resource to analyse personal in the form of symbolic notation along with their units of
policy combinations in terms of their appropriateness for measurement. These variables may be seen as number of
maintaining a steady-state flow of people into the firm and units to be ordered, number of units to be sold, number of
out of the firm. units to be manufactured etc.
4. Transportation Problem (b) Construction of Objective Function
LP helps in determining the optimum transportation The second step in the formulation of LPP is the
schedule with minimum total transportation cost of moving setting of goal or objective function. It may be in two forms
goods from various origins to various destinations. i.e., maximization or minimization form. If the data reveals
profit in the system, then the objective function will be in
5. Assignment Problem maximization form. For example, output, profit, sales etc.,
LP models give the best assignment schedule with are to be maximized. If the data is in terms of cost, then the
minimum total cost of assignment of various resources to objective function will be written in the minimization form.
various activities on a one-to-one basis. For example, production cost, spoilage, overheads, loss etc.
2.5 Problem
(c) Developing the Constraints Equation The relevant data is given in the table. If the whole labour
The third step in the formulation of LPP is the is engaged in manufacturing product A, 1600 units of this
development of the constraints equation. Constraints are product can be produced. There is demand for at least 300,
the restrictions applied to the objective function. These 250 and 200 units of products A, B and C and the profit
constraints may be demand, supply and equality constraints. earned per unit is ` 50, ` 40 and ` 70 respectively. Formulate
Demand constraints are represented by ‘ ’ sign and is used the problem as linear programming problem.
when a minimum restriction is applied to the function.
Requirement per
Supply constraints are represented by ‘ ’ sign and Raw unit of product (kg) Total
issued when a maximum restriction is applied to the function. material Availability
A B C
Equality constraints are represented by ‘=’ sign and P 6 5 9 5000
is used when an exact value should be used.
Q 4 7 8 6000
(d) Deciding the Variable Status
The fourth step in the formulation of LPP is deciding Assume the number of units to be produced of products
the status of decision variable. As the variables are generally A, B and C be x1, x2, x3 where x1 0, x2 0 and x3 0.
numbers, they may be positive, negative or zero. For most ∴ The objective is to maximize the profit.
of the cases it will be positive. For example, allocation of
material for transportation cannot be negative, it should be The objective function can be written as,
either zero or greater than that. But, in some cases, it may Max Z = 50x1 + 40x2 + 70x3
be positive or negative, like in inventory. Therefore, the
condition of the variable can be non-negative or unrestricted. Constraints can be formed as,
Mathematical Formulation of LPP 6x1 + 5x2 + 9x3 5000
LP is the optimum allocation of the available 4x1 + 7x2 + 8x3 6000
resources on the basis of the given criterion. The LPP can
be mathematically represented as follows, The constraints on the number of units to be
manufactured can be expressed as,
Let,
Z = Overall optimum value. 1 1
x1 + x2 + x3 1600
xi = Decision variables (Where, i = 1, 2, 3, ..., n). 2 3
Ci = Constants multiplied to each level of (Since, B requires 1/2 and C requires 1/3, the time
activity depending on their contribution required for A)
to Z.
Market demand requires,
bj = Resource available for the allocation.
x1 300
aij = Resource consumed by each activity.
x2 250
Therefore, the LPP form of the above data can be
written as, x3 200
(Max or Min ) Z = C1 x1 + C2 x2 + C3 x3 + ... + Cnxn As A, B and C are to be produced in the ratio of 3 : 4
Subject to, : 5, we get,
a11 x1 + a12 x2 + a13 x3 + .... or or = b1 x1 : x2 : x3 :: 3 : 4 : 5
a21 x1 + a22 x2 + a23 x3 + ... or or = b2
x1 x2
a31 x1 + a32 x2 + a33 x3 + ... or or = b3 ∴ =
3 4
Where, x1, x2, x3, ...., xn 0.
x2 x3
Q5. How do you formulate LPP in production and and =
4 5
marketing? Illustrate.
∴ Two additional constraints obtained are,
Answer :
4x1 – 3x2 = 0
LPP Formulation in Production – Illustration
5x2 – 4x3 = 0
A firm manufactures three products A, B and C. Time
to manufacture product A is twice that for B and thrice that ∴ The above relations obtained forms the linear
for C and they are to be produced in the ratio of 3 : 4 : 5. programming model.
LPP Formulation in Marketing – Illustration Thus, x1 + x2 + x3 = 50,00,000
A marketing manager wishes to allocate his annual Through bond issue, the company can obtain
advertisement budget of ` 20,000 in two media vehicles A `20 lakhs.
and B. The unit cost of a message in media A is ` 1000 and
Thus, x1 ≤ 20,00,000
that of B is ` 1500. Media A is a monthly magazine and not
more than one insertion is desired in one issue. At least 5 Financial corporation ‘A’ can provide loan upto
messages should appear in media B. The expected effective `30 lakhs.
audience for unit messages in the media A is 40,000 and Thus, x2 ≤ 30,00,000
for media B is 55,000. Develop a mathematical model for
maximizing total effective audience. The condition of A, states that the bond debt and loan
from B should not be more than twice the amount of A.
Maximize Z = 40000x + 55000y
Thus, x1 + x3 ≤ 2x2
Subject to 1000x + 1500y 20000
Financial corporation ‘B’ can also provide loan upto
x 12
`30 lakhs.
y 5
Thus, x3 ≤ 30,00,000
and x, y 0
Therefore,
Q6. Illustrate the formulation of LPP in financial
management. Min z = 0.12 x1 + 0.16 x2 + 0.18 x3
Answer : STC,
The board of directors of a company has given x1 + x2 + x3 = 50,00,000
approval for the construction of a new plant. The plant will
require an investment of `50 lakh. The required funds will x1 ≤ 20,00,000
come from the sale of a proposed bond issue and by taking x2 ≤ 30,00,000
loans from two financial corporations. For the company, it
x3 ≤ 30,00,000
will not be possible to sell more than `20 lakh worth of bonds
at the proposed rate of 12%. Financial corporation A will give x1 + x3 ≤ 2x2
loan upto `30 lakh at an interest rate of 16% but insists that and x1, x2, x3 ≥ 0.
the amount of bond debt plus the amount owned to financial
corporation B be no Z more than twice the amount owned Q7. Explain the formulation of LPP in personnel
to financial corporation A. Financial corporation B will loan management.
the same amount as that loaned by financial corporation A Answer :
but it would do so at an interest rate of 18%. Formulate this
A machine tool company conducts a job training
problem as an L.P model to determine the amount of funds
program for machinists. Trained machinists are used as
to be obtained from each source in a manner that minimizes
teachers in the programme at a ratio of one for every
the total annual interest charges.
ten trainees. The training lasts for one month. From past
Objective Function experience it has been found that out of ten trainees
In the given problem, the objective function is to hired, only six complete the programme successfully. The
unsuccessful trainees are released. Trained machinists are
minimizes the total annual interest charges, after fulfilling
needed for machining and company’s requirements for the
the requirement of `50 lakhs for constructing new plant. next three months are January: 120, February: 180 and
Let, March: 220. Further, the company requires 250 trained
‘x1’ – Be the bond debt machinists by April. There are 150 trained machinists
available at the beginning of the year. The relevant costs per
‘x2’ – Be the loan from A month are,
‘x3’ – Be the loan from B
Each trainee : ` 1,000.00
The interest rates of x1, x2 and x3 are 12,16 and 18
percents respectively. Each trained machinists : ` 2,000.00
Thus, minimize z = 0.12 x1 + 0.16 x2 + 0.18 x3 (Machining or Teaching)

Subject to Constraints Each trained machinist idle : ` 1,500.00
The requirements of new plant is `50 lakhs which has Formulate as an LPP that will result in minimum cost
to be fulfilled from these three sources. hiring and training schedule and meet the company’s need.
2.7 Problem
LP Model Formulation
Let,
x1, x2 = Trained machinist teaching and idle in January respectively
x3, x4 = Trained machinist teaching and idle in February respectively
x5, x6 = Trained machinist teaching and idle in March respectively.
LP Model
Minimize (total cost) Z = Cost of training programme (teachers and trainees) + Cost of idle machinists
+ Cost of machinists doing machine work (constant)
= 1000 (10x1 + 10x3 + 10x5) + 2000 (x1 + x3 + x5) + 1500 (x2 + x4 + x6)
Subject to the constraints,
(i) Total trained machinists available at the beginning of January = Number of machinists doing machining + Teaching + Idle
150 = 120 + x1 + x2
∴ x1 + x2 = 150 – 120 = 30 ... (1)
(ii) Total trained machinists available at the beginning of February= Number of machinists in January + Joining
after training programme
150 + 6x1 = 180 + x3 + x4
6x1 – x3 – x4 = 180 – 150 = 30 ... (2)
(iii) In January there are 10x1 trainees in the programme and out of those only 6x1 will become trained machinists.
Total trained machinists available at the beginning of March = Number of machinists in January + Joining after
training programme in January and February
150 + 6x1 + 6x3 = 220 + x5 + x6
6x1 + 6x3 – x5 – x6 = 220 – 150
6x1 + 6x3 – x5 – x6 = 70 ... (3)
(iv) Company requires 250 trained machinists by April
150 + 6x1 + 6x3 + 6x5 = 250
6x1 + 6x3 + 6x5 = 250 – 150
6x1 + 6x3 + 6x5 = 100 ... (4)
LP Model (Simplification)
Minimize (Cost of hiring and training schedule)
Z = 1000 (10 x1 + 10 x3 + 10 x5) + 2000 (x1 + x3 + x5) + 1500 (x2 + x4 + x6)
= 10000 x1 + 10000 x3 + 10000 x5 + 2000 x1 + 2000 x3 + 2000 x5 + 1500 x2 + 1500 x4 + 1500 x6
Z = (10000 x1 + 2000 x1) +( 1500 x2) + (10000 x3 + 2000x3) + (1500x4) + (10000x5 + 2000x5) + (1500 x6)
Z = 12000 x1 + 1500 x2 + 12000 x3 + 1500 x4 + 12000 x5 + 1500 x6
Subject,
x1 + x2 = 30 ... (1)
6x1 – x3 – x4 = 30 ... (2)
6x1 + 6x3 – x5 – x6 = 70 ... (3)
6x1 + 6x3 + 6x5 = 100 ... (4)
Standard form of LP Model
Minimize,
Z = 12000 x1 + 1500 x2 + 12000 x3 + 1500 x4 + 12000 x5 + 1500 x6

Subject to the constraints, = ` [2,000x1 + 2,250x2 + 2,000x3] –
x1 + x2 ≥ 30 ` [150x1 + 170x2 + 125x3]
6x1 – x3 – x4 ≥ 30 = ` (1,850x1 + 2,080x2 + 1,875x3)
6x1 + 6x3 – x5 – x6 ≥ 70
Therefore, the objective function is,
6x1 + 6x3 + 6x5 ≥ 100
Maximize, z = 1,850x1 + 2,080x2 + 1,875x3 ... (1)
x1, x2, x3, x4, x5, x6 ≥ 0
Constraints
Q8. Write about the formulation of LPP in agricultural
Since, the total area is restricted to 100 acres and
aspect.
the total man-days labour is restricted to 400 man-days, the
Formulation of LPP in Agriculture: constraints are,
A farmer has a 100 acre farm. He can sell all the x1 + x2 + x3 ≤ 100
tomatoes, lettuce, radishes he can raise. The
price he can obtain is ` 1 per kg for tomatoes, 5x1 + 6x2 + 5x3 ≤ 400 ... (2)
` 0.75 a head for lettuce and ` 2 per kg for Non-negativity Restrictions
radishes. The average yield per acre is 2000 kg
of tomatoes, 3000 heads of lettuce and 1000 kg x1, x2, x3 ≥ 0 ... (3)
of radishes. Fertilizer is available at ` 0.5 per kg The relations equations (1) and (2) give a complete
and the amount required per acre is 100 kg each
linear programming model for the problem.
for tomatoes and lettuce and 50 kg for radishes.
Labour required for sowing, cultivating and
harvesting per acre is 5 man-days for tomatoes
2.2Solving of LPP by graphical
and radishes and 6 man-days for lettuce. A method – Extreme point method
total of 400 man-days of labour are available Q9. What is graphical method of LPP? What are
at ` 20 per man-day. Solve the LPP in order its characteristics and limitations. Explain the
to maximize the farmer’s total profit. procedure for generating an extreme point
Answer : solution to an LPP by graphical method.
Decision Variables Answer :
Let, x1 denote the number of acres allotted for tomato Graphical Method
cultivation, x2 denote the number of acres allotted for lettuce Graphical method is a simple method to understand
cultivation and let x3 denote the number of acres allotted for and also to use. This is effectively used in LPP’s which
radish cultivation. involves only 2 variables. It gives the graphical representation
Objective Function of the solutions.
All types of solutions are highlighted in this method
The objective is to maximize the farmer’s profit. The
very clearly. The only drawback is that more the number of
farmer’s total profit = Sale – Expenditure. constraints, more will be the straight lines which makes the
Total sale of farmer, graph difficult to understand.
= ` [1×2,000x1+0.75×3,000x2+2×1,000x3] Characteristics of Graphical Method
= ` 2,000x1 + 2,250x2 + 2,000x3 The following are the characteristics of graphical
method of LPP,
Fertilizer expenditure,
1. Method is very simple and easy to understand.
= ` 0.50 (100x1 + 100x2 + 50x3)
2. Very sensitive analysis and can be illustrated very
= ` (50x1 + 50x2 + 25x3) easily by drawing graphs.
Labour expenditure, 3. Very easy to obtain optimal solution.
= ` 20.00 (5x1 + 6x2 + 5x3) 4. It consumes very less time.
= ` (100x1 + 120x2 + 100x3) Limitations of Graphical Method
Total expenditure is, The greatest drawback of this procedure is that it can
be used for LPP with not more than two decision variables,
` (150x1 + 170x2 + 125x3)
1. Limited to problems of two decision variables only.
The total profit, The reason is graphical method is 2D and only two
` [Sale – expenditure] dimensions (2 variable) can be displayed on the graph.
2. When the number of constraints and their type (≤, =, ≥) (ii) Draw lines parallel to this line. These are called Iso
are many and mixed respectively, identifying feasible profit or Iso cost lines.
solution region becomes complex. (iii) Choose the line which touches the feasible region at
3. The accuracy of the result is lesser when the solutions the farthest point for maximization and nearest point
are not integers i.e., it is difficult to read a value of for minimization case.
say variable x1 = 2.47. (iv) This point gives the optimal solution.
Procedure for Graphical Method (v) Compute Z using the coordinates of this optimal point.
The steps involved in generating extreme point Note
solution to an LPP in graphical method are as follows, The simplest approach is extreme point enumeration
(a) Formulate the problem in terms of a series of method.
mathematical constraints and an objective function. Problems on Extreme Point
(b) Plot Each of the Constraints Method
Consider each inequality constraint as equation. Give 1. Given the LP as
any arbitrary value (say zero) to one variable and get the
value of the other variable by solving the equation. Max 3A + 2B
Such that,
Similarly, give an arbitrary value to another variable
and find the corresponding value of the other variable. A+B $ 4
Plot these two sets of values. Connect these points by 3A + 4B # 24
a straight line. The procedure is repeated for every constraint A $ 2
of the problem. A–B # 0
(c) Identify the Feasible Region (Solution Space) A, B $ 0
Feasible region is the area which satisfies all the Solve the above problem graphically, mark the
constraints simultaneously. feasible region and explain the optimal solution.
To identify this region, adopt the following points, Solution : (Model Paper-I, Q4(b) | Aug./Sept.-16, Q4)
For > or ≥ type constraints, shade the region which Step 1
lies above the constraint lines.
Objective Function
For < or ≤ type constraints, shade the region which
lies below these lines. Max 3A +2B
The common shaded region is the feasible region. Such that,
The process of optimal solution determination proceeds A + B $ 4 ...eq(1)
as follows, 3A + 4B # 24 ... eq (2)
(d) Extreme Point Approach A $ 2 ... eq (3)
(i) Identify each of the corner or extreme points of the A – B # 0 ... eq (4)
feasible region either by visual inspection or by the A, B $ 0
method of simultaneous equations.
Step 2
(ii) Compute the profit or cost at each corner point by
substituting the point coordinates into the objective Converting inequality constraints into equality
function. constraints
(iii) The corner point which gives highest value of Z Equation I
for maximization case or lowest value of Z for By solving equation (1), we get the values of P and
minimization case is the optimal solution. Q points.
The coordinates of that optimal point are the values Case I
of the decision variables.
If A = 0, then by solving equation (1), we get,
(e) Iso Profit/Iso Cost Line Approach
A+B=4
(i) Assume zero profits or costs to draw the objective
function line. This is obtained by assuming the value =0+B=4
of one variable and from the assumed value getting B=4
the value of the other variable. Draw a line through
\ P = (0, 4)
this point and origin and extend.
Case II
If B = 0, then by solving equation (1), we get,
A+B=4
A+0=4
A=4
\ Q = (4, 0)
Equation II
By solving equation (2), we get the values of R and S points.
Case I
If A = 0, then by solving equation (2), we get
3A + 4B = 24
3(0) + 4B = 24
4B = 24
B = 24
4
B=6
\ R = (0, 6)
Case II
If B = 0, then by solving equation (2), we get,
3A + 4B = 24
3A + 4(0) = 24
3A = 24
A = 24
3
A=8
\ S = ^8, 0h
Equation III
By solving equation (3), we get,
A=2
\ T = (2, 0)
Equation IV
By solving equation (4), we get,
U Þ A – B = 0
(or)
\ J = (2, 0) –
Step 3
Plot the points P, Q, R, S, T and U on the graph to obtain a feasible solution.
P(0, 4), Q(4, 0), R(0, 6), S(8, 0), T(2, 0), U(0, 0)
Step - 4
The common feasible region is V, W and X
V = (2, 2)
W = (3, 4, 3.4)
X = (2, 4.5)
2.11 Problem
Step - 5
Substitute the coordinates of feasible region in objective function.
Max 3A + 2B
V (2,2) = 3(2) = 2(2) = 6 + 4 = 10
W (3.4. 3.4) = 3(3.4) + 2 (3.4) = 10.20 + 6.80 = 17
X (2, 4.5) = 3(2) + 2 (4.5) = 6 + 9 = 15
\ Maximum value of ` is obtainal at W (3.4, 3.4) = 17.
2. The manufacturer of patent medicines has proposed to prepare a production plan for medicines A and
B. There are sufficient ingredients available to make 20,000 bottles of medicine A and 40,000 bottles of
medicine B but there are only 45,000 bottles into which either of the medicines can be filled. Further, it
makes three hours to prepare enough material to fill 100 bottles of medicine A and one hour to prepare
enough material to fill 1000 bottles of medicine B and there are 66 hours available for this operation.
The profit is ` 8 bottle for medicine A and ` 7 per bottle for medicine B. Formulate this problem as a
LPP and solve it by graphical method.
Solution : Sept./Oct.-13, Q1(b)
Formulating LPP
Let us assume that the two types of medicines A and B are x1 and x2.
The sufficient ingredients available to make x1 is 20,000.
Thus, x1 ≤ 20,000.
The sufficient ingredients available to make x2 is 40,000.
Thus, x2 ≤ 40,000.
Total number of bottles that can be filled is 45000.
Thus, x1 + x2 ≤ 45000.
Number of hours needed to fill x1 is 3 and x2 is 1.
Availability of hours is 66.
Thus, 3x1 + x2 ≤ 66
Profit on x1 is 8 and on x2 is 7.
x1 = 100; x2 = 1000
Thus, z = 8 × 100 x1 + 7 × 1000 x2
= 800 x1 + 7000 x2
This can be summarized as,
Objective function, max z = 800 x1 + 7000 x2.
STC, 3x1 + x2 ≤ 66 ... (1)
x1 + x2 ≤ 45 (in 000’s) ... (2)
x1 ≤ 20 (in 000’s) ... (3)
x2 ≤ 40 (in 000’s) ... (4)
x1, x2 ≥ 0
Solving the LPP with Graphical Method
Considering equation (1),
3x1 + x2 = 66
Put x1 = 0; 3(0) + x2 = 66, x2 = 66
∴ (0, 66)
66
Put x2 = 0; 3x1 + 0 = 66; x1 = = 22
3
∴ (22, 0)
Considering equation (2),
x1 + x2 = 45
Put x1 = 0; 0 + x2 = 45, x2 = 45
∴ (0, 45)
Put x2 = 0, x1 + 0 = 45, x1 = 45
(45, 0)
x1 = 20 – Equation (3)
x2 = 40 – Equation (4)
Plotting points on a graph to obtain a feasible region.

2.13 Problem
The points obtained from the feasible region on graph are,
0 (0, 0)
A (0, 40)
B (5, 40)
C (12, 35)
D (20, 7)
E (20, 0)
Substituting the obtained points in objective function (Max z = 800 x1 + 7000 x2)
Points Co-ordinates Objective Function

0 (0, 0) z = 800 (0) + 7000(0) = 0
A (0, 40) z = 800(0) + 7000(40) = 280000
B (5, 40) z = 800(5) + 7000(40) = 284000*
C (12, 35) z = 800(12) + 7000(35) = 254600
D (20, 7) z = 800(20) + 7000(7) = 65000
E (20, 0) z = 800(20) + 7000(0) = 16000
Since, it is a profit function, maximum value is to be considered which is obtained at point B.
Max z = 284000
x1 = 5, x2 = 40
2.3 Simplex Method

Q10. Explain the simplex method and its standard form. What is the terminology used in simplex method
for solving LPP.
Answer :
Simplex Method
Simplex method was introduced by G. Dantzig. The simplex method is a suitable method to provide an algorithm
which have its basis related to ‘Fundamental theorem of linear programming’. It is a systematic algorithm which moves
from one basic solution to another basic solution. It is a systematic method through which improved objective function
can be obtained.
The general linear programming problem in the standard form is,
a1x1 + a2x2 ≤ b (b ≥ 0)
It is changed in the standard form to
a1x1 + a2x2 + S1 = b, where S1 ≥ 0
p1x1 + p2x2 ≥ q (q ≥ 0)
It is changed to p1x1 + p2x2 – S2 = q, where S2 ≥ 0.
The quantities S1 and S2 are slack variables and their values depends upon the values assumed by other in a
particular equation.
Characteristics of the Standard Form
(i) The objective function is of maximization type
(ii) All constraints are expressed as equations
(iii) Right hand side of each constraint is non-negative
(iv) All variables are non-negative.
Example 7. Product Mix
Express the following LPP in standard form, A column in the simplex table that contains all of the
variables in the solution.
Minimize Z = 5x1 + 7x2
Subject to the constraints 8. Basic
x1 + x2 ≤ 8 A set of variables which are not restricted or equal

to zero in the current basic solution are listed in the product
3x1 + 4x2 ≥ 3 mix column. These variables are called basic variables.
6x1 + 7x2 ≥ 5 9. Iteration
and x1, x2 ≥ 0. The sequence of steps performed in moving from one
Solution : basic feasible solution to another.
Since, Min Z = – Max (Z) = Max Z* 10. Zj Row
The given LPP becomes, The row containing the figures for gross profit or loss
Maximize Z* = – 5x1 – 7x2 – 0S1 + 0S2+ 0S3 given up by adding one unit of a variable into the solution.
Subject to constraints 11. Cj – Zj / Net Evaluation / Index Row
x1 + x2 + S1 = 8 The row containing the net profit or loss that will
result from introducing one unit of the variable indicated in
3x1 + 4x2 – S2 + A1 = 3
that column in the solution. Number in the index row are
6x1 + 7x2 – S3 + A2 = 5 also shadow prices or according prices.
[ A1, A2 = Artificial variables] 12. Pivot/Key Column
and x1, x2, S1, S2, S3 ≥ 0. The column with the largest positive number in the
Terminology of Simplex Method Cj – Zj row of a maximization problem (largest negative for
The following are the terminologies used in solving minimization problems). It indicates which variables will
LPP through simplex method, enter the solution next (i.e., entering variable/incoming
1. Standard Form variable).
A LPP in which all the constraints are written as 13. Pivot/Key Row
equalities. The row corresponding to the variable that will leave
2. Slack Variable the basis in order to make room for the entering variable.
A variable added to LHS of ≤ constraints The departing variable (outgoing variable) will correspond
(maximizations LPP) to convert the constraints into equality. to the smaller positive ratio found by dividing the quantity
The value of this variable is equal to the amount of unused column values by the key column values for each row.
resource.
14. Pivot/Key Element
3. Surplus Variable
The element of the intersection of key row and key
A variable subtracted from the LHS of ≥ constraints
column.
(minimization LPP) to convert the constraints into equality.
The value of this variable is equal to the amount over and
above the required minimum level.
2.3.1 Algorithm of Simplex
Method (Maximization Case in
4. Basic Solution
Simplex Method)
For a general LP with ‘n’ variables and ‘m’ constraints,
a basic solution can be found by setting (n – m) variables Q11. Write down the algorithm for simplex method
equal to zeros and solving the constraint equations for the for solving the LPP.
value of other m variables. If a unique solution exists, it is
a basic solution. Answer :
5. Basic Feasible Solution Following steps must be followed in order to solve
It refers to the set of constraints corresponding to the an LPP using simplex method,
extreme point of the feasible region. Step 1: Formulation of LPP
6. Simplex Table Let objective function be,
A table used to keep track of the calculations made at
each iteration when the simplex solution method is employed. Maximize Z = c1 x1 + c2 x2

2.15 Problem
Subject to constraints Step 3: Determine Initial Basic Feasible Solution
a1 x1 + a2 x2 ≤ b1 Initially, we assign zero to all the decision variables
and it results in obtaining the values of slack/surplus
a3 x1 + a4 x2 ≤ b2 variables.
So, if x1 = 0; x2 = 0 then
and x1, x2 ≥ 0
a1 x1 + a2 x2 + S1 ⇒ 0 + 0 + S1 = b1
Where, Cj = c1, c2 (j = 1, 2)

⇒ S1 = b1
Step 2: Converting Given LPP into Standard Form a x + a x + S ⇒ 0 + 0 + S2 = b2
3 1 4 2 2
1. An objective function must be converted into ⇒ S2 = b2

maximization problem by simply multiplying the Step 4
objective function with minus sign.
Set up the table as follows,
Maximize Z = Minimize (– Z) C C C 0 0
j 1 2
i.e., if Minimize Z = c1 x1 + c2 x2 is given then it can
be converted as, cCBB B X
xBB x1 x2 sS11 Ss22
Maximize Z = – c1 x1 – c2 x2
0 s1 b1 a1 a2 1 0
2. Convert the inequality constraints into equations by
adding or subtracting the variables depending on 0 sS22 b2 a3 a4 0 1
inequalities.
Maximization Zj 0 × a1 + 0 × a2+ 0×1 0×0

0 × a3 = 0 0 × a4 = 0 = 0 =0
v If a1 x1 + a2 x2 ≤ b1 is given then we add slack
Z j −Cj
C −C1 −C2 0−0 0−0
variable to LHS to convert it into an equation. j – Zj
⇒ a1 x1 + a2 x2 + S1 = b1 (S1 ≥ 0) Here, Cj – Coefficient of decision variable and basic

variable in objective
Minimization
Z = c1 x1 + c2 x2 + 0S1 + 0S2
v If a1 x1 + a2 x2 ≥ b1 is given then we subtract
{Cj ⇒ c1, c2, 0, 0}
surplus variable from LHS or add artificial
variables to RHS. XB − Right hand side values of constraints
(equation).
⇒ a1 x1 + a2 x2 – S2 = b1 B − Basic variables (here they are S1, S2).
⇒ a1 x1 + a2 x2 = b1 + A [(S2 ≥ 0) (A ≥ 0)] CB – Coefficient of basic variables in objective
function (i.e.,) coefficient of S1 and S2 are zeroes.
v If a1 x1 + a2 x2 = b1 is given then we add artificial a1, a2, a3 and a4 are coefficients of x1 and x2 in the
variable to RHS. given constraints (equation).
⇒ a1 x1 + a2 x2 = b1 + A (A ≥ 0) Step 5
(i) Determine Zj = Σ CB.Xj (j = 1, 2....)
So, for the example considered standard form
and also Zj = Cj
would be,
(ii) After finding C j – Z j, find the key column by
Maximize Z = c1 x1 + c2 x2 + 0S1 + 0S2 considering highest positive value.
Subject to constraint (iii) Find the minimum ratio of XB and Xi of key column
to obtain key row, the corresponding variable would
a1 x1 + a2 x2 + S1 = b1 be leaving.
Minimum ratio,

a3 x1 + a4 x2 + S2 = b2
XB
=
 x1 , x2 ≥ 0 Corresponding key column value in a row
 

s1 , s2 ≥ 0  (iv) The intersection of key row and key column would
result in key element or pivot element.
The table would be,
Cj C1 C2 0 0
xB
cCBB B xXBB x1 x2 Ss11 sS2
2
xi
0 S1 b1 a1 a2 1 0 b1 /a2
Key ⇐ Min
0 S2 b2 a3 element 0 1 b2 /a4 Ratio
a4
(Leaving)
Zj a1 × 0 a2 × 0 0×1 0×0
=0 =0 =0 =0
CZj j–− CZj j C01−–C10 C0−

2 –C20 0−0 0−0
= −=CC
1 1
==− CC2 2 =0 =0
⇑
Assume – C2 as most negative
( it is the key column)
centering variable
Step 6
Next iteration is performed by replacing the entering and leaving variable.
Values of rows and columns change after each iteration, for simplicity as follow,
old row value
1. New values of key row are ⇒
key element
2. New values of remaining rows are,
corresponding corresponding 
 
old row value – old value in × new value in 
key column key row in same column 

Step 7
After performing each iteration check for optimality,
i.e., once the Cj – Zj ≤ 0 and minimum ratio cannot be determined, then that would be the final iteration and it gives
out the final optimal solution.
Problems on Simplex Method (maximization case)
1. Maximize Z = 5x1 – 2x2 + 3x3,
Subject to,
2x1 + 2x2 – x3 > 2,
3x1 – 4x2 ≤ 3,
x2 + 3x3 ≤ 5
and x1, x2, x3 > 0.
Solution : (Model Paper-II, Q4(a) | Aug.-17, Q4(b))
Given that,
Maximize Z = 5x1 – 2x2 + 3x3
Subject to,
2x1 + 2x2 – x3 > 2
3x1 – 4x2 < 3
x2 + 3x3 < 5
x1, x2, x3 > 0

Standardization
Introducing surplus variables OS1, OS2, OS3 and artifical variable A1. LPP can be written as follows,
Max Z = 5x1 – 2x2 + 3x3 + OS1 + OS2 + OS3 – MA1
Subject to, 2x1 + 2x2 – x3 + S1 + A1 = 2
3x1 – 4x2 + S2 = 3
x2 + 3x3 + S3 = 5
and x1, x2, x3, S1, S2, S3, A1 > 0
Let x1 = 0, x2 = 0, S1 = 0
Thus, S2 = 3, S3 = 5, A1 = 2
Setting up Initial Simplex Table
Table-I
Cj 5 –2 3 0 0 0 –M Remarks
CB BV XB X1 X2 X3 S1 S2 S3 A1 (Min. Ratio)
2 ←
–M A1 2 2 2 –1 –1 0 0 1 =1
2
3
0 S2 3 3 –4 0 0 1 0 0 =1
3
5
0 S3 5 0 1 3 0 0 1 0 =0
0
Z = – 2M Zj – 2M – 2M M –M 0 0 –M
Cj – Zj 5 + 2M – 2 + 2M 3–M –M 0 0 0
Entering variable is x1
Droping variable is A1
New R1 = R1 (old) ¸ 2 (Key element)
2
= = 1 and vice versa
2
New R2 = old R2 – (3 × New R1)
R3 = old R3 – (0 × New R1)
Table-II
Cj 5 –2 3 0 0 0
Remarks (Min.Ratio)
CB BV XB X1 X2 X3 S1 S2 S3
–1 –1 –1 –1
5 X1 1 1 1 0 0 1÷ =
2 2 2 2
0 S2 0 0 –7 3 3 1 0 3 3
0÷ =
2 2 2 2
5
0 S3 5 0 1 3 0 0 1 5÷3=
3
–5 –5
Z=5 Zj 5 5 0 0
2 2
11 –5
Cj – Zj 0 –7 0 0
2 2

Leaving variable is S2
3
New R2 = R2 (old) ¸ (Key element)
2
–1
R1 = old R1 ¸ f × Rew R 2 p
2
R3 = old R3 ¸ (3 × Rew R2)
Table-III
Cj 5 –2 3 0 0 0
Remarks (Min. Ratio)
CB BV XB X1 X2 X3 S1 S2 S3
5 X1 1 1 –4 0 0 1 0 1 ÷ – 4 = – ve
3 3 3
– 14 2
3 X3 0 0 1 1 0 0 ÷ – 14 = – 14 – ve
3 3 3 3
0 S3 5 0 15 0 –3 –2 1 5 ÷ 15 = 1
3
– 62 11
Z=5 Zj 5 3 3 0
3 3
56 – 11
Cj – Zj 0 0 –3 0
3 3
Leaving variable is S2
New R3 = R3 (old) ¸15 (Key element)
–4
R1 = old R1 – f ×New R3 p
3
– 14
R2 = old R2 – f ×New R3 p
3
Table-IV
Cj 5 –2 3 0 0 0 Remarks
CB BV XB X1 X2 X3 S1 S2 S3 (Min. Ratio)
–4 13 –4
5 X1 13 1 0 0 7 4 ÷ = – ve
15 9 5
9 45 45
14 1 2 14 14 1
3 X3 0 0 1 ÷ = 70
9 15 45 45 19 15 3
1 –1 –2 1 1 –1
–2 X2 0 1 0 ÷ = – ve
3 5 15 15 3 5
101 – 11 53 56
Z= Zj 5 –2 3
9 15 45 45
11 – 53 – 56
Cj –Zj 0 0 0
15 45 45
Entering variable is S1
Leaving variable is X3
2.19 Problem
1
New R2 = R2 old ¸ (Key element)
15
–4
R1 = old R1 – f × New R 2 p
15
–1
R3 = old R3 – f × New R 2 p
5
Table-V
Cj 5 –2 3 0 0 0 Remarks
CB BV XB X1 X2 X3 S1 S2 S3 (Min. Ratio)
5 X1 23 1 0 4 0 1 4
3 3 3
70 2 14
0 S1 0 0 15 1
3 3 3
–2 X2 5 0 1 3 0 0 1
85 5 14
Z= Zj 5 –2 14 0
3 3 3
0 –5 – 14
Cj – Zj 0 0 – 11
3 3
Since all Cj – Zj < 0, the current basis feasible solution is optimal.
85 23
\ The optimal solution is Max Z = , x1 = , x2 = 5, x3 = 0.
3 3
Working Notes
1. Calculation of Zj = CB × jth Column of Matrix
For Table-I = – M × 2 + 0 × 3 + 0 × 0
= – 2M.
Calculate the remaining Zj values in the same way.
2. Calculation of Z value.
Z = CB × XB
For Table-I, Z = – M × 2 + 0 × 3 + 0 × 5
= – 2M.
2. Solve the following LP problem using Simplex method.
Maximize Z = 10X1 + 15X2 + 20X3
Subject to,
2X1+4X2+6X3 < 24
3X1+9X2+6X3 < 30
X1, X2, and X3 > 0
Solution : Feb.-17, Q5
Given that,
Maximize Z = 10x1 + 15x2 + 20x3

Subject to,
2x1 + 4x2 + 6x3 # 24
3x1 + 9x2 + 6x3 # 30
x1, x2 and x3 $ 0
Standardization Since ‘Cj – Zj’ value is highest for ‘x1’, thus enter
the basis for ‘x1’ and leaves through ‘s2’ as it has lowest
Maximize = 10x1 + 15x2 + 20x3 + 0s1 + 0s2

minimum ratio of ‘6’.
Subject to,
Interation II
2x1 + 4x2 + 6x3 + s1 = 24
Old R 2 Old R 2
3x1 + 9x2 + 6x3 + s2 = 30 New R2 = Key element i.e, 1

x1, x2, x3, s1, s2 $ 0

New R1 = Old R1 – 13 × New R2
Initial Solution
Simplex Table
Let, x1 = 0 x2 = 0, x3 = 0,
we get, Cj 10 15 20 0 0
Min ratio
s1 = 24, s2 = 30 CB Bv XB x1 x2 x3 s1 s2
Develop initial simplex table 20 x3 2 0 –1 1 1 –1
Basic variables : S1, S2 2 3
Calculation of Zj : SCB Xij 10 x1 6 1 5 0 –1 1
Index row : Cj – Zj zj 10 30 20 0 10
3
Cj 10 15 20 0 0 cj – zj 0 – 15 0 0 – 10
3
CB Bv XB x1 x2 x3 s1 s2 Min ratio
0 s1 24 2 4 6 1 0 24 = Since, all ‘Cj – Zj’ values are # 0 , thus the
6 4 optimality reached with,
0 s2 30 3 9 6 0 1 30 = x1 = 6, x2 = 0, x3 = 2
6 5
zj 0 0 0 0 0 Substituting the values of x1 and x2, x3, in the objective
cj – zj 10 function.
15 20 0 0
Max Z = 10X1 + 15 X2 + 20 X3
= 10 (6) + 15 (0) + 20 (2)
Since, Cj – Zj value is highest for ‘x3’, thus enter the
basis through x3 and leaves through ‘s1’ as it has lowest = 60 + 0 + 40
minimum ratio of ‘4’. = 100
Iteration - I
2.3.2 Minimization Case – Big-M
Old R1 Old R1
New R1 = Key element i.e., 6 Method
New R2 = Old R2 – Key element × New R1 i.e., Q12. Explain about the minimization case of LPP.
Old R2 – 6 × New R1 Write about the Big-M method used in simplex
problem.
Simplex Table
Answer :
Cj 10 15 20 0 0 Minimization Case of LPP
Min ratio
CB Bv XB x1 x2 x3 s1 s2 In simplex method, starting the simplex iterations at
20 x3 4 1 2 1 1 0 12 a basic feasible solution guarantees that all the succeeding
3 3 6 iterations will be feasible.
0 s2 6 1 5 0 –1 1 6
For the LPs in which all the constraints are of the (≤)
zj 20 40
40 20 10 0 type (with non-negative right hand sides), the slacks offer a
3 33 3
0 – 10 0 convenient starting basic feasible solution. A natural question
cj – zj 10 5
3 3 3 then arises.
How can we find a starting basic solution for models
that involve (=) and (> =) constraints?
The most common procedure for starting LPs that do not have convenient slacks is to use artificial variables.
These are variables that assume the role of slacks at the first iteration, only to be disposed off at a later iteration.
Two closely related methods are proposed for effecting this result.
1. The Big-M method (Penalty method) and
2. Two phase method.
Big-M Method (Penalty Method)
The Big-M method starts with the LP in the standard form. For any equation i that does not have a slack, we augment
an artificial variable Ri. Such a variable then becomes part of the starting basic solution.
However, because artificial are extraneous to the LP model, we assign them a penalty in the objective function to
force them to zero level at a later iteration of the simplex algorithm.
Computational Steps of Big-M Method
1. Express the LPP in the standard form.
2. Add non-negative artificial variables to the left side of each of the equations corresponding to constraints of the type
(≥) and ‘=’. When artificial variables are added, it causes violation of the corresponding constraints. This difficulty is
removed by introducing a condition which ensures that artificial variables will be zero in the final solution (provided
the solution of the problem exists). On the other hand, if the problem does not have a solution, at least one of the
artificial variables will appear in the final solution with positive value. This is achieved by assigning a very large
price (per unit penalty) to these variables in the objective function. Such large price will be designated by –M for
maximization problems (+M for minimization problems), where M > 0.
3. In the last, use the artificial variables for the starting solution and proceed with the usual simplex routine until the
optimal solution is obtained arrived.
Problems on Minimization Case (Big-M Method)
1. Solve the following LPP by simplex method,
Maximize Z = 3x1 + 2x2
Subject to 2x1 + x2 ≤ 2
3x1 + 4x2 ≥ 12 and x1, x2 ≥ 0
Solution :
By introducing the non-negative slack variable S1 and surplus variable S2, the standard form of the LPP becomes,
Maximize Z = 3x1 + 2x2 + 0S1 + 0S2
STC,
2x1 + x2 + S1 = 2
3x1 + 4x2 – S2 = 12
x1, x2, S1, S2 ≥ 0
But this will not yield a basic feasible solution. To get the basic feasible solution, add the artificial variable R1 to the
left hand side of the constraint equation which does not possess the slack variable and assign –M to the artificial variable
in the objective function.
The LPP becomes,
Maximize Z = 3x1 + 2x2 + 0S1 + 0S2 – MA1
STC,
2x1 + x2 + S1 = 2
3x1 + 4x2 – S2 + A1 = 12
x1, x2, S1, S2, A1 ≥ 0
The initial basic feasible solution is given by,
S1 = 2, R1 + 12 (basic) (x1 = x2 = S2 = 0, non-basic)
Note
While constructing simplex table in Big-M method, some constraints will be having both slack and artificial
variables, in such a situation, only artificial variable must be taken as a non basic variable. The same condition is applied
in 2nd constraint i.e., “3x1 + 4x2 + 0S1 – S2 + A1 = 12”.
Initial Iteration
Cj (3 2 0 0 – M) Min Ratio
CB BV XB x1 x2 S1 S2 A1
0 2 2 1 1 0 – 2
S1 =2
1
–M A1 12 3 4 0 –1 1 12
=3
4
Zj – 3M – 4M 0 M –M
Cj – Zj 3 + 3M 2 + 4M 0 –M 0
Since there are some Cj – Zj > 0. The current basic feasible solution is not optimal.
The non-basic variable x2 enters into the basic and the basic variable S1 leaves the basis.
First Iteration
New X2 value = 2, 2, 1, 1, 0
New A1
12 – 4(2) = 4
3 – 4(2) = – 5
4 – 4(1) = 0
0 – 4(1) = – 4
– 1 – 4(0) = – 1
1 – 4(0) = 1
Cj 3 2 0 0 –M
2 x2 2 2 1 1 0 –
– M A1 4 –5 0 –4 –1 1
Zj 4 + 5M 2 2 + 4M M –M
Cj – Zj – 5M – 1 0 – 4M – 2 – M 0
Since, all Cj – Zj ≤ 0 and an artificial variable A1 appears in the basic with non-zero level, the given LPP does not
posses any feasible solution. But the LPP possess a pseudo optimal solution.
2. Solve the following LPP
Max Z = 6x1 + 4x2
STC, 2x1 + 3x2 ≤ 30
3x1 + 2x2 ≤ 24
x1 + x2 ≤ 3 and
xi ≥ 0

2.23 Problem
Solution :
2.3.3 Two Phase Method
Standardization
Q13. What is a two phase method? Write the steps
Max Z = 6x1 + 4x2 + 0S1 + 0S2 + 0S3 involved in the two-phase simplex method to
STC solve a LPP.
2x1 + 3x2 + S1 = 30 Answer :

Two Phase Method
3x1 + 2x2 + S2 = 24
The two-phase method is another method to solve
x1 + x2 + S3 = 3 a given problem in which some artificial variables are
Where, x1, x2 ≥ 0 involved. The solution is obtained in two phases as follows,
Simplex Tableau-I Phase-I
In this phase, the simplex method is applied to a
Cj 6 4 0 0 0
Min ratio (XB/X1) specially constructed auxiliary linear programming problem
CB BV XB x1 x2 S1 S2 S3 leading to a final simplex table containing a basic feasible
0 S1 30 2 3 1 0 0 30/2 = 15 solution to the original problem.
0 S2 24 3 2 0 1 0 24/3 = 8 Steps of Phase-I
0 S3 3 1 1 0 0 1 The following are the steps involved in phase-I,
3/1 = 3
Step-1
Zj 0 0 0 0 0
Assign a cost-1 to each artificial variable and a cost-0
Cj – Zj 6 4 0 0 0 to all other variable (in place of their original cost) in the
objective functions. Thus, the new objective function is
Z* = a1 – a2 – a3...an.
Enter x1, drop S3
Where, an are the artificial variables.
x1 values = 3, 1, 1, 0, 0, 1 Step-2
New S1 New S2 Construct the auxiliary LPP in which the new
objective function Z* is to be maximized subject to the given
30 – (2 × 3) = 24 24 – (3 × 3) = 15
set of constraints.
2 – (2 × 1) = 0 3 – (3 × 1) = 0
Step-3
3 – (2 × 1) = 1 2 – (3 × 1) = –1
Solve the auxiliary LPP by simplex method until
1 – (2 × 0) = 1 0 – (3 × 0) = 0 either of the following three possibilities do arise,
0 – (2 × 0) = 0 1 – (3 × 0) = 1 (i) Maximum Z* = 0 and atleast one artificial variable
0 – (2 × 1) = –2 0 – (3 × 1) = –3 appears in the optimum basic at a nonzero level. In
this case the given LPP does not possess any feasible
Simplex Tableau-II
solution, stop the procedure.
Cj 6 4 0 0 0 (ii) Maximum Z* = 0 and no artificial variable appears in
the optimum basic. In this case proceed to phase-II.
CB BV XB x1 x2 S1 S2 S3 Min ratio
(iii) Maximum Z* = 0 and at least one artificial variable
0 S1 24 0 1 1 0 –2
is present in the basis at zero values then the feasible
0 S2 15 0 –1 0 1 –3 solution is also a feasible solution to the original
6 x1 3 1 1 0 0 1 LPP problem. Now to obtain basic feasible solution
proceed to phase-II or remove the artificial basic
Zj 6 6 0 0 6
variable and then proceed to phase-II.
Cj – Zj 0 –2 0 0 –6 Phase-II
Since, all the Cj – Zj values are either negative or 0’s, the Use the optimum basic feasible solution of phase-I
as a starting solution for the originals LPP. Assign the actual
problem has reached optimality.
costs to the variables in the objective function and a cost-0
x1 = 3; x2 = 0 to every artificial variable that appears in the basic at the
S1 = 24; S2 = 15, S3 = 0 zero level. Use simplex method to the modified simplex
table obtained at the end of phase-I, till an optimum basic
Max Z = 6x1 + 4x2 = 6 × 3 + 4 × 0 = 18 feasible solution (if any) is obtained.
Steps of Phase-II
Following are the steps involved in phase-II,
Step-1
In phase-1, the iterations are stopped as soon as the value of the new objective function becomes zero because this is
its maximum value. There is no need to continue till the optimality is reached if this value becomes zero earlier than that.
Step-2
The new objective function is always of maximization type regardless of whether the original problem is of
maximization or minimization type.
Step-3
Before starting phase-II remove all artificial variables from the table which were non-basic at the end of phase-I.
PRoblems on Two phase method
1. Use two-phase simplex method to solve,
Maximize Z = 5x1 + 8x2
Subject to the constraints,
3x1 + 2x2 ≥ 3
x1 + 4x2 ≥ 4
x1 + x2 ≤ 5
x1, x2 ≥ 0
Solution :
By introducing the non-negative slack, surplus and artificial variables, the standard form of the LPP becomes.
Maximize Z = 5x1 + 8x2 + 0S1 + 0S2 + 0S3
Subject to
3 x1 + 2x2 – S1 + A1 = 3
x1 + 4x2 – S2 + A2 = 4
x1 + x2 + S3 = 5
and x1, x2, S1, S2, S3, A1 ≥ 0
(Here: S1, S2 – surplus, S3 – slack, A1, A2 – artificial)
The initial basic feasible solution is given by,
A1 = 3, A2 = 4, S3 = 5 (basic) (x1 = x2 = S1 = S2 = 0, non-basic)
Phase-I
Assigning a cost-1 to the artificial variable and cost-0 to all other variables, the objective function of the auxiliary
LPP becomes,
Max Z* = –A1 – A2
2.25 Problem
Subject to the given constraints.
The iterative simplex table for the auxiliary LPP are,

Initial Iteration
Cj (0 0 0 0 0 –1 – 1) Min.Ratio
CB BV XB x1 x2 S1 S2 S3 A1 A2 (XB/X2)
–1 3 3 2 –1 0 0 1 0 3
A1 = 1.5
2
0 4
–1 A2 4 1 4 –1 0 0 1 =1→
4
0 S3 5 1 1 0 0 1 0 0 5
=5
1
Zj – 4 –6 1 1 0 –1 –1
Cj – Zj 4 6 –1 –1 0 0 0
↑
First Iteration
Introduce x2 and drop A2.
New R2 = old R2 ÷ 4
New R1 = old R1 – (2 × new R2)
New R3 = old R3 – (1 × new R2)
Cj (0 0 0 0 0 –1 –1) Min. Ratio

CB BV XB x1 x2 S1 S2 S3 A1 A2 (XB/X1)
5 1 −1 2
–1 A1 1 0 –1 0 1 →
2 2 5
2
0 1 1 1 0 −1 0 0 1 4
x2
4 4 4
3 1 −1 16
0 S3 4 0 0 1 0
4 4 4 3
Zj –5/2 0 1 –1/2 0 –1 1/2
5 0 –1 1 0 0 3
−
Cj – Zj 2 2 2
↑
Second Iteration
Introduce x1 and drop A1.
New R1 = old R1 ÷ 5/2
New R2 = old R2 – (1/4 × new R1)
New R3 = old R3 – (3/4 × new R1)

Cj (0 0 0 0 0 –1 –1)
CB BV XB x1 x2 S1 S2 S3 A1 A2
2 −2 1 2 −1
0 x1 1 0 0
5 5 5 5 5
9 1 −3 −1 3
0 x2 0 1 0
10 10 10 10 10
37 3 1 −3 −1
0 S3 0 0 1
10 10 10 10 10
Zj 0 0 0 0 0 0 0
Cj – Zj 0 0 0 0 0 –1 –1
Since, all Cj – Zj ≤ 0, the current basic feasible solution is an optimum solution. Further more, no artificial variable
appears in the optimum basic so we proceed to phase-II.
Phase-II
Here, we consider actual costs associated with the original variables. The new objective function then becomes.
Max Z = 5x1 + 8x2 + 0S1 + 0S2 + 0S3
The initial basic feasible solution for this phase is the one obtained at the end of phase-I.
The iterative simplex tables for this phase are,
Initial Iteration
Cj (5 8 0 0 0) Min. Ratio
CB YB XB x1 x2 S1 S2 S3 (XB/S2)
2 −2 1
5 x1 1 0 0 2→
5 5 5
9 1 −3
8 x2 0 1 0 ––
10 10 10
37 3 1
0 S3 0 0 1 37
10 10 10
Zj 5 8 –6/5 –7/5 0
6 7
Cj – Zj 0 0 0
5 ↑5
Since, there are some Cj – Zj > 0, the current basic feasible solution is not optimal.
Enter S2 and drop x1

New R1 = old R1 ÷ 1/5
New R2 = old R1 + (3/10 × new R1)
New R3 = old R3 – (1/10 × new R1)

2.27 Problem
Cj (5 8 0 0 0) Min ratio
(XB/S2)
CB BV XB x1 x2 S1 S2 S3
2
0 S2 2 5 0 –2 1 0 = –1
−2
3 3 −1
8 x2 1 0 0 –3
2 2 2
7 −1 1
0 S3 0 0 1 7→
2 2 2
Zj 12 8 –4 0 0
Cj – Zj –7 0 4 0 0
↑
Enter S1 and drop S3

New R3 = old R3 ÷ 1/2
New R1 = old R1 + (2 × new R3)
New R2 = old R2 + (1/2 × new R3)
Cj (5 8 0 0 0)
CB BV XB x1 x2 S1 S2 S3
0 S2 16 3 0 0 1 4
8 x2 5 1 1 0 0 1
0 S1 7 –1 0 1 0 2
Zj 8 8 0 0 8
Cj – Zj –3 0 0 0 –8
Since, all Cj – Zj ≤ 0, the current basic feasible solution is optimal.

∴ The optimal solution is Max Z = 40, x1 = 0, x2 = 5.
2.3.4 Special Cases in Simplex: Unbounded Solution, Unrestricted Variable,

Multiple Optimal Solution and Degeneracy
Q14. What is an unbounded solution in simplex method?
Answer :
Unbounded solution indicates linear programming problem that do not have finite solutions.
For example if in a maximization problem the solution variable can be made infinitely large without violating a
constraint it indicates unbounded solution. In simplex method, this situation can be recognized at the time of deciding the
leaving variables.
If at any iteration, the minimum ratio columns consists of either negatives or infinities, then no variable can be
choosen to leave the basis. Such a situation is called unbounded solution to LPP.
The reasons for not choosing negatives or infinities in min ratio column is as follows,
(a) If a negative value occurs, it indicates the entry of a non-basic variable in the basis with a negative value (an unfeasible
solution will occur).
(b) An infinity occurs when the denominator is zero, and this indicates that the entering variable could be increased
indefinitely with any of the current basic variables being removed from the basis.
Note
If min ratios are either negative or infinite, then the LPP has unbounded solution.
Problem on Unbound Solution

1. Solve the following LP problem using simplex method,
Minimize Z = 3x1 + 5x2 + 4x3
Subject to constraints, 2x1 + 3x2 ≤ 8
2x1 + 5x3 ≤ 10
3x1 + 2x2 + 4x3 ≤ 15
Where x1, x2, x3 ≥ 0
Solution :
The given problem is a minimization case, so convert it into a maximization case in the following manner.
Maximize (– Z) = – 3x1 – 5x2 – 4x3
By introducing non-negative slack variables S1, S2 and S3 the standard form of given LPP becomes,
Maximize Z = – 3x1 – 5x2 – 4x3 + OS1 + OS2 + OS3
Subject to constraints: 2x1 + 3x2 + S1 = 8
2x1 + 5x3 + S2 = 10
3x1 + 2x2 + 4x3 + S3 = 15
x1, x2, x3, S1, S2, S3 ≥ 0
Therefore, the initial feasible solution is given by S1 = 8, S2 = 10, S3 = 15
Iteration 1
Z
C Z
Since all Cj – Zj ≤ 0, the problem cannot be solved further. All values in the CB column are ‘zero’. Hence, feasible
solution to the given original LP problem does not exist.
Q15. What are unrestricted variables? Give example.
Answer :
In LPP along with the functional constraints, a non-negative constraint is given which states that the decision variable
can take only zero or positive values.
The simplex procedure was evolved keeping in mind this constraint. But, in many practical situations, one or more
of the variables may be unrestricted i.e., the variable may take either positive, negative or zero value.
In order to convert an LP problem involving unrestricted variables into an equivalent problem having only restricted
variables, the following operation is carried out,
2.29 Problem
Each unrestricted variable is expressed as the difference of two non-negative variables.
i.e., Let xr – Unrestricted variable
Then xr can be represented as a difference of two non-negative variables.
' "
i.e., xr = xr − xr
Where xr' ≥ 0 ; xr" ≥ 0

Example
x2 – Unrestricted variable
' "
x2 = x2 − x2
Now the unrestricted variable is replaced as above by the two new variables in the objective function as well as in
the constraints of the LP problem.
Variables x2' and x2" cannot appear simultaneously in the basis and one of the following may happen in optimal
solution,
(a) xr' = 0 ⇒ xr = − xr"
(b) xr'' = 0 ⇒ xr = xr'
(c) xr' = xr" = 0 ⇒ xr = 0

Problem on Unrestricted Variables
1. Solve the following linear programming problem,

Maximize z = 8x1 – 4x2
Subject to, 4x1 + 5x2 ≤ 20
– x1 + 3x2 ≥ – 23
x1 ≥ 0 ; x2 unrestricted
Solution : (Model Paper-II, Q4(b) | Sept.-14, Q4)
Standard Form of LPP

x2 is unrestricted
∴ x2 = x2′ − x2′′
Max z = 8x1 – 4 x2′ + 4 x2′′+ 0 S1 + 0 S2
4 x1 + 5 x2′ + 5 x2′′+ S1 = 20
STC,
– x1 + 3x2′ + 3x2′′+ S2 = −23

and
′ ′′
x2 , x2 ≥ 0
The initial basic feasible solution is,
s1 = 20, s2 = – 23 (basic) [ x2′ = x2′′ = 0(Non - basic)]

Initial Iteration
Cj 8 –4 4 0 0
CB BV XB x1 x '2 x '2' S1 S2 θ
0 S1 20 4 5 5 1 0 5
0 S2 – 23 – 1 3 3 0 1 – 23
Zj 0 0 0 0 0
Cj – Zj
8 –4 4 0 0
Since there are some (Cj – Zj) > 0, the current basic feasible solution is not optimal.
First Iteration
Introduction X1 and drop S1
Cj 8 –4 4 0 0
CB BV XB X1 X ′2 X ′′2 S1 S2
5 5 1
8 X1 5 1 0
4 4 4
17 17 1
0 S2 – 18 0 1
4 4 4
Zj 8 10 10 2 0
Cj – Zj
0 – 14 –6 –2 0

Since all (Cj – Zj) ≤ 0, the current solution is optimal.
X1 = 5
X2 = x2′ − x2′′ = 0 – 0 = 0
∴ Max z = 8X1 – 4X2
= 8 × 5 – 4 × 0
= 40 – 0
= 40
Q16. Explain the situation in which an infeasible solution occur in simplex method.
Answer :
Infeasible solution is a situation that occurs in LPP when the problem has no feasible solution that satisfies all the
constraints including the non-negativity constraint. In simplex method, the infeasibility condition can be recognized by
the presence of artificial variables i.e., if atleast one of the artificial variables appears with a positive value at the optimal
simplex table then no feasible solution exists.
This is so, because the presence of artificial variable will drastically alter the objective function value as its contribution
is very high (i.e., M). The solution obtained with artificial variables in the basis will, however satisfy all the constraints.
But, it is not a feasible solution. It is known as pseudo-optimal solution, as it satisfies the constraints but not objective
function.
Note
If artificial variable with positive value exists in the final (optimal) simplex table, then the LPP is said to have
infeasible solution.
2.31 Problem
Problem on Infeasible Solution

1. Solve using simplex method,
Max Z = 3x1 + 2x2

Subject to,
2x1 + x2 ≤ 2
3x1 + 4x2 ≥ 12
and x1, x2 ≥ 0
Solution :
Step 1: Formulation of LPP
Step 2: Standard Form
Max Z = 3x1 + 2x2 + 0S1 + 0S2 – MA1
STC 2x1 + x2 + S1 = 2
3x1 + 4x2 – S2 + A1 = 12
And x1, x2, S1, S2, A1 ≥ 0
Step 3: IBFS
Let x1 = 0, x2 = 0, S2 = 0. Thus, S1 = 2, A1 = 12
Step 4: Setting up Initial Simplex Tableau
Tableau-I
Cj 3 2 0 0 –M Min
Formulae Remarks
CB BV XB x1 x2 S1 S2 A1 Ratio
0 S1 2 2 1 1 0 0 2 Leaving
–M A1 12 3 4 0 –1 1 3 Variable
Zj – 3M – 4M 0 M –M
Cj – Zj 3 + 3M 2 + 4M 0 –M 0
↑
Tableau-II
Cj 3 2 0 0 –M
Min Ratio Formulae Remarks
R1
2 x2 2 2 1 1 0 0 R1′ = No entering
1 variable could
–M A1 4 –5 0 –4 –1 1 R ′2 = R 2 – (4 × R1′ ) be identified
Zj 4 + 5M 2 2 + 4M M –M
Cj – Zj – 5M – 1 0 – 4M– 2 –M 0
Since, all Cj – Zj ≤ 0, the iteration stops. The solution is optimal. However, this solution is not a feasible solution
because the artificial variable A1 is still present in the basis of the optimal simplex table.
Also the optimal solution x1 = 0 and x2 = 2 violates the second constraint (3x1 + 4x2 ≥ 12). This also indicates that
the LPP has infeasible solution.
Q17. Describe the simplex problem in which multiple solution occur.
Answer :
An LPP can be solved by simplex method or by graphical method. While solving by simplex method in each iteration
index row values (Cj – Zj) are calculated.
In the last iteration i.e., optimal simplex table, all the Cj – Zj values ≤ 0 for maximization case (≥ 0 for minimization case).

At this situation if one or more non-basic variable columns have Cj – Zj = 0 then the LPP is said to have multiple
optimal solutions. This is because the index row indicates the contribution per unit of a particular variable if it is entered
into the basis.
Thus, when a non-basic variable corresponding to which Cj – Zj = 0 is entered into the basis, a new solution will be
arrived at, but the value of the objective function will not change.
This is an alternative optimal solution. So, such an LPP will have multiple optimal solutions.
Note
If Cj – Zj = 0 for non-basic variable, LPP has multiple optimal solutions.
Problem on Multiple Optimal Solutions
1. Use penalty method to solve the LPP.
Max Z = 6x1 + 4x2
Subject to 2x1 + 3x2 30
3x1 + 2x2 24
x1 + x2 3
and x1, x2 0.
Solution :
Step 1
Formulation of LPP (Given).
Step 2: Standard Form
Max Z = 6x1 + 4x2 + 0S1 + 0S2 + 0S3 – MA1
Subject to 2x1 + 3x2 + S1 = 30
3x1 + 2x2 + S2 = 24
x1 + x2 – S3 + A1 = 3
x1, x2, S1, S2, S3, A1 0
Step 3
IBFS
Let, x1 = 0, x2 = 0, S3 = 0
Thus, S1 = 30, S2 = 24, A1 = 3
Step 4
Setting up initial simplex tableau.
Tableau-I
Cj
6 4 0 0 0 –M Min. Remarks
CB BV XB
x1 x2 S1 S2 S3 A1 Ratio
0 S1 30 2 3 1 0 0 – 15
0 S2 24 3 2 0 1 0 – 8
–M A1 3 1 1 0 0 –1 1 3 Leaving
variable
Z = –3M Zj –M –M 0 0 M –M
Cj – Zj 6 + M 4 + M 0 0 –M 0
Entering variable

2.33 Problem
Tableau-II
Cj
6 4 0 0 0 Mi n. Formul ae Remarks
CB BV XB
x1 x2 S1 S2 S3 Ratio
0 S1 24 0 1 1 0 2 12 R1* = R1 − ( 2 × R3* )
R2 = R2 − ( 3 × R3 )
* *
0 S2 15 0 –1 0 1 3 5→ Leaving
R
R3 =
* 3
6 x1 3 1 1 0 0 –1 – variable
1
Z = 18 Zj 6 6 0 0 –6
C j – Zj 0 –2 0 0 6
[Note: ‘*’ denotes new]
Tableau-III
Cj
6 4 0 0 0 Remarks Formuale
CB B XB x1 x2 S1 S2 S3
0 S1 14 0 5/3 1 –2/3 0 No entering R1 = R1 − (2 × R2 )
* *
0 S3 5 0 –1/3 0 1/3 1 variable is R2 = R1 / 3

*
R3 = R3 + (1 × R2 )
* *
6 x1 8 1 2/3 0 1/3 0 identified
Z = 48 Zj 6 4 0 2 0
Cj – Zj
0 0 0 –2 0
Since, all Cj – Zj 0, the iteration stops.

Optimal Solution
x1 = 8, x2 = 0
Max Z = 6 × 8 + 4 × 0 = 48
Alternative Solution
Presence of Cj – Zj = 0 for non basic variable indicates that the LPP has alternative solutions.
In the final simplex table Cj – Zj row is observed, it is found that for the non basic variable x2, the value of Cj – Zj =
0. Hence, by making this variable enter the basis, alternative optimal solution can be obtained as follows,
Tableau–IV
Cj
6 4 0 0 0 Min. Remarks
CB BV xB
x1 x2 S1 S2 S3 Ratio
0 S1 14 0 5/3 1 –2/3 0 42/5 Leaving
variable
0 S3 5 0 –1/3 0 1/3 1 –
6 x1 8 1 2/3 0 1/3 0 12
Z = 48 Zj 6 4 0 2 0
Cj – Z j
0 0 0 –2 0
Tableau-V
R1* = R1 Old × 3 / 5
R2* = R2 Old + 1 / 3 × R1*
R3* = R3 Old – 2 / 3 × R1*

Cj
6 4 0 0 0 Remarks
CB BV xB x1 x2 S1 S2 S3
4 x2 42/5 0 1 3/5 –2/5 0 No entering
0 S3 39/5 0 0 1/5 3/15 1 variable
6 x1 12/5 1 0 –2/5 9/15 0 could be
identified
Z = 48 Zj 6 4 0 2 0
C j – Zj 0 0 0 –2 0
Since all Cj – Zj 0, the iteration stops.

Optimal Solution
12 42
x1 = , x2 =
5 5
6x1 + 4x2
12 42
Max Z = 6 × +4×
5 5
= 48
This Z value is same as one obtained earlier.
Q18. Explain briefly the concept of degeneracy.
Answer :
Degeneracy
The concept of obtaining a degenerate basic feasible solution in a LPP is known as degeneracy. Degeneracy in a
LPP may arise,
(i) At the initial stage when atleast one basic variable is zero in the initial basic feasible solution.
(ii) At any subsequent iteration when more than one basic variable is eligible to leave the basis and hence one or more
variables becoming zero in the next iteration and the problem is said to be degenerate. There is no assurance that
the value of the objective function will improve, since the new solutions may remain degenerate. As a result it is
possible to repeat the same sequence of simplex iterations endlessly without improving the solutions. This concept
is known as cycling or circling.
Perturbation Rule to Avoid Cycling
(a) Divide each element in the tied rows by the positive coefficients of the key column in that row.
(b) Compare the resulting ratios, column by column, first in the identity and then in the body, from left to right.
(c) The row which first contains the smallest algebraic ratio contains the leaving variable.
Problems
1. Solve the following LPP,
Max Z = 3x1 + 9x2
Subject to constraints
x1 + 4x2 ≤ 8
x1 + 2x2 ≤ 4
and x1, x2 ≥ 0
Solution :
Max. Z = 3x1 + 9x2
Subject to constraints,
x1 + 4x2 + 1S1 + 0S2 = 8
x1 + 2x2 + 0S1 + 1S2 = 4
2.35 Problem
Initial Iteration
Cj 3 9 0 0
CB BV XB x4 x2 S1 S2 Min ratio
0 S1 8 1 4 1 0 8/4=2
0 S2 4 1 2 0 1 4/2=2 ←
Zj 0 0 0 0
Cj – Zj 3↑
3 9↑ 0 0
Since, there is a tie in the minimum ratio column it is an indication of degeneracy. To resolve degeneracy we select
the least value from the entering column that is from (4 and 2). As the least value is 2 we select the row 2 as the key row.
Working Notes
New x2
4/2 = 2, 1/2, 2/2 = 1, 0/2 = 0, 1/2
New S1
8 – (4 × 2) = 0
1
1 – (4 × )=–1
2
4 – (4 × 1) = 0
1 – (4 × 0) = 1
1
0 – (4 × )=–2
2
First Iteration
Cj 3 9 0 0
CB BV XB x1 x2 S1 S2
0 S1 0 –1 0 1 –2
9 x2 2 1/2 1 0 1/2
Zj 9/2 9 0 9/2
Cj – Zj –3/2 0 0 –9/2
Since, all ∆j ≤ 0, an optimal solution is reached, x1 = 0, x2 = 2, Max Z = 18

Max Z = 3(0) + 9(2)
= 18
2.4 Converting Primal LPP to Dual LPP

Q19. Explain duality principle. What is primal dual relationship. Write the rules for converting primal to dual.
Answer :
Duality Principle
For every linear programming problem there is a unique linear programming problem associated with it, involving
the same data and closely related optimal solutions. The original (given) problem is called “the primal problem” while the
other is called its “dual”. But, in general, the two problems are said to be duals to each other.

The importance of the duality concept is due to two (b) The RHS values of dual is taken from the contributions
main reasons. Firstly, if the primal contains a large number in objective function of primal LPP.
of constraints and a smaller number of variables, the work of
a computations can be considerably reduced by converting (c) The equation type of constraints is decided as follows,
into the dual problem and then solving it. Secondly, the
interpretation of the dual variables from the cost or economic
Primal Dual
Objective Primal variable constraints
point of view proves extremely useful in making future
decisions in the activities being programmed.
Max Z (i) ≥ ≥
Primal-Dual Relationship
(ii) unrestricted =
If Primal Then Dual
Min Z (i) ≥ 0 ≤
1. Objective : Maximize Objective : Minimize
(ii) unrestricted =
2. Variable xj Constraint, j
3. Constraint, i Variable yi Step 5
4. Constraint i: ‘=’ type Constraint j: ‘=’ type
The status of the variables are decided as follows,
5. Variable xj: unrestricted Variable yi: unrestricted
6. ≤ type constraints ≥ type constraints
Primal Dual
Objective Primal variable constraints
7. ≥ 0 variables ≥ inequality constraints
8. ≤ inequality constraints ≥ 0 variables Max Z (i) ≤ ≥ 0
Rules for Converting Primal to Dual (ii) = unrestricted
Following rules/steps are followed for converting Min Z (i) ≥ 0 ≥ 0
primal to dual,
(ii) = unrestricted
Step 1
Check the following condition, Step 6: Testing of Optimality
For Max Z, all constraints are “≤” type.
For Min Z, all constraints are “≥” type. In case of maximization, Cj – Zj values are either negative
If the condition is not satisfied convert the constraints or zero then the given solution is said to be an optimal one.
sign by multiplying by ‘–1’. Obtaining Primal Solution from Dual Solution
If any of the constraint is equal to (‘=’) type, ignore
at this step. This will be considered later. The steps of obtaining an optimum solution to one of
Step 2 the problems from an optimum solution to the dual may be
Decide on the decision variables of the dual LPP. The given as follows,
number of decision variables should be equal to the number
of constraints of the primal LPP. (a) First locate the starting solution primal variables.
Step 3 These variables are corresponding to the primal basic
(a) If the objective function is maximization convert it variables in the optimum solution.
to minimization and vice versa.
(b) Read the net evaluations or index row corresponding
(b) The contributions of variables to the objective
to the above located starting variables in the optimum
function of the dual will be the RHS (Right Hand
dual simplex table. The optimal solution to the primal
Side) values of the constraints of the primal.
is given by the optimal dual index row directly. If the
Step 4
dual is a maximization problem, then the positive of
(a) The values of the coefficients of variables in the
constraints for the dual LPP is obtained from that of the Cj – Zj values will give the optimal values.
primal LPP but transposed.
Primal (Max) Dual (Min)
Example
Primal Dual Basic variable values (C j – Z j ) under starting
1 2  solution columns.
 3 5  → 1 3 6  (Cj + Zj) under starting Basic variable values solution
  2 5 7 
6 7    columns.

2.37 Problem
Problems on Primal and dual LPP

1. Apply the principle of duality to solve the LP problem,
Max. z = 3x1 – 2x2, subject to: x1 + x2 ≤ 5, x1 ≤ 4, 1 ≤ x2 ≤ 6 and x1, x2 ≥ 0.
Solution : (Aug./Sept.-15, Q4 | March-15, Q4)
Given that,
Max. z = 3x1 – 2x2
Subject to,
x1 + x2 ≤ 5
x1 ≤ 4
1 ≤ x2 ≤ 6
and x1, x2 ≥ 0
The given problem can be reformed as,
Max. z = 3x1 – 2x2
Subject to:
x1 + x1 ≤ 5
x1 + 0x1 ≤ 4
0x1 + x2 ≤ 6
0x1 + x2 ≤ – 1
and x1, x2 ≥ 0
This reformed problem is already in a standard primal form, therefore we can now apply the principle of duality to
form dual LPP.
Step 1: Dual Formulation
Let, w1, w2, w3 and w4 be the decision variables of dual LPP.
Objective Function
Min. z = 5w1 + 4w2 + 6w3 – w4
Subject to,
w1 + w2 + 0w3 + 0w4 ≥ 3
w1 + 0w2 + 1w3 – 1w4 ≥ – 2
and w1, w2, w3, w4 ≥ 0
Step 2
Changing the objective function into maximization form and introducing surplus variable S1 ≥ 0 and slack variable S2 ≤ 0
Max. Zw = – 5w1 – 4w2 – 6w3 + w4 + 0S1 + 0S2
Subject to,
w1 + w2 + 0w3 + 0w4 – S1 + 0S2 = 3
–w1 + 0w2 – 1w3 + 1w4 + 0S1 + S2 = 2
w1, w2 , w3 , w4 , S1 , S2 ≥ 0
Note: Introducing artificial varaible is not requried as W2 can be considered as artificial variable
Step 3
Cj –5 –4 –6 1 0 0 Minimum Ratio
CB BV XB w1 w2 w3 w4 S1 S2 (XB/Key column)
–4 W2 3 1 1 0 0 –1 0
1 W4 2 –1 0 –1 1 0 1
Zj –5 –4 –1 1 4 1
Cj – Zj 0 0 –5 0 –4 –1

Since, all Cj – Zj ≤ 0, an optimum solution is obtained for the dual problem. The optimum solution is,
w1 = 0, w2 = 3, w3 = 0 and w4 = 2,
Min. ZW = 5w1 + 4w2 + 6w3 – w4
= 5(0) + 4(3) + 6(0) – 2(1)
= 10
The primal variables X1 and X2 correspond to slack and surplus variables S1 and S2 of the dual problem. In the above
table, the Cj – Zj values of S1 and S2 are – 4 and – 1 respectively.
Therefore, optimum solution to the original primal will be X1 = – 4, X2 = – 1, Max. Z = 10.
2. Use duality to solve the following L.P.P.
Maximize z = 2x1 + x2 subject to the constraints,
x1 + 2x2 ≤ 10, x1 + x2 ≤ 6, x1–x2 ≤ 2, x1 – 2x2 ≤ 1; x1,x2 ≥ 0
Solution : (Model Paper-III, Q5(a) | Feb./March-16, Q4)
Given that,
Maximize z = 2x1 + x2
Subject to constraints: x1 + 2x2 ≤ 10
x1 + x2 ≤ 6
x1 – x2 ≤ 2
x1 – 2x2 ≤ 1
x1, x2 ≥ 0
Dual of LPP,
Let w1, w2, w3, w4 be the decision variables of dual LPP.
Primal: Max. Z
Dual: Min. Z*
The objective function is,
Min. z* = 10w1 + 6w2 + 2w3 + w4
Primal
 x1 x2  Dual
1 2
   1 w2 w3 w4 
w
1 1 1 1 1 1 
  
 1 –1  2 1 –1 –2 
 1 –2

Subject to,
w1 + w2 + w3 + w4 ≥ 2
2w1 + w2 – w3 – 2w4 ≥ 1
⇒ Introducing surplus (S1, S2) and artificial variables in the dual LPP, as follows,
Min. z* = 10w1 + 6w2 + 2w3 + w4 + OS1 + OS2 + MA2 + MA3
Subject to: w1 + w2 + w3 + w4 – S1 + MA1 = 2
2w1 + w2 – w3 – 2w4 – S2 + MA2 = 1
w1, w2, w3, w4, S1, S2, A1, A2 ≥ 0.
⇒ Setting variables w1, w2, w3, w4, S1, S2 all equals to zero, then the initial feasible solution is A1 = 2, A2 = 1.
2.39 Problem
Iteration: 1
Cj 10 6 2 1 0 0 M M Minimum
ratio = XB/
CB BV XB w1 w2 w3 w4 S1 S2 A1 A2 key column
M A1 2 1 1 1 1 –1 0 1 0 2/1 = 2
M A2 1 2 1 –1 –2 0 –1 0 1 1/2 = 0.5 →
Zj 3M 2M 0 –M –M –M M M
Cj – Zj 10 – 3M 6 – 2M 2 M M M 0 0
↑
w1 is an entering variable
A2 is a leaving variable, key element is ‘2’.
Modified Values
Key row is,
Old row values
New row values =
Key element

R2 = R2 (old) ÷ 2
(New)
XB w1 w2 w3 w4 S1 S2 A1 A2
1 1 –1 –1 1
1 –1 0 0
2 2 2 2 2

Values of other rows = Old values – [Corresponding old value in key column × Corresponding new value in key
row in same column]
XB w1 w2 w3 w4 S1 S2 A1 A2
R1 new, 3 0
1 3
2 –1
1
1
–1
2 2 2 2 2
Iteration: 2
A1 is a leaving variable
w4 is an entering variable, key element is ‘2’.
Modified Values
Old row values
Key row values =
Key element

 3
 
2
R1 new =
2

XB w1 w2 w3 w4 S1 S2 A1 A2
R1 new, 3 1 3 –1 1 1 –1
0 1
4 4 4 2 4 2 4
Other row values = Old values – [Corresponding old value in key column × Corresponding new value in key row
in same column]
 1  3
R2 new =   –  –1 × 
 2  4
XB w1 w2 w3 w4 S1 S2 A1 A2
5 3 1 –1 –1 1 1
1 0
4 4 4 2 4 2 4

Iteration: 3
Cj 10 6 2 1 0 0 M M Min ratio
CB BV XB w1 w2 w3 w4 S1 S2 A1 A2 (XB/w2)
3/ 4
1 w4 3/4 0 1/4 3/4 1 –1/2 1/4 1/2 –1/4 =3
1/ 4
10 w1 5/4 1 3/4 1/4 0 –1/2 –1/4 1/2 1/4 5/ 4 5

= or 1.67 →
3/ 4 3
Zj 10 1
Cj – Zj 0 0
↑
w1 is a leaving variable
3
w2 is an entering variable, key element is .
4
Modified Values
 5
 
4
Key row values = R2 =
3
4
XB w1 w2 w3 w4 S1 S2 A1 A2
R2 new, 5 4 1 –2 –1 2 1
1 0
3 3 3 3 3 3 3
Values of other row,
 3  1 5
R1 new =   –  × 
4 4 3
XB w1 w 2 w3 w4 S1 S2 A1 A2
1 –1 2 –1 1 1 –1
0 1
3 3 3 3 3 3 3

2.41 Problem
Iteration: 4
Cj 10 6 2 1 0 0 M M Min ratio
CB BV XB w1 w2 w3 w4 S1 S2 A1 A2 (XB/w3)
5/3
6 w2 5/3 4/3 1 1/3 0 –2/3 –1/3 2/3 1/3 =5
1/ 3
1 w4 1/3 –1/3 0 2/3 1 –1/3 1/3 1/3 –1/3 1/ 3 1

= or 0.5 →
2/3 2
Zj 6 1
Cj – Zj 0 0
↑
w4 is a leaving variable
2
w3 is an entering variable, key element is .
3
Modified Values
 1
  1
3
Key row values = R2 = =
 2 2
 
3
XB w1 w 2 w3 w4 S1 S2 A1 A2
R2 new, 1 –1 3 –1 1 1 –1
0 1
2 2 2 2 2 2 2
Values of other row,
 5  1 1
R1 new =   –  × 
3 3 2
XB w1 w2 w3 w4 S1 S2 A1 A2
3 3 –1 –1 –1 1 1
1 0
2 2 2 2 2 2 2

Iteration: 5
Cj 10 6 2 1 0 0 M M
Min ratio
CB BV XB w1 w2 w3 w4 S1 S2 A1 A2
2 w3 1/2 –1/2 0 1 3/2 –1/2 1/2 1/2 –1/2
6 w2 3/2 3/2 1 0 –1/2 –1/2 –1/2 1/2 1/2
Zj 10 8 6 2 0 –4 –2 4 2
Cj – Zj 2 0 0 1 4 2 4M 2M
Since, all Cj – Zj ≥ 0, the current solution is optimal. The optimum solution of the dual is,
3 1
w1 = 0, w2 = , w3 = , w4 = 0 and minimum of w = 10
2 2
  1  3 
 i.e.,  2 ×  +  6 ×  = 10
2 2
 
∴ The optimal solution of the primal is,
x1 = 4, x2 = 2 with maximum of z = 10.
2.5 Advantages and Limitations 2.6 Transportation Problem

of LPP – Definition, Assumptions
Q20. What are the advantages and limitations of
and Mathematical model of
LPP? Model Paper-I, Q5 transportation problem
OR Q21. Define transportation problem. State the
assumptions and mathematical model of
What are the advantages of linear programming transportation problem.
techniques? Explain briefly.
Answer :
(Refer Only Topic: Advantages of Linear Programming
Problem) Transportation Problem
Answer : Aug.-17, Q4(a) Transportation problem is a special class of L.P.P, as
the transportation problem is used to provide large quantity
Advantages of Linear Programming Problem of a single similar commodity, which are stored at different
The advantages of LPP are as follows, places or origins in the beginning, to transport the goods
1. It provides an insight view into the problem. This to different destinations in such a manner that the total
generally results in clear picture of the true problem. transportation cost should be reduced.
2. It is helpful in conducting scientific and mathematical Example
analysis of the problematic situations. A spare part manufacturing firm has m factories
3. It gives an opportunity to the decision maker to situated in m different cities. The total supply potential of
formulate his strategies consistent with the constraints the finished product is utilized by ‘ns’ retailer in different
and the objective function. places of the country, then the problem of transportation is
4. It deals with changing situations. Once a plan is to find the transportation pattern that reduces the total cost
prepared through the linear programming it can also of transporting the spare parts from various factory locations
be revaluated depending on changing conditions. to various retail dealers.
5. By using linear programming the decision maker Assumptions of Transportation Problem
makes sure that he is considering the best optimal
solution. Transportation model is based on following
assumptions,
6. Available resources can be utilize in a best possible
way through LPP techniques. 1. It is assumed that quantity available at sources is equal
to the quantity required at destinations.
7. Decision quality can also be improved with the help
of LPP techniques. 2. Items are easily transported from source to destination.
8. It helps in optimum utilization of manpower and 3. Transportation cost per unit from source to destination
productive resources. is well known.
Disadvantages of Linear Programming Problem 4. The main aim is to reduce total transportation cost
for the whole organization.
Following are the disadvantages of LPP,
5. There is direct relationship between transportation
1. The major limitation of linear programming is that it cost of specific route and number of units shipped
treats all relationship as linear. for that specific route.
2. The decision variables in some LPP would be Mathematical Model at Transportation Problem
meaningful only if they have integer values. But,
sometimes we get fractional values to the optimal Transportation problem is a special case of linear
solution, where in practise only integer values are programming problem. Thus, TP can be formulated as an
used. L.P.P. The classical transportation problem can be stated
3. All the parameters in the linear programming model mathematically as follows,
are assumed to be known as constants. But, in real Let,
life they may not be known completely or they may ai = Quantity of product available at origin i
be probabilistic and they may be liable for changes
from time to time. bj = Quantity of product required at destination j
4. The problems are complex if the number of variables Cij = Cost of transporting one unit of product from
and constraints are quite large. origin i to destination j
5. Linear programming deals with only a single objective xij = Quantity from origin i to destination j.
problems, whereas in real life situations, their may be Let, there be “m” sources and “n” destinations so that
more than one objective. i = 1, 2, 3, ..., m and j = 1, 2, 3, ..., n.
The objective is to determine the number of units to (ii) If the demand for the first destination (column)
be transported from source i to destination j. So that, the is exhausted, then move horizontally and make
total transportation cost is minimized. allocation in the first row and second column,
Assume that Σai = Σbj i.e., total quantity available at i.e., cell (1, 2).
the origins is precisely equal to the total amount required at (iii) If both supply and demand are exhausted, then
the destinations. move diagonally and make allocation in the
The linear programming problem is, second row and second column, i.e., cell (2, 2).
Minimize total cost, 4. Continue the procedure until the total available
m n quantity is fully allocated to the cells as required. This
Minimize Z = ∑∑ Cij xij happens when the procedure is repeated till south east
i =1 j =1 corner cell is reached.
Subject to,
n
Problem on NWCM
∑x
j =1
ij = ai for i = 1, 2,..., m 1. Consider the following transportation problem.
m Destination
∑x
i =1
ij = bj for j = 1, 2,..., n
Source
D1 D2 D3 D4
Total
And, xij ≥ 0 for i = 1, 2,..., m O1 1 2 1 4 30

j = 1, 2,..., n
O2 3 3 2 1 50
The objective function and constraint equations
stated above are all linear in xij and thus, resembles a linear O3 4 2 5 9 20
programming problem. Total 20 40 30 10 100
2.7 Methods for Finding Initial
Feasible Solution – North West Determine the initial feasible solution.
Corner Method (NWCM) Solution :
Q22. What are the different methods of finding To obtain IBFS, any one of the three methods viz
initial feasible solution? Discuss briefly North NWCM, LCM, VAM can be choosen.
West Corner Method. Let us solve using NWCM.
Answer :
Step 1
Methods for Finding Initial Feasible Solution
The following are the methods of finding initial The north west corner cell, i.e., (O1, D1) is choosen.
feasible solution, Step 2
1. North West Corner Method (NWCM). The minimum of supply and demand (30, 20) is 20.
2. Least Cost Method (LCM). Allocate 20 units to cell (O1, D1).
3. Vogel’s Approximation Method (VAM).
Step 3
North West Corner Method
Adjust supply and demand. Since, demand for first
North West Corner Method (NWCM Algorithm) is
given below, destination is exhausted move to cell (O1, D2).
1. The cell (1, 1) in the north west corner (i.e., upper Tableau-I
left) of the transportation matrix is choosen first and Total
as many units as possible is allocated to the cell. This D1 D2 D3 D4
will be equal to the minimum of available supply and
demand requirement. O1 1 20 2 1 4 30 10
2. Adjust the supply and demand as per the allocations O2 3 3 2 1 50
made.
O3 4 2 5 9 20
3. (i) If the supply for the first source (row) is
exhausted, then move vertically down and make Total 20 40 30 10 100
allocation in the second row and first column,
0
i.e., cell (2, 1).
Step 4 Tableau-VI
Repeat the procedure of allocation as in step 2 and D1 D2 D3 D4 Total
step 3. Now cell O1, D2 is choosen. The minimum of supply O1 1 2 1 4 30 10 0
and demand (10, 40) is 10. Allocate 10 units to cell O1, D2. 20 10
Adjust the supply and demand. O2 3 3 2 1 50 20 0
Tableau-II 30 20
D1 D2 D3 D4 Total O3 4 2 5 9 20 10 0
10 10
O1 1 2 1 4 30 10 0 Total 20 40 30 10
20 10 0 30 10 0
0 0
O2 3 3 2 1 50
Total cost of transportation = ΣCij .xij
O3 4 2 5 9 20
Where,
Total 20 40 30 10
Cij = Unit cost of transportation
0 30
Tableau-III xij = Number of units allocated.
D1 D2 D3 D4 Total = (1 × 20) + (2 × 10) + (3 × 30) + (2 × 20)
+ (5 × 10) + (9 × 10)
O1 1 2 1 4 30 10 0
= 20 + 20 + 90 + 40 + 50 + 90 = ` 310
20 10
O2 3 3 2 1 50 20 2.7.1 Least Cost Method (LCM) or
30 Matrix Minimum Method
Q23. Explain the method to get an initial basic
O3 4 2 5 9 20
feasible solution by least cost entry method.
Total 20 40 30 10 Answer :
0 30 Least cost method is also known as lowest cost entry
0 method or matrix minima method. To achieve the objective
Tableau-IV of minimum transportation cost, this method consider those
D1 D2 D3 D4 Total routes (or cells) with least unit transportation cost to transport
the goods. The steps of LCM are as follows,
O1 1 2 1 4 30 10 0
20 10 Step 1
O2 3 3 2 1 50 20 0 The cell with the smallest unit cost of transportation
30 20 is choosen and as many units as possible are allocated to that
cell. If there is a tie in the least cost, the cell where maximum
O3 4 2 5 9 20 allocation is possible is choosen.
Total 20 40 30 10 Step 2
0 30 10
Adjust the supply and demand for the allocation
0
made and eliminate (strike out) the row or column in which
Tableau-V either supply or demand is exhausted. If both are exhausted
D1 D2 D3 D4 Total simultaneously, both must be eliminated simultaneously.
O1 1 2 1 4 30 10 0 Step 3
20 10
Repeat steps (1) and (2) with the next smallest unit
O2 3 3 2 1 50 20 0
cost of transportation among the remaining cells (uncrossed-
30 20
out cells).
O3 4 2 5 9 20 10
10
Step 4
Total 20 40 30 10 Continue the procedure until the entire available
0 30 10 supply at various sources and entire demand at various
0 0 destinations are satisfied.
2.45 Problem
Tableau-II
Problem on LCM
D1 D2 D3 D4 Total
1. Consider the following transportation problem.
O1 1 2 1 4 30 0
Destination 30
Source Total O2 3 3 2 1 50 40
D1 D2 D3 D4 10
O1 1 2 1 4 30 O3 4 2 5 9 20
Total 20 40 30 10
O2 3 3 2 1 50 0 0
O3 4 2 5 9 20 Tableau-III
D1 D2 D3 D4 Total
Total 20 40 30 10 100
O1 1 2 1 4 30 0
Determine the initial feasible solution. 30
O2 3 3 2 1 50 40
Solution :
10
This problem was solved using NWCM. Now, let us 0
O3 4 2 5 9 20
solve by LCM. 20
Least Cost Method Total 20 40 30 10
Step 1 20 0 0
The cell with smallest unit cost is choosen. There are Tableau-IV
three cells with cost ‘1’ i.e., (O1, D1), (O1, D3) (O2, D4). The D1 D2 D3 D4 Total
cell with maximum allocation among the least cost cells is O1 1 2 1 4 30 0
chosen. 30
O2 3 3 2 1 50 40 20
Cell Allocation 20 10
(O1, D1) Min(30, 20) = 20 O3 4 2 5 9 20 0
20
(O1, D3) Min(30, 30) = 30 ← Max Total 20 40 30 10
0 20 0 0
(O2, D4) Min(50, 10) = 10
Tableau-V
The cell (O1, D3) is choosen for allocation of 30 units. D1 D2 D3 D4 Total
Step 2 O1 1 2 1 4 30 0
30
Adjust the supply and demand. Eliminate the row or
column for which supply or demand is exhausted. O2 3 3 2 1 50 40 20
20 20 10
Tableau-I O3 4 2 5 9 20 0
D1 D2 D3 D4 Total 20
O1 1 2 1 4 30 Total 20 40 30 10
0
0 20 0 0
30
0
O2 3 3 2 1 50
Note
O3 4 2 5 9 20
In tableau-IV, the least cost cells are (O 2, D 1)
Total 20 40 30 10
0 and (O2, D2). The allocations in these cells are (20, 20)
respectively. As there is tie in not only in least cost cell but
Note also in maximum possible allocation, the choice was made
Since, both O 1 and D 3 are exhausted both are arbitrarily.
eliminated. Total cost of transportation = ΣCij .xij
Step 3 = (1 × 30) + (3 × 20) + (3 × 20) + (1 × 10) + (2 × 20)
Identify the least cost cell among the remaining cells = 30 + 60 + 60 + 10 + 40
i.e., uncrossed cells. Repeat steps with choosen cell until all
the demand and supply are exhausted. = ` 200

Comparing the total cost of transportation, we get,
By NWCM ` 310
By LCM ` 200
Always, cost by NWCM ≥ cost by LCM

2.7.2 Vogel’s Approximation Method (VAM)
Q24. Write the steps for Vogel’s approximation method.
Answer :
VAM or Vogel’s Approximation Method is also known as penalty or regret method. It is basically a heuristic method.
Allocation is done on the basis of opportunity cost (penalty) that would have incurred if allocation in certain cells with
minimum costs were missed.
The procedure of VAM is as follows,
Step 1
Compute the penalties for each row and column by taking the difference between the smallest and next smallest unit
transportation cost in that particular row or column can be determined.
Step 2
Choose the row or column with the largest penalty and allocate as much as possible in the cell having the least cost
in the selected row or column.
If a tie occurs in the penalties, select the one which has the minimum cost.
If there is a tie in the minimum cost also, select that row/column which will have maximum possible allocation.
Step 3
Adjust the supply and demand for the allocation made and eliminate (strike out) the row or column in which either
supply or demand is exhausted. If both are exhausted, eliminate both the rows and columns.
Step 4
Recompute the penalties for the remaining rows and columns (uncrossed rows and columns).
Step 5
Repeat the procedure until the entire available supply at various sources and entire demand at various destinations
are satisfied.
Problems on VAM
1. A company has factories at F1, F2 and F3 which supply warehouses at W1, W2 and W3. Weekly factory
capacities are 200, 160 and 90 units respectively. Weekly warehouse requirements are 180, 120 and
150 units respectively. Unit shipping costs (in rupees) are as follows,
Warehouse
Factory W1 W2 W2 Supply
F1 16 20 12 200
F2 14 8 18 160
F3 26 24 16 90
Demand 180 120 150 450

Determine the optimum distribution for this company to minimize shipping costs.

2.47 Problem
Solution : Feb./March-16, Q5
Warehouses
Supply
W1 W2 W3
F1 16 20 12 200
Factories F2 14 8 18 160
F3 26 24 16 90
Demand 180 120 150 450
IBFS using VAM
Penalty
RP4
16 ←
14
40 0
Penalty
CP4 2 – –
Optimal Distribution/Allocation
For Factory F1
X11 = 140, X12 = 0, X13 = 60
For Factory F2
X21 = 40, X22 = 120, X23 = 0
For Factory F3
X31 = 0, X32 = 0, X33 = 90
Total transportation cost is,
= (16 × 140) + (12 × 60) + (14 × 40) + (8 × 120) + (16 × 90)
= ` 2240 + 720 + 560 + 960 + 1440 = ` 5920.
2. Consider the following transportation problem. Determine the initial feasible solution.
Destination
Source Total
D1 D2 D3 D4
O1 1 2 1 4 30
O2 3 3 2 1 50
O3 4 2 5 9 20
Total 20 40 30 10 100

Solution :
Vogel’s Approximation Method (VAM)
Step 1
Compute row and column penalties by taking the difference between the smallest and next smallest unit cost in each
row and each column.
Step 2
Choose the row or column with largest penalty and make allocations to the least cost cell in that row or column.
Step 3
Adjust the supply and demand. Eliminate the row or column which is exhausted.
Tableau-I
D1 D2 D3 D4 Total Penalty
O1 1 2 1 4 30 0 (1 – 1)
O2 3 3 2 1 50 40 1 (2 – 1)
10
O3 4 2 5 9 20 2 (4 – 2)
Total 20 40 30 10 100
0
Penalty 2 0 1 3 Largest penalty
Tableau-II
Penalty
D1 D2 D3 D4 Total Rp1 Rp2
O1 1 2 1 4 30 10 0 0
20
O2 3 3 2 1 50 40 1 1
10
O3 4 2 5 9 20 2 2
Total 20 40 30 10 100
Largest penalty
0 0
Cp1 2 0 1 3
Penalty
Cp2 2 0 1 –
Largest penalty
Note
Tie in largest penalty. Choice is made based on least cost cell i.e., column 1.
2.49 Problem
Tableau-III
Penalty
D1 D2 D3 D4 Total Rp1 Rp2 Rp3
O1 1 2 1 4 30 10 0 0 1
20
O2 3 3 2 1 50 40 1 1 1
10
O3 4 2 5 9 20 0 2 2 3
20
Total 20 40 30 10 100 Largest penalty
0 20 0
Cp1 2 0 1 3
Penalty Cp2 2 0 1 –
Cp3 – 0 1 –
Tableau-IV
Tableau-V
Penalty
D1 D2 D3 D4 Total Rp1 Rp2 Rp3 Rp4 Rp5
O1 1 2 1 4 30 10 0 0 0 1 1 –
20 10
O2 3 3 2 1 50 40 20 1 1 1 1 1
20 10
O3 4 2 5 9 20 0 2 2 3 – –
20
Total 20 40 30 10 100
0 20 20 0
0
Cp1 2 0 1 3
Cp2 2 0 1 –
Penalty Cp3 – 0 1 –
Cp4 – 1 1 –
Cp5 – 3 2 –

Tableau-VI
0
20
Only left out cell is cell (O2, D3) make the allocation of balance 20 units.
Total transportation cost = ΣCij.xij
= (1 × 20) + (1 × 10) + (3 × 20) + (2 × 20) + (1 × 10) + (2 × 20)
= 20 + 10 + 60 + 40 + 10 + 40 = 180
2.8 Test of Optimality by MODI method

Q25. Explain MODI method for obtaining optimal solution.
Answer :
MODI method or u-v method is an optimal solution technique for Transportation Problem (TP). The algorithm of
MODI method is as follows,
(a) Determine an Initial Basic Feasible Solution (IBFS) using any one of three methods viz NWCM, LCM, VAM.
(b) Ensure that the number of occupied cells is exactly equal to (m + n – 1) where m is number of rows and n is number
of columns.
(If number of occupied cells < (m + n – 1) it is called degeneracy and separate procedure has to be adopted to solve).
(c) Determine a set of numbers for each row and each column. To compute ui and vj form from the equations Cij = ui + vj for
each of the occupied cell. Assign zero to one of the variables (ui or vj) and solve the equations to get the set of
numbers.
(d) Compute the opportunity cost (improvement index) using ∆ij = Cij – (ui + vj) for each of the unoccupied cell.
(e) Check the sign of each opportunity cost. If the opportunity costs of all unoccupied cells is ≥ 0, negative, then the
current solution is the optimal solution. If not, then it implies that the current solution can be improved i.e., total
transportation cost can be reduced and go to step ‘f ’.
(f) Select the unoccupied cell with the largest negative opportunity cost.
(g) Trace a closed path (loop) for the chosen unoccupied cell.
(h) Start with a plus (+) sign at the unoccupied cell and assign alternate plus (+) and minus (–) signs on each corner of the
closed path.
(i) Determine the minimum number of units that should be shipped to this unoccupied cell. The smallest cell with a
negative position is chosen. The quantity is added to all the cells on the path marked with plus (+) sign and subtracted
from those cells marked with minus (–) sign.
(j) Repeat the whole procedure until an optimum solution is attained i.e., when all opportunity costs are either positive
or zero.

2.51 Problem
Problems on MODI Method
1. A cement company has three factories which manufacture cement which is then transported to four
distribution centres. The quantity of monthly production of each factory, the demand of each distribution
centre and the associated transportation cost per quintal are given below:
Distribution Centres Monthly production

Factory (quintals)
W X Y Z
A 10 8 5 4 7,000
B 7 9 15 8 8,000
C 6 10 14 8 10,000
Monthly demand 6,000 6,000 8,000 5,000 25,000
(in quintals)
(a) Suggest the optimum transportation schedule.
(b) Is there any other transportation schedule which is equally attractive? If so, write that.
(c) If the company wants that atleast 5,000 quintals of cement are transported from factory C to
distribution centre Y, will the transportation schedule be any different? If so, what will be the new
optimum schedule and the effect on cost?
Solution : (Model Paper-III, Q4 | Aug.-17, Q5)
(a) The initial feasible schedule using VAM method is as follows,
Monthly Penalties
Production
Distribution Centres
(quintals)
Factory W X Y Z RP1 RP2 RP3 RP4 RP5
A 7000 1 - - - - -
10 8 5 4 7000
B 6000 1000 1000 1000 1 1 1 1 1 8
7 9 15 8 7000
8000
C 6000 4000 2 2 2 - - -
6 10 14 8 4000
10,000
Monthly
demand 1000 1000
6000 6000 8000 5000
(in
quintals)
CP1 1 1 9 4
CP2 1 1 1 0
CP3 - 1 1 0
CP4 - 9 15 8
CP5 - 9 - 8
CP6 - - - 8
Number of allocation = 6
M + n – 1 = 4 + 3 – 1
=7–1=6
∴ Total transportation cost
= (5 × 7000) + (9 × 6000) + (15 × 1000) + (8 × 1000) + (6 × 6000) + (8 × 4000)
= 35000 + 54000 + 15000 + 8000 + 36000 + 32000
= 1, 80, 000
Optimality Check through Modi Method Using Occupied Cells
Cij = ui + vj ; Assume u1 = 0
C13 = u1 + v3 = 5 ; 0 + v3 = 5; v3 = 5
C23 = u2 + v3 = 15; u2 + 5 = 15; u2 = 10
C22 = u2 + v2 = 9; 10 + u2 = 9; v2 = – 1
C24 = u2 + v4 = 8; 10 + v4 = 8; v4 = – 2
C34 = u3 + v4 = 8; u3 – 2 = 8; u3 = 10
C31 = u3 + v1 = 6; 10 + v1 = 6; v1 = – 4
Determining opportunity cost for unoccupied cells using the equations as,
∆ij = Cij – (ui + vj)
∆11 = 10 – (0 – 4) = 10 + 4 = 14
∆12 = 8 – ( 0 – 1) = 8 – 1 = 9
∆14 = 4 – (0 – 2) = 4 + 2 = 6
∆21 = 7 – (10 – 4) = 7 – 6 = 1
∆32 = 10 – (10 –1) = 10 – 9 = 1
∆33 = 14 – (10 + 5) = 14 – 15 = – 1
Since, we have a negative value in cell (3, 3), the solution is not optimum. Construct a loop from cell (3, 3)
with ‘+’ sign.
14 9 7000 6
10 8 5 4 u1 = 0
1 6000 1000 1000

u = 10
7 9 15 8 +θ 2
–θ
6000 1 1 4000
6 10 14 8 – θ u3 = 10
+θ
V1 = – 4 V 2 = – 1 V 3 = 5 V4 = – 2
θ = Min (allocations with – θ assignments)

θ = Min (4,000, 1000)
θ = 1000
The transportation table with modified allocations are as follows,
The minimum negative allocation is 1000. Add this to cells with ‘+’ sign and subtract from cells with ‘–’ sign as
follows,
(2,3) : 1000 – 1000 = 0
(2,4) : 1000 + 1000 = 2000
(3,3) : 0 + 1000 = 1000
(3,4) : 4000 – 1000 = 3000
7000
10 8 5 4 u1 = 0
6000 2000 u2 = 9
7 9 15 8
6000 1000 3000 u3 = 9
6 10 14 8
v1 = – 3 v2 = 0 v3 = 5 v4 = – 1
2.53 Problem
Occupied Cells
Assume u1 = 0
u1 + v3 = 5; 0 + v3 = 5; v3 = 5
u3 + v3 = 14; u3 + 5 = 14; u3 = 9
u3 + v4 = 8; 9 + v4 = 8; v4 = – 1
u2 + v4 = 8; u2 + ( – 1) = 8 ; u2 = 9
u2 + v2 = 9; 9 + v2 = 9; v2 = 0
u3 + v1 = 6; 9 + v1 = 6; v1 = – 3
Unoccupied Cells
∆ij = Cij – (ui + vj)
∆11 = 10 – (0 – 3) = 10 + 3 = 13
∆12 = 8 – (0 + 0) = 8
∆14 = 4 – (0 – 1) = 4 + 1 = 5
∆21 = 7 – (9 – 3) = 7 – 6 = 1
∆23 = 15 – (9 + 5) = 15 – 14 = 1
∆32 = 10 – (9 – 0) = 10 – 9 = 1
Since there are no negative values in unoccupied cells, the solution is optimal,
Total transportation cost,
= ( 5 × 7000) + (9 × 6000) + (8 × 2000) + (6 × 6000) + (14 × 1000) + (8 × 3000)
= ` 1,79,000/-
(b) The above solution is unique there is no other transportation schedule which is equality attractive.
(c) 5000 quintals of cement are to be transported from c to y. The balance suply from factory c will be 5,000 quintals
and balance demand at y will be 3000 quintals as shown below.
Distribution Centres Monthly Penalties (RP)
Factory Production RP1 RP2 RP3 RP4 RP5 RP6
W X Y Z (Quintals)
3000 4000
↑
4000 1 4 – – – –
A 10 8 5 4
7000
1000 6000 1000 1000
↑
B 7 9 15 8 2000 1 1 1 1 1 7
8000
5000
↑
2 2 2 – – –
C 6 10 14 8 5000
6000 5000
6000 3000 20000
1000 1000
Penalty
CP1 1 1 9↑ 4
(CP)
CP2 1 1 – 4
CP3 1 1 – 0
CP4 7 1 – 8
CP5 7 9↑ – 8↑
CP6 7↑ – – –

Check for Optimality Using MODI Method
=m+n–1
=3+4–1
=7–1
=6
Determining Opportunity Cost for Occupied Cells Using Cij = ui + vj , Assume u1 = 0
C13 = u1 + v3 = 0 + 5 Þ v3 = 5
C14 = u1 + v4 = 0 + 4 Þ v4 = 4
C24 = u2 + v4 = u2 + 4 Þ u2 = 4
C21 = u2 + v1 = 4 + v1 = 7 Þ v1 = 3 (7 – 4)
C22 = u2 + v2 = 4 + v2 = 9 Þ v2 = 5 (9 – 4)
C31 = u3 + v1 = u3 + 3 Þ u3 = 3
For Unoccupied Cells Using the Equations as Dij = Cij – (ui + vj)
D11 = C11 – (u1 + v1) Þ 10 – (0 + 3) = 7
D12 = C12 – (u1 + v2) Þ 8 – (0 + 5) = 3
D23 = C23 – (u2 + v3) Þ 15 – (4 + 5) = 6
D32 = C32 – (u3 + v2) Þ 10 – (3 + 5) = 2
D33 = C33 – (u3 + v3) Þ 14 – (3 + 5) = 6
D34 = C34 – (u3 + v4) Þ 8 – (3 + 4) = 1
3000 4000
10 8 5 4 u1 = 0
1000 6000 1000
7 9 15 8 u2 = 4
5000 5000
u3 = 3
6 10 14 8
v1 = 3 v2 = 5 v3 = 5 v4 = 4
Since there are no negative values in unoccupied cells, the solution is optimal,
= (5 × 3000) + (4 × 4000) + (7 × 1000) + (9 × 6000) + (8 × 1000) + (6 × 5000) + (14 × 5000) (Route C – Y)
= 15,000 + 16,000 + 7,000 + 54,000 + 8,000 + 30,000 + 70,000
= 2,00,000.
2. For the transportation problem given by the following tableau, find an initial basic feasible solution by
the North-West corner method and then find an optimal solution.
D E F G H Supply
A 10 15 10 12 20 8
B 5 10 8 15 10 7
C 15 10 12 12 10 10
Demand 5 9 2 4 5

2.55 Problem
Solution : Aug./Sept.-16, Q5 Unoccupied cells
Given, T ij = Cij – (ui + vj)
Supply, M = 8 + 7 + 10 = 25 T13 = C13 – (u1 + v3) = 10 – (0 + 13) = – 3
Demand, N = 5 + 9 + 2 + 4 + 5 = 25 T14 = C14 – (u1 + v4) = 12 – (0 + 13) = – 1
As, supply = Demand the given problem is a balanced T15 = C15 – (u1 + v5) = 20 – (0 + 11) = 9
transportation problem.
T 21 = C21 – (u2 + v1) = 5 – (– 5 + 10) = 0
D E F G H Supply
5 3 3 T 24 = C24 – (u2 + v4) = 15 – (– 5 + 13) = 7
A
10 15 10 12 20 8
6 1
15
1 T 25 = C25 – (u2 + v5) = 10 – (– 5 + 11) = 4
B 5 10 8 7
1 4 5 945 T31 = C31 – (u3 + v1) = 15 – (– 1 + 10) = 6
C 15 10 12 12 10 10
6 1
T32 = C32 – (u3 + v2) = 10 – (– 1 + 15) = – 4
Demand 5 9 2 4 5
Since, the opportunity cost for two cells is negative,
the current solution can be improved.
m + n – 1 = 3 + 5 – 1 = 7 = Allocation (7)
Draw a loop from cell (3, 2) having the largest
\ The total transportation cost, negative opportunity cost – 4
= (5 × 10) + (3 × 15) + (6 × 10) + (1 × 8)
+ (1 × 12) + (4 × 12) + (5 × 10) 5 3

10 15 10 12 20
= 50 + 45 + 60 + 8 + 12 + 48 + 50 – 6 1 +
5 10 8 15 10
= ` 273
+ 1 – 4 5
15
Optimality check using MODI Methods 10 12 12 10
5 3
10 u1 = 0 The Minimum negative allocation is ‘– 1’. Add this
10 15 12 20
to cell with ‘+’ sign and subtract from cell with ‘–’ sign as
6 1 follows,
5 10 8 15 10 u2 = – 5
1 4 5 (3, 2): 0 + 1 = 1
15 u =–1
10 12 12 10 3
(3, 3): 1 – 1 = 0
v1 =10 v2 = 15 v3 = 13 v4 = 13 v5 = 11
(2, 2 ): 6 – 1 = 5
Occupied cells
(2, 3): 1 + 1 = 2
Cij = ui + vj (Assume, u1 = 0)
The improved feasible solution is (with modified
C11 = u1 + v1 = 10 Þ 0 + v1 = 10 Þ v1 = 10 allocations)
C12 = u1 + v2 = 15 Þ 0 + v2 = 15 Þ v2 = 15
5 3
10 15 10 12 20 u1 = 0
C22 = u2 + v2 = 10 Þ u2 + 15 = 10 Þ u2 = – 5
5 5 2
C23 = u2 + v3 = 8 Þ – 5 + v3 = 8 Þ v3 = 13 10 8 15 10 u2 = – 5
1 4 5
C33 = u3 + v3 = 12 Þ 43 + 13 = 12 Þ u3 = – 1 u =–5
15 10 12 12 10 3
v1 =10 v2 = 15 v3 = 13 v4 = 17 v5 = 15
C34 = u3 + v4 = 12 Þ – 1 + v4 = 12 Þ v4 = 13
C35 = u3 + v5 = 10 Þ – 1 + v5 = 10 Þ v5 = 11 Testing the solution for optimality.

Determine ui and vj using occupied cells assuming, The improved feasible solution is (with modified
allocation),
u1 = 0
C11 = u1 + v1 = 0 + v1 = 10 Þ v1 = 10 5 3
20 u1 = 0
10 15 10 12
C12 = u1 + v2 = 0 + v2 = 15 Þ v2 = 15 5 5 2
10 8 15 10 u2 = 0
C22 = u2 + v2 = u2 + 15 = 10 Þ u2 = – 5 4 1 5
u3 = 0
15 10 12 12 10
C23 = u2 + v3 = – 5 + v3 = 8 Þ v3 = 13 v1 =10 v2 = 10 v3 = 8 v4 = 12 v5 = 10
C32 = u3 + v2 = u3 + 15 = 10 Þ u3 = – 5
Testing the solution for optimality.
C34 = u3 + v4 = – 5 + v4 = 12 Þ v4 = 17
Determine ui and vj using occupied cells assuming
C35 = u3 + v5 = – 5 + v5 = 10 Þ v5 = 15 u1 = 0,
Determining opportunity cost for unoccupied cells. C11 = u1 + v1 = 0 + v1 = 10 Þ v1 = 10

The equation is T ij = Cij – (ui + vj),
C14 = u1 + v4 = 0 + v4 = 12 Þ v4 = 12
T13 = C13 – (u1 + v3) = 10 – (0 + 13) = – 3 C22 = u2 + v2 = u2 + 10 = 10 Þ u2 = 0
T14 = C14 – (u1 + v4) = 12 – (0 + 17) = – 5 C23 = u2 + v3 = 0 + v3 = 8 Þ v3 = 8
T15 = C15 – (u1 + v5) = 20 – (0 + 15) = 5 C32 = u3 + v2 = 0 + v2 = 10 Þ v2 = 10
T 21 = C21 – (u2 + v1) = 5 – (– 5 + 10) = 0 C34 = u3 + v4 = u3 + 12 = 12 Þ u3 = 0
T 24 = C24 – (u2 + v4) = 15 – (– 5 + 17) = 3 C35 = u3 + v5 = 0 + v5 = 10 Þ v5 = 10
T 25 = C25 – (u2 + v5) = 10 – (– 5 + 15) = 0 Determining opportunity cost for unoccupied cells.
The equation is T ij = Cij – (ui + vj),
T31 = C31 – (u3 + v1) = 15 – (– 5 + 10) = 10
T12 = C12 – (u1 + v2) = 15 – (0 + 10) = 5
T33 = C33 – (u3 + v3) = 12 – (– 5 + 13) = 4
T13 = C13 – (u1 + v3) = 10 – (0 + 8) = 2
Since the opportunity cost for the two cell is negative, T15 = C15 – (u1 + v5) = 20 – (0 + 10) = 10
the current solution can be improved. Draw a loop from cell
(1, 4) having the largest negative opportunity cost ( – 5). T 21 = C21 – (u2 + v1) = 5 – (0 + 10) = – 5
5 – 3 + T 24 = C24 – (u2 + v4) = 15 – (0 + 12) = 3

10 15 10 12 20
T 25 = C25 – (u2 + v5) = 10 – (0 + 10) = 0
5 2
5 10 8 15 10 T31 = C31 – (u3 + v1) = 15 – (0 + 10) = 5
1 + 4 – 5
13 T33 = C33 – (u3 + v3) = 12 – (0 + 8) = 4
10 12 12 10
Since, the opportunity cost for one cell is negative.
The minimum negative allocation is ‘–3’. Add this to The current solution can be improved. Draw a loop from cell
cell with ‘+’ sign and subtract from cell with ‘–’ as follows, (2, 1) having the largest negative opportunity cost ( – 5)
(1, 2) : 3–3=0 5 – 3 +
20
10 15 10 12
(1, 4) : 0+3=3 + 5 5 – 2
10 8 15 10
(3, 2) : 1+3=4
4 1 – 5
(3, 4) : 4–3=1 15 + 10 12 12 10

2.57 Problem
The minimum negative allocations is ‘1’. Add this to 4 – + 9
cells with ‘+’ sign and subtract from cells ‘–’ as follows, 20
10 15 10 12
1 4 – 2 –
(1, 1) = 5 – 1 = 4 8
+ 5 10 15 10
(1, 4) = 3 + 1 = 4 5 5
15 10 12 12 10
(2, 1) = 0 + 1 = 1
(2, 2) = 5 – 1 = 4 (1, 1) : 4–2=2
(3, 2) = 4 + 1 = 5 (1, 3) : 0+2=2
(3, 4) = 1 – 1 = 0 (2, 1) : 1+2=3
The improved feasible solution is (with modified (2, 3) : 2 – 2 = 0
allocation),
The improved feasible solution is (with modified
4 4 u1 = 0 allocations),
10 15 10 12 20
1 4 2 u2 = – 5 2 2 4
5 10 8 15 10 10 15 10 12 20 u1 = 0
5 5 3 4
u3 = – 5
15 10 12 12 10 5 10 8 15 10 u2 = – 5
v1 =10 v2 = v3 = 13 v4 = 12 v5 = 15 5 5
u3 = – 5
15 15 10 12 12 10
v1 =10 v2 = 15 v3 = 10 v4 = 12 v5 = 15
Testing the solution for optimality. Determine ui and
vj occupied cells Assuming u1 = 0.
Testing the solution for optimality. Detrmine ui and
C11 = u1 + v1 = 0 + v1 = 10 Þ v1 = 10 vj occupied cells Assuming u1 = 0.
C14 = u1 + v4 = 0 + v4 = 12 Þ v4 = 12 C11 = u1 + v1 = 0 + v1 = 10 Þ v1 = 10
C21 = u2 + v1 = u2 + 10 = 5 Þ u2 = – 5
C13 = u1 + v3 = 0 + v3 = 10 Þ v3 = 10
C22 = u2 + v2 = – 5 + v2 = 10 Þ v2 = 15
C14 = u1 + v4 = 0 + v4 = 12 Þ v4 = 12
C23 = u2 + v3 = 0 –- 5+ v3 = 8 Þ v3 = 13
C21 = u2 + v1 = u2 + 10 = 5 Þ u2 = – 5
C32 = u3 + v2 = u3 + 15 = 10 Þ u3 = – 5
C22 = u2 + v2 = - 5+ v2 = 10 Þ v2 = 15
C35 = u3 + v5 = – 5 + v5 = 10 Þ v5 = 15
C32 = u3 + v2 = u3 + 15 = 10 Þ u3 = – 5
Determining opportunity cost for unoccupied cells.
The equation is T ij = Cij – (ui + vj), C35 = u3 + v5 = – 5 + v5 = 10 Þ v5 = 15
T12 = C12 – (u1 + v2) = 15 – (0 + 15) = 0 Determining opportunity cost for unoccupied cells.
T13 = C13 – (u1 + v3) = 10 – (0 + 13) = – 3 The equation is T ij = Cij – (ui + vj),
T15 = C15 – (u1 + v5) = 20 – (0 + 15) = 5 T12 = C12 – (u1 + v2) = 15 – (0 + 15) = 0
T 24 = C24 – (u2 + v4) = 15 – (– 5 + 12) = 8 T15 = C15 – (u1 + v5) = 20 – (0 + 15) = 5

T 25 = C25 – (u2 + v5) = 10 – (– 5 + 15) = 0 T 23 = C23 – (u2 + v3) = 8 – (– 5 + 10) = 3
T31 = C31 – (u3 + v1) = 15 – (– 5 + 10) = 10 T 24 = C24 – (u2 + v4) = 15 – (– 5 + 12) = 8
T33 = C33 – (u3 + v3) = 12 – (– 5 + 13) = 4 T 25 = C25 – (u2 + v5) = 10 – (– 5 + 15) = 0
T34 = C34 – (u3 + v4) = 12 – (– 5 + 12) = 5 T31 = C31 – (u3 + v1) = 15 = – (– 5 + 10) = 10
Since the opportunity cost for one cell is negative,
T33 = C33 – (u3 + v3) = 12 – (– 5 + 10) = 7
the current solution can be improved. Draw a loop from
cell (1, 3) having the largest opportunity cost ( – 3) T34 = C34 – (u3 + v4) = 12 – (– 5 + 12) = 5

Since, all the T ij values are positive, optimality is reached.
Total transportation cost = 2(10) + 2(10) + 4(12) + 3(5) + 4(10) + 5(10) + 5(10)
= 20 + 20 + 48 +15 + 40 + 50 + 50
= ` 243.
3. Solve the following transportation problem:
Destination
Source Supply
1 2 3 4
1 15 18 22 16 30
2 15 19 20 14 40
3 13 16 23 17 30
Demand 20 20 25 35 100
Is the optimal solution obtained by you a unique one? If not, why? What are the alternate optima
then?
Solution : Feb.-17, Q4
Let us obtain optimal transportation problem by using VAM to get IBFs and Modi to get optimal solution.
Vogel’s Approximation Method
Tableau - I
Source Destination Supply Penalty
1 2 3 4 Rp1 Rp2 Rp3 Rp4
10 20 ¬
1 15 16 30 20 1 2 4 4
18 22
5 35 ¬
2 40 50 1 5 1 1
15 19 20 14
20 10 ¬ ¬ –
3 30 10 0 3 1 7
13 16 23 17
Demand 20 0 20 10 25 20 0 35 0 100
Penalty Cp1 2 2 2 2
Cp2 – 3 2 2
Cp3 – 3 2 –
Cp4 – 1 2 –
The total cost of transportation is,

= (10 × 18) + (20 × 22) + (5 × 20) + (35 × 14) + (20 × 13) + (10 × 16)
= ` 1630
Modified Distribution Method (MODI method) checking the IBFS
Number of occupied cells = 6
m+n–1=3+4–1=6
Since, the number of occupied cells is equal to m + n – 1, then,
Determining uij and vij values using occupied cells
cij = ui + vi Assume, u = 0
c12 = u1 + v2 = 18 Þ 0 + v2 = 18 Þ v2 = 18
c13 = u1 + v3 = 22 Þ 0 + v3 = 22 Þ v3 = 22
c23 = u2 + v3 = 20 Þ u2 + 22 = 20 Þ u2 = –2
c24 = u2 + v4 = 14 Þ –2 + v4 = 14 Þ v4 = 16
c31 = u3 + v1 = 13 Þ –2 + v1 = 13 Þ v1 = 15
c32 = u3 + v2 = 16 Þ u3 + 18 = 16 Þ u3 = –2
Determining opportunity cost for unoccupied cells Tableau - III
using the equation as, 1 2 3 4
T ij = cij – (ui + v j) 10 20
1 18 22
T11 = c11 – (u1 – v1) = 15 – (0 + 15) = 0 15 16 30
T14 = c14 – (u 2 – v1) = 16 – (0 + 16) = 0 5 35 40
2 20
15 19 14
T 21 = c 21 – (u 2 – v1) = 15 – (–1 + 15) = 2
10 20
T 22 = c 22 – (u 2 – v 2) = 19 – (–2 + 18) = 3 3 13 16 23 17 30
T33 = c33 – (u3 – v3) = 23 – (–2 + 22) = 3 20 20 25 35
T34 = c34 – (u3 – v 4) = 17 – (–2 + 16) = 3
Optimal Solution
Tabulating these ui and vj values and T ij values in the
transportation table Total transportation cost is,
Tableau - II = (10 × 15) + (20 × 22) + (5 × 20) + (35 × 14)
+ (10 × 13) + (20 × 16)
1 2 3 4
10 20 = ` 1630
1 0 0
15 18 22 16 u1 = 0 It is noted that this, cost is same as obtained in
5 tabaleau - II. Thus alternate optimal solution is ` 1630,
2 2 3 35
15 19 20 14 2 u = – 2 as m + n – 1 = 6 i.e., 3 + 4 – 1 = 6
3 20
13
10
16
3
23
3
17 3
u =–2 2.9 Unbalanced supply and demand
v1 =15 v2 = 18 v3 = 22 v4 = 16 Q26. What do you mean by an unbalanced
transportation problem? How you can modify
unbalanced TP into balanced TP?
Since, all T ij values are $ 0 the current solution is
optimal one. Answer :
Unbalanced Transportation Problem
Total cost of transportation is,
An unbalanced transportation problem refers to a
= (10 × 18) + (20 × 22) + (5 × 20) + (35 × 14) +
situation where in the sum of supplies of all the sources does
(20 × 13) + (10 × 16)
not match with the sum of the demands of all the destinations.
= ` 1,630 Mathematically, it can be represented as,
Check the values of T ij in the iteration it can be seen m n
in tableau II that the cell (1, 1) and cell (1, 4) are equal to
zero. This indicates the presence of multiple optimal solution.
∑ ai ≠ ∑b
j =1
j
i =1
Alternative Optimal Solution When this rim condition is violated, i.e., the total
Draw a closed (loop) path for cell (1, 1) and modify supply is not equal to the total demand, the transportation
allocations as in any iteration of MODI method. problem is said to be an unbalanced one. There are two types
1 2 3 4 of unbalanced TP are,
+ 10 – 20 (a) Total supply exceeds total demand
1 15 18 22 16 (b) Total demand exceeds total supply.
5 35 Modifying Unbalanced TP to Balanced TP
2 5 9 20 14 (a) When total supply exceeds total demand, an additional
20 column (a dummy demand centre) should be added to
10 +
– 13 the transportation table to absorb the excess supply.
3 16 23 17
The unit transportation cost for the cells in this column
The minimum negative allocation is 10. Add this to is set equal to zero because these products are neither
cells with ‘+’ sign and subtract from cells with ‘–’ as follows, made nor transported. By this way, the unbalanced
(1,1) : 0 + 10 = 10 TP becomes a balanced TP.
(1,2) : 10 – 10 = 0 (b) When total demand exceeds total supply an additional
row (a dummy supply centre) should be added to the
(3,1) : 20 – 10 = 10
transportation table to absorb the excess demand.
(3, 2) : 10 + 10 = 20 The unit transportation cost of the cells in this row is
The transportation table with modified allocations is set equal to zero and the unbalanced TP is modified
as follows, to a balanced TP.
Problem on Unbalanced TP (Problems on Unbalanced Supply and Demand)
1. Solve the following transportation problem,

Warehouse
Plant Availability
W1 W2 W3
A 25 17 25 300
B 15 10 18 500
Requirement 300 300 500
Solution : Model Paper-II, Q5
Total demand exceeds total supply.
Tableau-I
Warehouse
Plant Availability
W1 W2 W3
A 25 17 25 300
B 15 10 18 500
Total supply = 800
Total demand = 1100
Since, total supply ≠ total demand, the given TP is an unbalanced one. To covert it to balanced TP add a dummy
supply centre ‘C’ with availability (1100 – 800) = 300 units and unit transportation cost for supply centre ‘C’ as zero. This
balanced TP is solved using VAM to obtain IBFS.
Tableau-II
8 7
100
0
– –
– –
– 17 25 ↑
– 17 ↑ –
ΣCij.xij = (17 × 100) + (25 × 200) + (15 × 300) + (10 × 200) + (0 × 300)
= 1700 + 5000 + 4500 + 2000 + 0
= 13,200
2.61 Problem
Optimal Solution using MODI Method
Checking the IBFS
m + n – 1 = 3 + 3 – 1 = 5
Since, the number of occupied cells is equal to m + n – 1.
(a) Determining ui and vj values for occupied cells using, Cij = ui + vj
(b) Determining ∆ij for unoccupied cells using,
∆ij = Cij – (ui + vj) or ∆ij = Cij – ui – vj
The results are shown in tableau-III.
Tableau-III
W1 W2 W3 Availability ui
A 25 17 25 300 0
3 100 200
B 15 10 18 500 –7
300 200 0
C 0 0 0 300 – 25
3 8 300

vj 22 17 25
Since, ∆ij ≥ 0 for all unoccupied cells, the current solution is optimal one.
Optimal Solution
Cell Number of units Unit cost Total

A – W2 100 17 1700
A – W3 200 25 5000
B – W1 300 15 4500
B – W2 200 10 2000
C – W3 300 0 0
13200
Total cost of shipping = 13,200.
2.10 Degeneracy and its Resolution

Q27. What is degeneracy in transportation problem? How it is resolved?
Answer :
Degeneracy in TP
Degeneracy in TP is said to occur when the number of allocated cells is less than m + n – 1.
Where,
m = Number of rows
n = Number of columns.

If degeneracy is occurred in a TP, then it is not possible to draw a closed loop for every occupied cell while solving
the problem by MODI method. This degeneracy may occur during:
(a) Initial stage
(b) Testing of optimal solution.
Resolving Degeneracy TP
In order to resolve degeneracy, an artificial quantity ε with a ‘o’ cost is added to one or more of the unoccupied cells
such that total allocated cells will be equal to m + n – 1. While placing ε in unoccupied cells, it should be considered that it
has been placed in the cell containing the lowest transportation cost. After placing ε in the unoccupied cell, the problem
can be solved as usual by MODI or stepping stone method. The ε will remain in the problem until degeneracy is removed
or a final solution is obtained.
Problems on degeneracy
1. A company has three plants at locations A, B and C which supply to warehouses located at D, E, F, G and
H. MOnthly plant capacities are 800, 500 and 900 units respectively. Monthly warehouse requirements
are 400, 400, 500, 400 and 800 units respectively. Unit transportation costs (in `) are given below,
To
D E F G H
A 5 8 6 6 3
From B 4 7 7 6 5
C 8 4 6 6 4

Determine an optimum distribution for the company in order to minimize the total transportation cost.
(Model Paper-III, Q5(b) | Aug./Sept.-15, Q5)
OR
Solve the following transportation problem for optimal solution,
P Q R S T Supply
A 5 8 6 6 3 8
B 4 7 7 6 5 5
C 8 4 6 6 4 9
Demand 4 4 5 4 8 22
25
Solution : Sept.-14, Q5
Note: The only difference between two questions i.e., Sept.-14, Q5 and Aug./Sept.-15 is that demand and supply values
are given in hundreds.
Stores
Production
D E F G H capacity
A 5 8 6 6 3 800
Factories
B 4 7 7 6 5 500
C 8 4 6 6 4 900
Requirement 400 400 500 400 800 2200

2500
As total production capacity (2200 units) total requirement (2500 units).

Hence, the given transportation problem is unbalanced. To convert into a balanced transportation problem, add a dummy
factory located at D with production capacity 3 units and unit cost of supplying from various factories is equal to zero.
This balanced transportation problem is then solved using VAM to obtain IBFS.
2.63 Problem
D E F G H Penalty
5 3 ←
A 5 8 6 6 3 8 2 2 2 3 3 –
3
4 1 ←
B 4 7 7 6 5 5 1 1 1 1 1 1 1 1
4 5 ←
C 8 4 6 6 4 9 0 0 0 0 2 2
5
D 3
0 0 0 0 0 3 0 – – – – –
(Dummy)
4 4 5 41 8
5
4 4 6 6↑ 3
4 4 6↑ 0 1
Penalty 4↑ 4 – 0 1
– 4↑ – 0 1
– – – 0 1
– – – 0 1
Total transportation cost = ΣCij . xij

= (6 × 5) + (3 × 3) + (4 × 4) + (6 × 1) + (4 × 4) + (4 × 5) + (0 × 3)
= ` 97
Optimal Solution using MODI Method
m+n–1=4+5–1=9–1=8
Since the number of occupied cells < m + n – 1 there is degeneracy. To resolve degeneracy, ‘∈’ is added to unoccupied
cells with least cost [i.e., cell (AS1)].
Note
∈ is very small value but not equal to zero and
a+∈=a–∈=a
Compute ui, vj, (ui + vj) and ∆ij values and check for optimality.
D E F G H Supply
5 8 6 6 3
A ε 5 5 8
4 7 7 6 5
B 4 1 5
3 4 6 6 4
C 4 5 9
0 0 0 0 0 4
D 3
25
Demand 4 4 5 4 8
25

Determining ui and vj values for occupied cells using, Consider the most negative all C34 (∆ij = – 2 ) and
C = u + v use loop method to find optimum solution.
ij i j
u1 + v1 = 5 [(θ − ∈, 500 − θ, 100 − θ)] select minimum positive

value θ = ∈
u1 + v3 = 6
u1 + v5 = 3 Modified allocation table,
u2 + v1 = 4
u2 + v4 = 6
u3 + v2 = 4
u3 + v5 = 4
u4 + v4 = 0
There are 9 shadow values, but 8 equations available.
Assume u1 = 0 and solve remaining values.
u1 = 0 v1 = 5
u2 = – 1 v2 = 3
u3 = 1 v3 = 6
Optimality Test for Occupied Cells
u4 = – 7 v4 = 7, v5 = 3
Consider unoccupied cells and find their net evaluations. Cij = ui + vj for allocated cells
∆ij = Cij – (ui + vj) Assume u3 = 0

or u1 + v3 = 6 u1 = – 1 v1 = 4
∆ij = Cij – uj – vj u1 + v5 = 3 u2 = 0 v2 = 4
C12 – u1 – v2 = 8 – 0 – 3 = +5 u2 + v1 = 4 u3 = 0 v3 = 7
C14 – u1 – v4 = 6 – 0 – 7 = –1
u2 + v4 = 6 u4 = – 6 v4 = 6
C22 – u2 – v2 = 7 + 1 – 3 = +5
u3 + v2 = 4 v5 = 4
C23 – u2 – v3 = 7 + 1 – 6 = +2
u3 + v4 = 6
C25 – u2 – v5 = 5 + 1 – 3 = +3
C31 – u3 – v1 = 8 – 1 – 5 = + 2 u3 + v5 = 4
C33 – u3 – v3 = 6 – 1 – 6 = – 1 u4 + v4 = 0
C34 – u3 – v4 = 6 – 1 – 7 = – 2 For Unoccupied Cells
C41 – u4 – v1 = 0 + 7 – 5 = + 2 ∆11 = 5 + 1 – 4 = 2
C42 – u4 – v2 = 0 + 7 – 3 = + 4 ∆12 = 8 + 1 – 4 = 5
C43 – u4 – v3 = 0 + 7 – 6 = + 1
∆14 = 6 + 1 – 6 = 1
C45 – u4 – v5 = 0 + 7 – 3 = + 4
∆22 = 7 + 0 – 4 = 3
D E F G H
∆23 = 7 + 0 – 7 = 0
∆25 = 5 + 0 – 4 = 1
∆31 = 8 + 0 – 4 = 4
∆33 = 6 + 0 – 7 = – 1
∆41 = 0 + 6 – 4 = 2
∆42 = 0 + 6 – 4 = 2
∆43 = 0 + 6 – 7 = –1
∆45 = 0 + 6 – 4 = 2
2.65 Problem
Consider –ve cells & from loop from cell (3,3) Since, all ∆ij 0 the solution obtained here is an
[500 – θ , 500 – θ ] minimum positive value θ = 500 optimal one.
= [(3 × 8) + (4 × 4) + (6 × 1) + (4 × 4) + (6 × 3)] × 10
Modified allocation table,
= [24 + 16 + 6 + 16 + 12 + 18] × 10
Minimum total transportation cost = 9,20
Cell Allocation Cost

A F 6×∈ ∈
A H 3×8 24
B D 4×4 16
B G 6×1 6
C E 4×4 16
C F 6×5 30
As there occurs degeneracy at this stage, include
one more ‘∈’ at (3,1) cell to resolve it. C G 6×∈ ∈
Check optimality for Occupied Cells D G 0×3 O
u1 + v3 = 6 u1 = 0 v1 = 4 2. Solve the following transportation problem (cell
entries represent unit costs):
u1 + v5 = 3 u2 = 0 v2 = 4
u2 + v1 = 4 u3 = 0 v3 = 6 Source Warehouses
Availability
destination 1 2 3 4 5 6
u2 + v4 = 6 u4 = – 6 v4 = 6
Plant 1 5 3 7 3 8 5 3
u3 + v2 = 4 v5 = 3 Plant 2 5 6 12 5 7 11 4
u3 + v3 = 6 Plant 3 2 1 3 4 8 2 2
Plant 4 9 6 10 5 10 9 8
u3 + v4 = 6
Required 3 3 6 2 1 2
u4 + v4 = 0
Solution : March-15, Q5
For Unoccupied Cells
Given that,
∆11 = 5 – 0 – 4 = 1
Sources Warehouses
Availability
∆12 = 8 – 0 – 4 = 4
Destination 1 2 3 4 5 6
∆14 = 6 – 0 – 6 = 0
Plant 1 5 3 7 3 8 5 3
∆22 = 7 – 0 – 4 = 3 Plant 2 4
5 6 12 5 7 11
∆23 = 7 – 0 – 6 = 1 Plant 3 2 1 3 4 8 2 2
∆25 = 5 – 0 – 3 = 2 Plant 4 9 6 10 5 10 9 8
∆31 = 8 – 0 – 4 = 4 17
Required 3 3 6 2 1 2
17
∆33 = 4 – 0 – 3 = 1
∆41 = 0 + 6 – 4 = 2 Total availability = 17
∆42 = 0 + 6 – 4 = 2 Total requirement = 17

Total availability is equal to total requirement. Hence,
∆43 = 0 + 6 – 6 = 0
the given transportation is a balanced one. Apply VAM for
∆45 = 0 + 6 – 3 = 3 IBFS.
VAM Method
R6 R7
C6
C7 1

Checking the IBFS
m+n–1=4+6–1=9
Since the number of occupied cells is less than m + n – 1, degeneracy has occurred. To resolve degeneracy, add a
small value ‘∈’ (epsilon) to the least cost cell.
Adding ‘∈’ to (3, 2) cell.
1 2 3 4 5 6 Availability
Plant 1 1 3 7 3 8 5 3
1 2
Plant 2 5 6 12 5 7 11 4
3 1
Plant 3 2 1 3 4 8 2 2
∈ 2
Plant 4 9 6 10 5 10 9 8
2 4 2
Requirement 3 3 6 2 1 2 17

Therefore, the total transportation cost is,
= (3 × 1) + (5 × 2) + (5 × 3) + (7 × 1) + (1 × ∈) + (3 × 2) + (6 × 2) + (10 × 4) + (5 × 2)
= 3 + 10 + 15 + 7 + 0 + 6 + 12 + 40 + 10
= ` 103
2.67 Problem
Short Questions and answers

Q1. What are the characteristics/elements in a typical LLP? What are the related assumptions?
(Model Paper-II, Q1(a) | Aug./Sept.-16, Q1(b))
OR
What are the assumptions in a linear programming problem? Aug./Sept.-15, Q1(b)
(Refer Only Topic: Assumptions of LPP)

OR
What are the assumptions in a linear programming problem? Also briefly comment on the significance
of these assumptions.
(Refer Only Topic: Assumptions of LPP)
Answer : Sept.-14, Q1(b)

Characteristics/Components/Elements of LPP
The major components of LPP are depicted in the following figure,
Components of LPP
Objective Constraints Decision

function variables
Demand Supply Equality

constraints constraints constraints
Figure
Assumptions of LPP
A Linear Programming Problem (LPP) has the following assumptions,
1. Proportionality
This assumption is for both the objective function and the constraints.
2. Additivity
Every function in a linear programming model (whether the objective function or the function on the left hand side
of a functional constraint) is the sum of the individual contributions to the respective activities.
3. Divisibility
Decision variables in a linear programming problem are allowed to have any values including non-integer values
that satisfy the functional and non-negative constraints. Thus, these variables are not restricted to just integer values.
4. Certainty (Deterministic)
The value assigned to each parameter of an LP model is assumed to be known as constant. In real applications, the
certainty assumption is thus satisfied precisely.
5. Multiplicatively
It can be explained with the help of an example if one person takes two hours on one machine, 10 persons will take
20 hours on the same machine.
Every planning model require some basic assumptions likewise linear programming model is also based on the
above assumptions. The significance of these assumptions is dependent on the way which is in formulation of a matrix or
planning exercise. For the purpose of attaining best solution the assumptions to LPP can also be modified.
Q2. Discuss applications of linear programming.
Answer :
Applications of Linear Programming are as follows,
1. Production Management
Linear programming can be applied in production management for determining product mix, product smoothing,
assembly time balancing, production scheduling and blending problem.
2. Marketing Management
Linear programming can be applied in marketing area by helping in analysing the effectiveness of advertising
campaign, advertising media mix and the shortest route for travelling salesman.
3. Manpower Management
LP allows the human resource to analyse personal policy combinations in terms of their appropriateness for maintaining
a steady-state flow of people into the firm and out of the firm.
4. Transportation Problem
LP helps in determining the optimum transportation schedule with minimum total transportation cost of moving
goods from various origins to various destinations.
Q3. What do you mean by graphical method of LPP?
Answer :
Graphical method is a simple method to understand and also to use. This is effectively used in LPP’s which involves
only 2 variables. It gives the graphical representation of the solutions.
All types of solutions are highlighted in this method very clearly. The only drawback is that more the number of
constraints, more will be the straight lines which makes the graph difficult to understand. Some of the characteristics of
graphical method are as follows,
1. This method is very simple and easy to understand.
2. It is very sensitive analysis and can be illustrated easily by drawing graphs.
3. Under this it is easy to obtain optimal solution.
4. It consumes very less time.
Q4. Discuss any four limitations of linear programming.
Answer : March-15, Q1(b)
The following are the limitations of linear programming,
1. The major limitation of linear programming is that it treats all relationship as linear.
2. The decision variables in some LPP would be meaningful only if they have integer values. But, sometimes we get
fractional values to the optimal solution, where in practise only integer values are used.
3. All the parameters in the linear programming model are assumed to be known as constants. But, in real life they
may not be known completely or they may be probabilistic and they may be liable for changes from time to time.
4. The problems are complex if the number of variables and constraints are quite large.
Q5. (i) What is an artificial variable and why is it necessary to introduce it?
(ii) Why is an artificial vector which leaves the basis once never considered again for re-entry into
the basis?
Answer : Feb./March-16, Q1(b)
(i) In case of linear programming problems where constraints have ‘≥’ sings, a slack variable (s) is subtracted from
the constraint in order to convert the inequality constraint into equality constraints.
Example
Constraint, 3x + 4y ≥ 20
Now, subtracting slack variables from the constraint as follows,
3x + 4y – S = 20
Inequality constraint is converted into quality constraint. x and y are no-basic variables and S is a starting basic
variable. If we consider the value of x and y as 0 ‘zero’ than the value of S will be ‘–20’. (i.e., S = –20). Under this
situation, initial basic feasible solution cannot be obtained as the value of S is infeasible.
In order to get initial basic feasible solution and to continue further to obtain optimum solution, a new type of variable
known as the ‘artificial variable’ (A) is introduced. Artificial variables are fictitious variables. Big M method and
two phase method are the two methods wherein artificial variables are used.
(ii) In simplex method, an artificial vector which leaves the basis once never considered again for re-entry into the basis
because the coefficient of such variable becomes non-negative. The variables with negative coefficient can only
enter the basis in the next iteration.
Q6. Explain the following in the context of a LP problem:
(i) Basic solution
(ii) Basic Feasible Solution (BFS)
(iii) Unbounded solution.
Answer : (Model Paper-I, Q1(b) | Aug.-17, Q1(b))
(i) Basic Solution
For a general LP with ‘n’ variables and ‘m’ constraints, a basic solution can be found by setting (n – m) variables
equal to zeros and solving the constraint equations for the value of other m variables. If a unique solution exists, it is a
basic solution.
(ii) Basic Feasible Solution (BFS)
It refers to the set of constraints corresponding to the extreme point of the feasible region.
(iii) Unbounded Solution
Unbounded solution indicates linear programming problem that do not have finite solutions.
Example if in a maximization problem the solution variable can be made infinitely large without violating a constraint
it indicates unbounded solution. In simplex method, this situation can be recognized at the time of deciding the leaving
variables.
If at any iteration, the minimum ratio columns consists of either negatives or infinities, then no variable can be
choosen to leave the basis. Such a situation is called unbounded solution to LPP.
Q7. Briefly explain the concept of degeneracy in LPP.
Answer :
The concept of obtaining a degenerate basic feasible solution in a LPP is known as degeneracy. Degeneracy in a
LPP may arise,
(i) At the initial stage when atleast one basic variable is zero in the initial basic feasible solution.
(ii) At any subsequent iteration when more than one basic variable is eligible to leave the basis and hence one or more
variables becoming zero in the next iteration and the problem is said to be degenerate. There is no assurance that
the value of the objective function will improve, since the new solutions may remain degenerate. As a result it is
possible to repeat the same sequence of simplex iterations endlessly without improving the solutions. This concept
is known as cycling or circling.
Q8. State duality principle.
Answer :
For every linear programming problem there is a unique linear programming problem associated with it, involving
the same data and closely related optimal solutions. The original (given) problem is called “the primal problem” while the
other is called its “dual”. But, in general, the two problems are said to be duals to each other.
The importance of the duality concept is due to two main reasons. Firstly, if the primal contains a large number of
constraints and a smaller number of variables, the work of a computations can be considerably reduced by converting into
the dual problem and then solving it. Secondly, the interpretation of the dual variables from the cost or economic point of
view proves extremely useful in making future decisions in the activities being programmed.
Q9. What are the types of transportation problem?
Answer :
There are two types of transportation problem. They are,
1. Balanced Transportation Problem
A balanced transportation problem is that in which the total of supplies of all the sources is equal to the total of the
demands of all the destinations. It can be mathematically represented as,
m n
ai = ∑i =1
∑b j
j =1
2. Unbalanced Transportation Problem
An unbalanced transportation problem refers to a situation where in the sum of supplies of all the sources does not
match with the sum of the demands of all the destinations. Mathematically, it can be represented as,
m n
∑ ai ≠ ∑b
j =1
j
i =1
When this rim condition is violated, i.e., the total supply is not equal to the total demand, the transportation problem
is said to be an unbalanced one. There are two types of unbalanced TP are,
(a) Total supply exceeds total demand
(b) Total demand exceeds total supply.
Q10. State the different steps required in solving LPP by graphic method.
Answer : (Model Paper-III, Q1(d) | Feb.-17, Q1(b))
The steps involved in generating extreme point solution to an LPP in graphical method are as follows,
(a) Formulate the problem in terms of a series of mathematical constraints and an objective function.
(b) Plot each of the Constraints.
(c) Identify the Feasible Region (Solution Space)
To identify this region, adopt the following points,
For > or ≥ type constraints, shade the region which lies above the constraint lines.
For < or ≤ type constraints, shade the region which lies below these lines.
The common shaded region is the feasible region.
(d) Extreme Point Approach
(i) Identify each of the corner or extreme points of the feasible region.
(ii) Compute the profit or cost at each corner point by substituting the point coordinates into the objective function.
(iii) The corner point which gives highest value of Z for maximization case or lowest value of Z for minimization
case is the optimal solution.
(e) Iso Profit/Iso Cost Line Approach
(i) Assume zero profits or costs to draw the objective function line.
(ii) Draw lines parallel to this line. These are called Iso profit or Iso cost lines.
(iii) Choose the line which touches the feasible region.
(iv) Compute Z using the coordinates of this optimal point.
2.71 Problem
Exercise Problems
1. Solve the following LP problem graphically:
Maximize Z = 20X1 + 80X2
Subject to
4X1 + 6X2 ≤ 90
8X1 + 6X2 ≤ 100
5X1 + 4X2 ≤ 80
X1 and X2 ≥ 0
(Ans: X1 = 0, X2 = 15 and Z = 1200).
2. Solve the following LP problem using simplex method:
Maximize Z = 3X1 + 2X2 + 5X3
Subject to
X1 + X2 + X3 ≤ 9
2X1 + 3X2 + 5X3 ≤ 30
2X1 – X2 – X3 ≤ 8
X1, X2 and X3 ≥ 0
(Ans: X1 = 5, X2 = 0, X3 = 4 and Z = 35).
3. Solve the following LP problem:
Subject to
2X1 + 3X2 ≤ 30
3X1 + 2X2 ≤ 24
X1 + X2 ≥ 3
X1 and X2 ≥ 0
Is the solution unique? If not, give other alternate optimum solutions.
(Ans: X1 = 8, X2 = 0 and Z = 48. The model has alternate optima at all points on the line joining the points (2.4, 8.4)
and (8, 0)).
4. Solve the following LP problem using two-phase method:
Minimize Z = 10X1 + 6X2 + 2X3
Subject to
– X1 + X2 + X3 ≥ 1
3X1 + X2 – X3 ≥ 2
X1, X2 and X3 ≥ 0
(Ans: X1 = 0.25, X2 = 1.25 and Z = 7).

5. Write the dual of the following LP problem:
Subject to
4X1 + 7X2 = 20
5X1 + 2X2 = 10
6X1 + 8X2 = 25
X1 and X2 ≥ 0
(Ans: Minimize Y = 20Y1 + 10Y2 + 25Y3, subject to: 4Y1 + 5Y2 + 6Y3 ≥ 5, 7Y1 + 2Y2 + 8Y3 ≥ 6, where Y1, Y2, Y3 are
unrestricted in sign).
6. Use duality to solve the following LP problem:
Maximize Z = 2X1 + X2
Subject to
X1 + 2X2 ≤ 10
X1 + X2 ≤ 6
X1 – X2 ≤ 2
X1 – 2X2 ≤ 1
X1 and X2 ≥ 0
(Ans: X1 = 4, X2 = 2 and Z = 10).
7. Use penalty (Big-M) method to solve the following LP problem:
Minimize Z = 3x1 – x2
Subject to 2x1 + x2 ≥ 2
x1 + 3x2 ≤ 3
x2 ≥ 4
x1, x2 ≥ 0
(Ans: x1 = 3, x2 = 0 and Max Z = 9).
8. Solve the following LP problem by using Big-M method:
Minimize Z = 2x1 + x2
Subject to 3x1 + x2 = 3
4x1 + 3x2 ≥ 6
x1 + 2x2 ≤ 4
x1, x2 ≥ 0
(Ans: x1 = 3/5, x2 = 6/5 and Min Z = 12/5).
9. A steel company has three open hearth furnaces and five rolling mills. The transportation costs (rupees per quintal)
for shipping steel from furnaces to rolling mills are given in the following table:
M1 M2 M3 M4 M5 Supply
F1 4 2 3 2 6 8
F2 5 4 5 2 1 12
F3 6 5 4 7 7 14
Demand 4 4 6 8 8

What is shipping schedule?
(Ans: x12 = 4, x14 = 4, x24 = 2, x25 = 8, x31 = 4, x33 = 6 and x36 = 4. Total cost = ` 80).

2.73 Problem
10. Consider the following unbalanced transportation problem.
To
I II III Supply
A 5 1 7 10
From B 6 4 6 80
C 3 2 5 15
Demand 75 20 50

Since there is not enough supply, some of the demands at these destinations may not be satisfied. Suppose there are
penalty costs for every unsatisfied demand unit which are given by 5, 3 and 2 for destinations I, II and III, respectively.
Find the optimal solution.
(Ans: x12 = 0, x21 = 60, x22 = 10, x23 = 10, x31 = 5 and x43 = 40. Transportation cost = ` 515. Penalty for transportation
of 40 units to destination 3 at the cost of ` 2 per unit = ` 80. Thus, total cost = 515 + 80 = ` 595 ).
11. ABC Tool Company has a sales force of 25 men, who operate from three regional offices. The company produces
four basic product lines of hand tools. Mr Jain, the sales manager, feels that 6 salesmen are needed to distribute
product line I, 10 to distribute product line II, 4 for product line III and 5 salesmen for product line IV. The cost (in `)
per day of assigning salesmen from each of the offices for selling each of the product lines are as follows:
Product Lines
I II III IV
A 20 21 16 18
Regional Office B 17 28 14 16
C 29 23 19 20

At the present time, 10 salesmen are allocated to office A, 9 to office B and 7 salesmen to office C. How many
salesmen should be assigned from each office to sell each product line in order to minimize costs? Identify alternate
optimum solutions, if any.
(Ans: x12 = 4, x13 = 1, x14 = 5; x21 = 6, x23 = 3, x32 = 2, x35 = 1 Total cost = ` 472
Alternative solution exists because opportunity cost is zero in cell (B, 4) and (C, 4)).
12. A manufacturer wants to ship 8 loads of his product as shown in the table. The matrix gives the mileage from origin
to destination. Shipping costs are ` 10 per load per mile. What shipping schedule should be used?
D1 D2 D3 Supply
O1 50 30 220 1
O2 90 45 170 3
O3 250 200 50 4
Demand 4 2 2

(Ans: x13 = 2, x22 = 1, x23 = 2, x31 = 4 and x33 = 1, Total cost = ` 33).

Very Short Questions and answers (Internal Assessment)
Q1. Define Linear Programming Problem.
Answer :
Linear programming problem is defined as a method which is adopted to optimize an objective function which is
subjected to a set of constraints.
Q2. What do you mean by a Feasibe Solution?
Answer :
A solution that satisfies the constraints of a linear programming problem is called a feasible solution. In other words,
it is also called as any solution to a general linear programming problem which also satisfies the non-negative restriction/
constraints of the problem.
Q3. What is an Objective Function?
Answer :
The general Linear Programming Problem (LPP) set of either maximizing or minimizing for optimizing the linear
function of variable is called an objective function.
Q4. Write a note on Simplex Method.
Answer :
A simplex method is a systematic algorithm which moves from one basic solution to another basic solution. Through
simplex method improved objective function can be obtained.
Q5. Write about Duality Principle.
Answer :
In the context of LPP, duality principle means that each linear programming problem can be analyzed in two different
ways but having equivalent solutions.
Q6. Define Optimality.
Answer :
In LPP, optimality refers to the best possible solution obtained for the current problem. Identification of optimality
principle is one of the important step in solving the LPP. LPP cannot be solved without identifying the optimality principle.
Q7. What is a Balanced Transportation Problem?
Answer :
A balanced transportation problem is that in which the total of supplies of all the sources is equal to the total of the
demands of all the destinations. It can be mathematically represented as,
m n
/a = /b
i j
i=1 j=1
Q8. Write a short note on Degeneracy in T.P.

Answer :
A solution to transportation problem is said to be a degeneracy one, when the number of occupied cells i.e., positive
allocation is less than (m + n – 1) where, ‘m’ is the number of rows and ‘n’ is the number of columns.
Q9. Write a note on Least Cost Method.
Answer :
Matrix minimum method is also known as least cost or lower cost method. In order to achieve the objective of
minimum transportation cost, this method consider those routes (or cells) with least unit transportation cost to transport
the goods.
Q10. Define Transportation Problem.
Answer :
Transportation problem is a special class of LPP as it is used to provide large quantity of a single similar commodity,
which are stored at different places or origins in the beginning to transport goods to different destinations in such a manner
that total transportation cost should be reduced.

Qabd Unit 2 PDF

Uploaded by

Copyright:

Available Formats

You might also like

Qabd Unit 2 PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Qabd Unit 2 PDF

Uploaded by

Copyright:

Available Formats

UNIT-2 Linear Programming Method and Transportation

v Concept of Unbalanced Supply and Demand and Degeneracy in TP.

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

2.1 Linear Programming – Concept, structure, Assumptions and

Objective Constraints Decision

Demand Supply Equality

This is the expression for objective function.

4. Transportation Problem (b) Construction of Objective Function

Thus, minimize z = 0.12 x1 + 0.16 x2 + 0.18 x3 (Machining or Teaching)

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

Points Co-ordinates Objective Function

2.3 Simplex Method

x1 + x2 ≤ 8 A set of variables which are not restricted or equal

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

Maximization Zj 0 × a1 + 0 × a2+ 0×1 0×0

⇒ a1 x1 + a2 x2 + S1 = b1 (S1 ≥ 0) Here, Cj – Coefficient of decision variable and basic

CZj j–− CZj j C01−–C10 C0−

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

x1, x2, x3, s1, s2 $ 0

Calculation of Zj : SCB Xij 10 x1 6 1 5 0 –1 1

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

2x1 + 3x2 + S1 = 30 Answer :

PRoblems on Two phase method

1. Use two-phase simplex method to solve,

Maximize Z = 5x1 + 8x2

Subject to the constraints,

Maximize Z = 5x1 + 8x2 + 0S1 + 0S2 + 0S3

and x1, x2, S1, S2, S3, A1 ≥ 0

(Here: S1, S2 – surplus, S3 – slack, A1, A2 – artificial)

The initial basic feasible solution is given by,

A1 = 3, A2 = 4, S3 = 5 (basic) (x1 = x2 = S1 = S2 = 0, non-basic)

The iterative simplex table for the auxiliary LPP are,

Cj (0 0 0 0 0 –1 –1) Min. Ratio

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

Max Z = 5x1 + 8x2 + 0S1 + 0S2 + 0S3

The iterative simplex tables for this phase are,

Enter S2 and drop x1

New R2 = old R1 + (3/10 × new R1)

New R3 = old R3 – (1/10 × new R1)

Enter S1 and drop S3

Since, all Cj – Zj ≤ 0, the current basic feasible solution is optimal.

2.3.4 Special Cases in Simplex: Unbounded Solution, Unrestricted Variable,

Problem on Unbound Solution

Where xr' ≥ 0 ; xr" ≥ 0

(a) xr' = 0 ⇒ xr = − xr"

(b) xr'' = 0 ⇒ xr = xr'

(c) xr' = xr" = 0 ⇒ xr = 0

1. Solve the following linear programming problem,

Solution : (Model Paper-II, Q4(b) | Sept.-14, Q4)

Standard Form of LPP

Max z = 8x1 – 4 x2′ + 4 x2′′+ 0 S1 + 0 S2

– x1 + 3x2′ + 3x2′′+ S2 = −23

SIA PUBLISHERS and DISTRIBUTORS PVT. LTD.

Problem on Infeasible Solution

Max Z = 3x1 + 2x2