A STUDY ON METHOD FOR SOLVING TRANSPORTATION

PROBLEMS USING STANDARD DEVIATION METHOD

content
 CONTENT

CHAPTER TITLE PAGE NO.

INTRODUCTION

CHAPTER - 1 BASIC DEFINITIONS

CHAPTER -2 UNBALANCED TRANSPORTATION PROBLEM

2.1 Mathematical Formulation of the Unbalanced

Transportation Problem
2.2 Existence of Feasible Solution for Unbalanced TP
2.3 Un Balanced Transportation Table (TT)

CHAPTER - 3 INITIAL BASIC FEASIBLE SOLUTION USING

STANDARD DEVIATION METHOD

3.1 Proposed Standard Deviation Methods to Find an

Initial Basic Feasible Solution.
3.2 Unbalanced Transportation Problems
3.3 Results and Discussions

CONCLUSION
BIBLIOGRAPHY
introduction
 INTRODUCTION

Operation research is a discipline that helps to make better decisions in complex

scenarios by the application of asset of advanced analytical methods. The central objective of
operations research is optimization (i.e.) “to do things best under the given circumstances”.
The primary purpose of operational research is to reduce complex business problems to
manageable mathematical constructs and works that outline expected behavior and goals. The
essential way operations research is used in assisting with business decisions involves
allocating resources like money, people and time.

Linear programming problem which may be solved using a simplified version of

simplex technique called is known as Transportation Problem. It is commonly significant in
the scheme of decision making. It is one of the fundamental problems of network flow problem
which is to transport various quantities of a single homogenous commodity that are initially
stored at various origins to different destinations in such a way that the transportation cost is
minimum.

To get this aim of cost, we have the quantity and locality of already accessible supplies
and the amount demanded on top of the participation united accompanied by each ‘sending’.

L. F. Hitchcock (1941) initially established the basic Transportation Problems.

The motive of transportation problems controls optimal distribution arrangements

between origins / sources and destinations, which are basically accumulated at many origins to
diverse destinations in such a way that the cost of total transportation is minimized.

A balanced condition (total demand=total supply) is assumed, then finding an optimal

schedule of shipment of the commodity with the satisfaction of demands at each destination is
the main goal of this problems

In chapter, I have discussed basic definitions of Transportation Problems. In chapter 2,

I have explained about the formulation of unbalanced transportation problems. In chapter 3, I
have solved the transportation problems using Vogel’s approximation method and standard
deviation method. Finally, we see the results that transportation problem using standard
deviation gives the minimum transportation cost compared to the Vogel’s approximation
method.
Chapter - 1
 BASIC DEFINITIONS

Definition 1.1
Operations Research (OR) may be defined as the science that aims for the application
of analytical and numerical techniques along with information technology to solve
organizational problems.

Definition 1.2
Linear Programming or Linear Optimization is a process which takes into
consideration certain linear relationships to obtain the best possible solution to a mathematical
model. It includes problems dealing with maximizing profits, minimizing costs, minimal usage
of resources, etc.

Definition 1.3

The Transportation Problem is a special type of Linear Programming Problem, where

the objective is to minimise the cost of distributing a product from a number
of sources or origins to a number of destinations.

Structure of the Problem

Let there be 𝑚 origins and 𝑛 destinations. Let the amount of supply at the 𝑖 th origin
is 𝑎𝑖 . Let the demand at 𝑗th destination is 𝑏𝑗 .

The cost of transporting one unit of an item from origin 𝑖 to destination 𝑗 is 𝑐𝑖𝑗 and is
known for all combinations (𝑖, 𝑗). Quantity transported from origin 𝑖 to destination 𝑗 be 𝑥𝑖𝑗 .

The objective is to determine the quantity 𝑥𝑖𝑗 to be transported over all routes (𝑖, 𝑗) so
as to minimize the total transportation cost. The supply limits at the origins and the demand
requirements at the destinations must be satisfied.
 The above transportation problem can be written in the following tabular form:

Destinations

1 2 3 … n Supply

1 (𝑥11 ) (𝑥12 ) (𝑥12 ) … (𝑥1𝑛 ) 𝑎𝟏

𝑐11 𝑐12 𝑐13 𝑐1𝑛
2 (𝑥21 ) (𝑥22 ) (𝑥23 ) … (𝑥2𝑛 ) 𝑎2
Origins

𝑐21 𝑐22 𝑐23 𝑐2𝑛

…

…
m (𝑥𝑚1 ) (𝑥𝑚2 ) (𝑥𝑚3 ) … (𝑥𝑚𝑛 ) 𝑎𝑚
𝑐𝑚1 𝑐𝑚1 𝑐𝑚3 𝑐𝑚𝑛

Demand 𝑏1 𝒃2 𝒃3 … 𝒃𝑛

Definition 1.4

A Feasible Solution to a transportation problem is a set of non-negative values

𝑥𝑖𝑗 (𝑖 = 1,2, … , 𝑚 ; 𝑗 = 1,2, … , 𝑛) that satisfies the constraints.

Definition 1.5
A feasible solution is called a Basic Feasible Solution if it contains not more than
𝑚 + 𝑛 − 1 allocations, where m is the number of rows and n is the number of columns in a
Transportation Problem.
Definition 1.6
Optimal Solution is a feasible solution (not necessarily basic) which optimizes
(minimize) the total transportation cost.

Definition 1.7

The Standard Deviation is defined as the deviation of the values or data from their
average mean. Lower standard deviation will conclude that the values are very close around
their average.

On the other hand, higher values mean the values are far from the mean value. It should
be noted here that the standard deviation value can never be negative.

Definition 1.8

A Balanced Transportation Problem is known as the summation of all supply bases

are equal to the summation of all demand purposes, i.e. ∑𝑚 𝑛
𝑖=1 𝐴𝑖 = ∑𝑗=1 𝐵𝑗 , otherwise it is

known as Unbalanced Transportation Problem i.e. ∑𝑚 𝑛

𝑖=1 𝐴𝑖 > ∑𝑗=1 𝐵𝑗 or ∑𝑚
𝑖=1 𝐴𝑖 <

∑𝑛𝑗=1 𝐵𝑗 .

Definition 1.9
The Vogel’s Approximation Method or VAM is an iterative procedure calculated to
find out the initial feasible solution of the transportation problem. Vogel’s Approximation
Method is also called as Penalty Method because the difference costs chosen are nothing but
the penalties of not choosing the least cost routes.
Chapter – 2
 UNBALANCED TRANSPORTATION PROBLEM

2.1 Mathematical Formulation of the Unbalanced Transportation Problem

Let there be 𝑚 sources of supply, 𝑆1 , 𝑆2 , … , 𝑆𝑚 having 𝑎𝑖 , (𝑖 = 1,2, … , 𝑚) units of
supply (or capacity) respectively, to be transported among 𝑛 destinations, 𝐷1 , 𝐷2 , … , 𝐷𝑛 with
(𝑗 = 1,2, … , 𝑛) units of demand (or requirement) respectively. Let 𝐶𝑖𝑗 be the cost of shipping
one unit of the commodity from source 𝑖 to destination 𝑗 for each route. If 𝑥𝑖𝑗 represents number
of units shipped per route from source 𝑖 to destination 𝑗, the problem is to determine the
transportation schedule so as to minimize the total transportation cost while satisfying the
supply and demand conditions. Mathematically, the problem, in general, may be stated as
follows:
𝑚 𝑛

𝑚𝑖𝑛𝑖𝑚𝑖𝑧𝑒 (𝑇𝑜𝑡𝑎𝑙 𝑐𝑜𝑠𝑡) 𝑍 = ∑ ∑ 𝐶𝑖𝑗 𝑥𝑖𝑗 (1)

𝑖=1 𝑗=1

Subject to the constraints

𝑚

∑ 𝑥𝑖𝑗 = 𝑎𝑖 , 𝑖 = 1,2, … , 𝑚 (𝑠𝑢𝑝𝑝𝑙𝑦 𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑠) (2)

𝑖=1

∑ 𝑥𝑖𝑗 = 𝑏𝑗 , 𝑗 = 1,2, … , 𝑛 (𝑑𝑒𝑚𝑎𝑛𝑑 𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑠) (3)

𝑗=1

And

𝑥𝑖𝑗 ≥ 0 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑖 𝑎𝑛𝑑 𝑗 (4)

In a transportation problem the sum of all available quantities is not equal to the sum of
requirements that is ∑𝑚 𝑛
𝑖=1 𝑎𝑖 ≠ ∑𝑗=1 𝑏𝑗 , then such problem is called an unbalanced

transportation problem
To modify unbalanced transportation problem

An unbalanced TP may occur in two different forms (I) Excess of availability (II)
Shortage in availability. Now discuss these two cases by considering our usual m-origin, n-
destination TP with the condition that ∑𝑚 𝑛
𝑖=1 𝑎𝑖 ≠ ∑𝑗=1 𝑏𝑗 .

Case (I): (𝑬𝒙𝒄𝒆𝒔𝒔 𝒂𝒗𝒂𝒊𝒍𝒂𝒃𝒊𝒍𝒊𝒕𝒚 ∑ 𝒂𝒊 > ∑ 𝒃𝒋 )

The general TP may be stated as follows

𝑚 𝑛

𝑚𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝑧 = ∑ ∑ 𝐶𝑖𝑗 𝑥𝑖𝑗 (5)

𝑖=1 𝑗=1

Subject to the constraints

𝑚

∑ 𝑥𝑖𝑗 < 𝑎𝑖 , 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚 (6)

𝑖=1

∑ 𝑥𝑖𝑗 = 𝑏𝑗 , 𝑓𝑜𝑟 𝑗 = 1,2, … , 𝑛 (7)

𝑗=1

And 𝑥𝑖𝑗 ≥ 0 (8)

Where 𝑐𝑖𝑛+1 = 0 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚 𝑎𝑛𝑑

𝑚 𝑛+1
∑ 𝑎𝑖 ≠ ∑ 𝑏𝑗
𝑖=1 𝑗=1

The problem with possess a feasible solution if ∑ 𝑎𝑖 > ∑ 𝑏𝑗 . In the first constraint, the
introduction of slack variable 𝑥𝑖𝑛+1 (𝑖 = 1,2, … , 𝑚) gives
𝑛

∑ 𝑥𝑖𝑗 + 𝑥𝑖𝑛+1 = 𝑎𝑖 , 𝑓𝑜𝑟 (𝑖 = 1,2, … , 𝑚)

𝑗=1

𝑚 𝑛 𝑚

∑ [∑(𝑥𝑖𝑗 + 𝑥𝑖𝑛+1 )] = ∑ 𝑎𝑖
𝑖=1 𝑗=1 𝑖=1
 𝑛 𝑚 𝑚 𝑚

∑ (∑ 𝑥𝑖𝑗 + ∑ 𝑥𝑖𝑛+1 ) = ∑ 𝑎𝑖
𝑗=1 𝑖=1 𝑖=1 𝑖=1

𝑚 𝑚 𝑛 𝑛

∑ 𝑥𝑖𝑛+1 = ∑ 𝑎𝑖 − ∑ 𝑏𝑗 = 𝐸𝑥𝑐𝑒𝑠𝑠 𝑜𝑓 𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑏𝑛+1 , 𝑠𝑎𝑦 ∑ 𝑥𝑖𝑗 = 𝑏𝑗

𝑖=1 𝑖=1 𝑗=1 𝑗=1

If this excess availability is denoted by 𝑏𝑛+1 the modified general TP can be reformulated as

𝑚 𝑛+1

𝑚𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝑧 = ∑ ∑ 𝐶𝑖𝑗 𝑥𝑖𝑗 (9)

𝑖=1 𝑗=1

∑ 𝑥𝑖𝑗 + 𝑥𝑖𝑛+1 = 𝑎𝑖 , 𝑓𝑜𝑟 (𝑖 = 1,2, … , 𝑚) (10)

𝑗=1

∑ 𝑥𝑖𝑗 = 𝑏𝑗 , 𝑓𝑜𝑟 (𝑖 = 1,2, … , 𝑛 + 1) (11)

𝑖=1

And 𝑥𝑖𝑗 ≥ 0 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚; 𝑗 = 1,2, … , 𝑛 (12)

Hence the above TP is balanced

𝑚 𝑛+1
∑ 𝑎𝑖 = ∑ 𝑏𝑗
𝑖=1 𝑗=1

Case (II): (𝒔𝒉𝒐𝒓𝒕𝒂𝒈𝒆 𝒊𝒏 𝒂𝒗𝒂𝒊𝒍𝒂𝒃𝒊𝒍𝒊𝒕𝒚 𝒊𝒆. ∑ 𝒂𝒊 > ∑ 𝒃𝒋 )

In the case the general TP becomes

𝑚 𝑛

𝑚𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝑧 = ∑ ∑ 𝐶𝑖𝑗 𝑥𝑖𝑗 (13)

𝑖=1 𝑗=1

Subject to the constraints

𝑚

∑ 𝑥𝑖𝑗 = 𝑎𝑖 , 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚 (14)

𝑖=1

∑ 𝑥𝑖𝑗 < 𝑏𝑗 , 𝑓𝑜𝑟 𝑗 = 1,2, … , 𝑛 (15)

𝑗=1
 𝑥𝑖𝑗 ≥ 0 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚; 𝑗 = 1,2, … , 𝑛 (16)

Now introducing the slack variable 𝑥𝑚+1𝑗 , 𝑓𝑜𝑟 𝑗 = 1,2, … , 𝑛 in the second constrained, we get

∑ 𝑥𝑖𝑗 + 𝑥𝑚+1𝑗 = 𝑏𝑗 , 𝑓𝑜𝑟 (𝑗 = 1,2, … , 𝑛)

𝑖=1

𝑛 𝑚 𝑛

∑ (∑(𝑥𝑖𝑗 + 𝑥𝑚+1𝑗 )) = ∑ 𝑏𝑗
𝑗=1 𝑖=1 𝑗=1

𝑛 𝑚 𝑚 𝑛

∑ (∑ 𝑥𝑖𝑗 + ∑ 𝑥𝑚+1𝑗 ) = ∑ 𝑏𝑗
𝑗=1 𝑖=1 𝑖=1 𝑗=1

𝑛 𝑛 𝑚

∑ 𝑥𝑚+1𝑗 = ∑ 𝑏𝑗 − ∑ 𝑎𝑖 = 𝑠ℎ𝑜𝑟𝑡𝑎𝑔𝑒 𝑖𝑛 𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑎𝑚+1 , 𝑠𝑎𝑦

𝑗=1 𝑗=1 𝑖=1

Thus, the modified TP can be reformulated as

𝑚+1 𝑛

𝑚𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝑧 = ∑ ∑ 𝐶𝑖𝑗 𝑥𝑖𝑗 (17)

𝑖=1 𝑗=1

Subject to the constraints

𝑚

∑ 𝑥𝑖𝑗 = 𝑎𝑖 , 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚 + 1 (18)

𝑖=1

∑ 𝑥𝑖𝑗 + 𝑥𝑚+1𝑗 = 𝑏𝑗 , 𝑓𝑜𝑟 𝑗 = 1,2, … , 𝑛 (19)

𝑗=1

𝑥𝑖𝑗 ≥ 0, 𝑓𝑜𝑟 𝑖 = 1,2, … , 𝑚 + 1; 𝑗 = 1,2, … , 𝑛 (20)

Where 𝑐𝑚+1𝑗 = 0 𝑓𝑜𝑟 𝑗 = 1,2, … , 𝑛 𝑎𝑛𝑑 𝑎𝑚+1 = ∑ 𝑏𝑗 − ∑ 𝑎𝑖

Hence, the above TP is balanced

𝑚+1 𝑛
∑ 𝑎𝑖 = ∑ 𝑏𝑗
𝑖=1 𝑗=1
2.2 Existence of Feasible Solution for Unbalanced TP

A necessary and sufficient condition for the existence of a feasible solution to the
unbalanced transportation problem is

𝑚 𝑛+1 𝑚+1 𝑛
∑ 𝑎𝑖 = ∑ 𝑏𝑗 (𝑜𝑟) ∑ 𝑎𝑖 = ∑ 𝑏𝑗
𝑖=1 𝑗=1 𝑖=1 𝑗=1

That is, the total supply must equal to total demand.

2.3 Un Balanced Transportation Table (TT)

Unbalanced Transportation Table: (Excess availability i.e. ∑ 𝑎𝑖 > ∑ 𝑏𝑗 )

Destination
source 𝑫𝟏 𝑫𝟐 𝑫𝟑 … 𝑫𝒋 … 𝑫𝒏 supply
𝑺𝟏 𝑥11 𝑥12 𝑥13 … 𝑥1𝑗 … 𝑥1𝑛 𝑎1
𝑐11
𝑐12 𝑐13 𝑐1𝑗 𝑐1𝑛
𝑺𝟐 𝑥12 𝑥22 𝑥23 … 𝑥2𝑗 … 𝑥2𝑛 𝑎2
𝑐21 𝑐22 𝑐23 𝑐2𝑗 𝑐2𝑛
… …
…

…
…

𝑺𝒊 𝑥𝑖1 𝑥𝑖2 𝑥𝑖3 … 𝑥𝑖𝑗 … 𝑥𝑖𝑛 𝑎𝑖

𝑐𝑖1 𝑐𝑖2 𝑐𝑖3 𝑐𝑖𝑗 𝑐𝑖𝑛
… …
…

𝑺𝒎 𝑥𝑚1 𝑥𝑚2 𝑥𝑚3 … 𝑥𝑚𝑗 … 𝑥𝑚𝑛 𝑎𝑚

𝑐𝑚1 𝑐𝑚2 𝑐𝑚3 𝑐𝑚𝑗 𝑐𝑚𝑛
𝑚 𝑛
Demand 𝑏1 𝑏2 𝑏3 … 𝑏𝑗 … 𝑏𝑛
∑ 𝑎𝑖 > ∑ 𝑏𝑗
𝑖=1 𝑗=1

Table 2.1: Transportation Table of Unbalanced Transportation Problem

Modified Balanced Transportation Table: ∑𝑚 𝑛+1
𝑖=1 𝑎𝑖 = ∑𝑗=1 𝑏𝑗

Destination
source 𝑫𝟏 𝑫𝟐 𝑫𝟑 … 𝑫𝒋 … 𝑫𝒏 𝑫𝒏+𝟏 supply
𝟎𝟏 𝑥11 𝑥12 𝑥13 … 𝑥1𝑗 … 𝑥1𝑛 𝒙𝟏𝒏+𝟏 𝑎1
𝑐11 𝑐12 𝑐13 𝑐1𝑗 𝑐1𝑛 𝒙𝟏𝒏+𝟏

𝟎𝟐 𝑥12 𝑥22 𝑥23 … 𝑥2𝑗 … 𝑥2𝑛 𝒙𝟐𝒏+𝟏 𝑎2

𝑐21 𝑐22 𝑐23 𝑐2𝑗 𝑐2𝑛 𝒙𝟐𝒏+𝟏

… …
…

…
…

…
𝟎𝒊 𝑥𝑖1 𝑥𝑖2 𝑥𝑖3 … 𝑥𝑖𝑗 … 𝑥𝑖𝑛 𝒙𝒊𝒏+𝟏 𝑎𝑖
𝑐𝑖1 𝑐𝑖2 𝑐𝑖3 𝑐𝑖𝑗 𝑐𝑖𝑛 𝒙𝒊𝒏+𝟏

… …
…

…
𝟎𝒎 𝑥𝑚1 𝑥𝑚2 𝑥𝑚3 … 𝑥𝑚𝑗 … 𝑥𝑚𝑛 𝒙𝒎𝒏+𝟏 𝑎𝑚
𝑐𝑚1 𝑐𝑚2 𝑐𝑚3 𝑐𝑚𝑗 𝑐𝑚𝑛 𝒙𝒎𝒏+𝟏

𝒃𝒏+𝟏 =
𝒎 𝒏
Demand 𝑏1 𝑏2 𝑏3 … 𝑏𝑗 … 𝑏𝑛 𝒎 𝒏

∑ 𝒂𝒊 > ∑ 𝒃𝒋 ∑ 𝒂𝒊 > ∑ 𝒃𝒋
𝒊=𝟏 𝒋=𝟏 𝒊=𝟏 𝒋=𝟏

Table 2.2: Transportation Table of Modified Balanced Transportation Problem

The 𝑚 × 𝑛 squares are called cells. The transportation cost 𝑐𝑖𝑗 from the 𝑖 𝑡ℎ source to the
𝑗 𝑡ℎ destination is displayed in the lower right side of the (𝑖, 𝑗)𝑡ℎ cell. Any feasible solution is
shown in the table by entering the values of 𝑥𝑖𝑗 in the upper left side of the (𝑖, 𝑗)𝑡ℎ cell.

The various 𝑎𝑖 ’s and 𝑏𝑗 ’s are called rim requirements. The feasibility of a solution can
be verified by summing the values of 𝑥𝑖𝑗 along the rows and down the columns (i.e., 𝑥𝑖𝑗
satisfying the rim conditions).

Any feasible solution for transportation problem must have exactly (𝑚 + 𝑛 −1) non –negative
basic variables (or) allocations.
Unbalanced Transportation Table: (Shortage in availability i.e., ∑ 𝑎𝑖 > ∑ 𝑏𝑗 )

Destination
source 𝑫𝟏 𝑫𝟐 𝑫𝟑 … 𝑫𝒋 … 𝑫𝒏 supply
𝟎𝟏 𝑥11 𝑥12 𝑥13 … 𝑥1𝑗 … 𝑥1𝑛 𝑎1
𝑐11
𝑐12 𝑐13 𝑐1𝑗 𝑐1𝑛
𝟎𝟐 𝑥12 𝑥22 𝑥23 … 𝑥2𝑗 … 𝑥2𝑛 𝑎2
𝑐21 𝑐22 𝑐23 𝑐2𝑗 𝑐2𝑛
… …
…

…
𝟎𝒊 𝑥𝑖1 𝑥𝑖2 𝑥𝑖3 … 𝑥𝑖𝑗 … 𝑥𝑖𝑛 𝑎𝑖
𝑐𝑖1 𝑐𝑖2 𝑐𝑖3 𝑐𝑖𝑗 𝑐𝑖𝑛
… …
…

…
𝟎𝒎 𝑥𝑚1 𝑥𝑚2 𝑥𝑚3 … 𝑥𝑚𝑗 … 𝑥𝑚𝑛 𝑎𝑚
𝑐𝑚1 𝑐𝑚2 𝑐𝑚3 𝑐𝑚𝑗 𝑐𝑚𝑛
𝑚 𝑛
Demand 𝑏1 𝑏2 𝑏3 … 𝑏𝑗 … 𝑏𝑛
∑ 𝑎𝑖 < ∑ 𝑏𝑗
𝑖=1 𝑗=1

Table 2.3: Transportation Table of Unbalanced Transportation Problem

Modified Balanced Transportation Table: ∑𝑚+1 𝑛
𝑖=1 𝑎𝑖 = ∑𝑗=1 𝑏𝑗

Destination
source 𝑫𝟏 𝑫𝟐 𝑫𝟑 … 𝑫𝒋 … 𝑫𝒏 supply
𝟎𝟏 𝑥11 𝑥12 𝑥13 … 𝑥1𝑗 … 𝑥1𝑛 𝑎1
𝑐11
𝑐12 𝑐13 𝑐1𝑗 𝑐1𝑛
𝟎𝟐 𝑥12 𝑥22 𝑥23 … 𝑥2𝑗 … 𝑥2𝑛 𝑎2
𝑐21 𝑐22 𝑐23 𝑐2𝑗 𝑐2𝑛
… …
…

…
𝟎𝒊 𝑥𝑖1 𝑥𝑖2 𝑥𝑖3 … 𝑥𝑖𝑗 … 𝑥𝑖𝑛 𝑎𝑖
𝑐𝑖1 𝑐𝑖2 𝑐𝑖3 𝑐𝑖𝑗 𝑐𝑖𝑛
… …
…

𝟎𝒎 𝑥𝑚1 𝑥𝑚2 𝑥𝑚3 … 𝑥𝑚𝑗 … 𝑥𝑚𝑛 𝑎𝑚

𝑐𝑚1 𝑐𝑚2 𝑐𝑚3 𝑐𝑚𝑗 𝑐𝑚𝑛

𝟎𝒎 𝑥𝑚+11 𝑥𝑚+12 𝑥𝑚+13 … 𝑐𝑚+1𝑗 𝑐𝑚+12 𝒂𝒎+𝟏 =

𝒏 𝒏
=0 =0
𝑐𝑚+11 = 0 𝑐𝑚+12 = 0 𝑐𝑚+12 = 0 𝑐𝑚+1𝑗 = 0 𝑐𝑚+1𝑛 = 0 ∑ 𝒂𝒊 − ∑ 𝒃𝒋
𝒋=𝟏 𝒋=𝟏

𝑚 𝑛
Demand 𝑏1 𝑏2 𝑏3 … 𝑏𝑗 … 𝑏𝑛
∑ 𝑎𝑖 < ∑ 𝑏𝑗
𝑖=1 𝑗=1

Table 2.4: Transportation Table of Modified Balanced Transportation Problem

CHAPTER – 3
 INITIAL BASIS FEASIBLE SOLUTION USING STANDARD
DEVIATION METHOD

3.1 Proposed Standard Deviation Methods to Find an Initial Basic Feasible

Solution
The stepwise procedure of Standard Deviation Method (SDM) is carried out as follows.
Step 1:
Check the given TP is unbalanced either excess availability or shortage in
availability.

Step 2:
For each row and column reaming under consideration calculate its Standard
1
Deviations (SD). In that rows SD is denoted by 𝜎𝑠𝑖 [= √𝑛 (𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 ] for

𝑖 = 1,2, … , 𝑚, where n is number of destinations and columns SD is denoted by

1
𝑐𝑖𝑗 2 ] for 𝑗 = 1,2, … . 𝑛, where m is number of origins. Identify the
𝜎𝐷𝑖 [= √𝑚 (𝑐𝑖𝑗 − ̅̅̅)

row (or) column with the largest SD and allocate maximum possible to the variable
with least unit TC in the selected row (or) column. So as to exhaust either the supply at
particular origin or satisfy demand at a destination.

Step 3: Rules for ties

1. If a tie occurs in the value of SD, select that row(column) which has minimum
transportation cost.
2. If there is a tie in the minimum cost also, select that row(column) which will
have maximum possible quantity to the cell in that selected row (or) column so as to
exhaust either the supply at particular origin (or) satisfy demand at a destination.
Step 4:
Adjust supply and demand and cross out the satisfied row (or) column. If a row
and a column are satisfied simultaneously only one of the two is crossed out, and
remaining row (column) is assigned zero supply (demand).
Any row (or) column with zero supply or demand should not be used in computing
future SD.
Step 5:
Re-compute the row and column SD for the reduced TT, omitting rows
(column) crossed out in the preceding step. Repeat steps 2 to 4 for the reduced table
until exact one row (or) column with zero supply (demand) remains uncrossed out step.

Step 6:
(I) If one row (column) with non-zero supply (demand) remains uncrossed out,
determine the basic variable in the row(column) by LCM step.
(II) If all the variable in uncrossed out rows and columns have zero supply
(demand), determine the zero basic variables by LCM step.

Step 7:
Given TP is unbalanced TP, there are two cases
(I) ∑ 𝑎𝑖 > ∑ 𝑏𝑗
(II) ∑ 𝑎𝑖 < ∑ 𝑏𝑗

Case(I):

Whenever ∑ 𝑎𝑖 > ∑ 𝑏𝑗 , introduce a dummy destination (𝑛 + 1)th

column in the transportation table. The unit transportation cost to this dummy
destination are the values of 𝑐𝑖𝑛+1=0 for 𝑖 = 1,2, … , 𝑚 the demand of this
dummy destination is 𝑏𝑛+1 = ∑ 𝑎𝑖 − ∑ 𝑏𝑗 . In this case subject to the constraints
th
is ∑𝑛+1
𝑗=1 𝑥𝑖𝑗 ≤ 𝑎𝑖 , 𝑖 = 1,2, … , 𝑚 The unit of amount shifting from 𝑖 origin to

(𝑛 + 1)th destination is 𝑥𝑖𝑛+1 = 𝑎𝑖 − ∑𝑛𝑗=1 𝑥𝑖𝑗 , 𝑖 = 1,2, … , 𝑚.

Case (II):
Whenever ∑ 𝑎𝑖 < ∑ 𝑏𝑗 , Introduce a dummy origin (𝑚 + 1)th row in the
transportation table. The unit transportation cost to this dummy origin is 𝑐𝑚+1𝑗
and the value of 𝑐𝑚+1𝑗 = 0 for 𝑗 = 1,2, … , 𝑛. The supply of this dummy origin
is 𝑎𝑚+1 = ∑ 𝑏𝑗 − ∑ 𝑎𝑖 . The unit of amount shafting from (𝑚 + 1)th origin to
𝑗th destination is 𝑥𝑚+1𝑗 = 𝑏𝑗 − ∑𝑚
𝑖=1 𝑥𝑖𝑗 , for 𝑗 = 1,2, … , 𝑛.
 Step 8:

The balanced transportation table having at least one basic cell in each
row(column) including dummy row(column) and also the basic cell do not form loop.
Hence the initial basic feasible solution is obtained and the minimum transportation
cost equal to

𝑚 𝑛+1 𝑚+1 𝑛 𝑚 𝑛

∑ ∑ 𝐶𝑖,𝑗 𝑥𝑖,𝑗 (𝑜𝑟) ∑ ∑ 𝐶𝑖,𝑗 𝑥𝑖,𝑗 (𝑜𝑟) ∑ ∑ 𝐶𝑖,𝑗 𝑥𝑖,𝑗

𝑖=1 𝑗=1 𝑖=1 𝑗=1 𝑖=1 𝑗=1

3.2 Unbalanced Transportation Problems

Problem 3.2.1
Consider the following transportation problem involving three sources and four
destinations.
Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

𝑆1 6 1 9 3 70

𝑆2 11 5 2 8 55

𝑆3 10 12 4 7 70

Demand 85 35 50 45

Solution:
To find the given problem is balanced or unbalanced transportation problem.
Now, Total no. of supply = ∑𝑖 𝑎𝑖 = 195
Total no. of demand = ∑𝑗 𝑏𝑗 = 215

It seems that total no. of supply and total no. of demand is not equal. Hence, the given
transportation problem is Unbalanced TP.
To change the Unbalanced TP into Balanced TP. Then add a dummy row (𝑆4 )
in the initial transportation table.
 Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

𝑆1 6 1 9 3 70

𝑆2 11 5 2 8 55

𝑆3 10 12 4 7 70

𝑆4 0 0 0 0 20

Demand 85 35 50 45 215

(Table 3.1.1)

Now, No. of supply = No. of demand. Hence the transportation problem is

Balanced TP.

METHOD 1
To find initial basis feasible solution using Vogel’s approximation or penalty
method as follows:

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply Row

difference

𝑆1 6 1 9 3 70 2

𝑆2 11 5 2 8 55 3

𝑆3 10 12 4 7 70 3
20
𝑆4 0 0 0 0 20(0) 0
Demand 85(65) 35 50 45

Column 6 1 2 3
difference
(Table 3.1.2)
 Source 𝐷1 𝐷2 𝐷3 𝐷4 supply Row
difference
65
𝑆1 6 1 9 3 70(5) 2

𝑆2 11 5 2 8 55 3

𝑆3 10 12 4 7 70 3
Demand 65(0) 35 50 45

Column 4 4 2 4
difference
(Table 3.1.3)

Source 𝐷2 𝐷3 𝐷4 supply Row

difference
5
𝑆1 1 9 3 5(0) 2

𝑆2 5 2 8 55 3

𝑆3 12 4 7 70 3
Demand 35(30) 50 45

Column 4 2 4
difference
(Table 3.1.4)
 Source 𝐷2 𝐷3 𝐷4 supply Row
difference
30
𝑆2 5 2 8 55(25) 3

𝑆3 12 4 7 70 3
Demand 30(0) 50 45

Column 7 2 1
difference
(Table 3.1.5)

Source 𝐷3 𝐷4 supply Row

difference
25
𝑆2 2 8 25(0) 6

𝑆3 4 7 70 3
Demand 50(25) 45

Column 2 1
difference
(Table 3.1.6)

Source 𝐷3 𝐷4 supply Row

difference
25
𝑆3 4 7 70(45) 3
Demand 25(0) 45

Column - -
difference
(Table 3.1.7)
 Source 𝐷4 supply Row
difference
45
𝑆3 7 45(0) -
Demand 45(0)

Column -
difference
(Table 3.1.8)

To find the total cost (IBFS) in the given transportation problem is as follows:
Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

65 5
𝑆1 6 1 9 3 70
30 25
𝑆2 11 5 2 8 55

25 45
𝑆3 10 12 4 7 70

20
𝑆4 0 0 0 0 20

Demand 85 35 50 45 215
(Table 3.1.9)

The transportation cost (IBFS) is

= (6 × 65) + (1 × 5) + (5 × 30) + (2 × 25) + (4 × 25) + (7 × 45) + (0 × 20)
= 1010
Hence, the total transportation cost (IBFS) is 1010
METHOD 2
To find initial basis feasible solution using Standard Deviation method as follows:

ITERATION 1

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

𝑆1 6 1 9 3 70

𝑆2 11 5 2 8 55

𝑆3 10 12 4 7 70

𝑆4 0 0 0 0 20

Demand 85 35 50 45

(Table 3.1.10)

𝑭𝒐𝒓𝒎𝒖𝒍𝒂:

1
𝒔𝒕𝒂𝒏𝒅𝒂𝒓𝒅 𝒅𝒆𝒗𝒊𝒂𝒕𝒊𝒐𝒏 (𝒓𝒐𝒘𝒔) = 𝜎𝑠𝑖 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2
𝑛

1
𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑖𝑡𝑜𝑛 (𝑐𝑜𝑙𝑢𝑚𝑛𝑠) = 𝜎𝐷𝑖 = √ 𝑐𝑖𝑗 2
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚

To find column difference using Standard Deviation:

SD (𝜎𝐷1 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

6 0.5625
11 18.0625
10 10.5625
0 45.5625

∑ 𝐶𝑖𝑗 = 27 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 74.75
 ∑ 𝐶𝑖𝑗 27
̅̅̅̅
𝐶𝑖𝑗 = = = 6.75 , where 𝑚 = 4
𝑚 4

1 1
𝜎𝐷1 = √ 𝑐𝑖𝑗 2 = √ (74.75) = 4.32
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟏 = 𝟒. 𝟑𝟐

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝐷2 ) ⟹
1 12.25
5 0.25
12 56.25
0 20.25

∑ 𝐶𝑖𝑗 = 18 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 89

∑𝐶 18
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 4 = 4.5 , where 𝑚 = 4
̅̅̅̅

1 1
𝜎𝐷2 = √ 𝑐𝑖𝑗 2 = √ (89) = 4.72
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟐 = 𝟒. 𝟕𝟐

SD (𝜎𝐷3 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

9 27.5625
2 3.0625
4 0.0625
0 14.0625

∑ 𝐶𝑖𝑗 = 15 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 44.75
 ∑𝐶 15
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 4 = 3.75 , where 𝑚 = 4
̅̅̅̅

1 1
𝜎𝐷3 = √ 𝑐𝑖𝑗 2 = √ (44.75) = 3.34
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟑 = 𝟑. 𝟑𝟒

SD (𝜎𝐷4 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

3 2.25
8 12.25
7 6.25
0 20.25

∑ 𝐶𝑖𝑗 = 18 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 41

∑𝐶 18
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 4 = 4.5 , where 𝑚 = 4
̅̅̅̅

1 1
𝜎𝐷4 = √ 𝑐𝑖𝑗 2 = √ (41) = 3.20
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟒 = 𝟑. 𝟐𝟎
To find row difference using Standard Deviation

SD (𝜎𝑆1 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

6 1.5625
1 14.0625
9 18.0625
3 3.0625

∑ 𝐶𝑖𝑗 = 19 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 36.75

∑𝐶 19
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 4 = 4.75 , where 𝑛 = 4
̅̅̅̅

1 1
𝜎𝑆1 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (36.75) = 3.03
𝑛 4

𝝈𝑺𝟏 = 𝟑. 𝟎𝟑

SD (𝜎𝑆2 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

11 20.25
5 2.25
2 20.25
8 2.25

∑ 𝐶𝑖𝑗 = 26 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 45

∑𝐶 26
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 4 = 6.5 , where 𝑛 = 4
̅̅̅̅
 1 1
𝜎𝑆2 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (45) = 3.35
𝑛 4

𝝈𝑺𝟐 = 𝟑. 𝟑𝟓

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝑆3 ) ⟹
10 3.0625
12 14.0625
4 18.0625
7 1.5625

∑ 𝐶𝑖𝑗 = 33 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 36.75

∑ 𝐶𝑖𝑗 33
̅̅̅̅
𝐶𝑖𝑗 = = = 8.25 , where 𝑛 = 4
𝑛 4

1 1
𝜎𝑆3 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (36.75) = 3.03
𝑛 4

𝝈𝑺𝟑 = 𝟑. 𝟎𝟑

SD (𝜎𝑆4 ) ⟹

𝝈𝑺𝟒 = 𝟎
 Source 𝐷1 𝐷2 𝐷3 𝐷4 supply SD row
difference
35
𝑆1 6 1 9 3 70 (35) 3.03

𝑆2 11 5 2 8 55 3.35

𝑆3 10 12 4 7 70 3.03

𝑆4 0 0 0 0 20 0
Demand 85 35 (0) 50 45

SD column 4.32 4.72 ↑ 3.34 3.20

difference
Table 3.1.11

The highest SD occurs in column 𝐷2 . Allocate maximum possibilities (i.e.) 35

eliminates column 𝐷2 .

ITERATION 2
Source 𝐷1 𝐷3 𝐷4 supply

𝑆1 6 9 3 35

𝑆2 11 2 8 55

𝑆3 10 4 7 70

𝑆4 0 0 0 20
Demand 85 50 45

Table 3.1.12
Column difference using Standard Deviation
𝝈𝑫𝟏 = 𝟒. 𝟑𝟐 𝝈𝑫𝟑 = 𝟑. 𝟑𝟒 𝝈𝑫𝟒 = 𝟑. 𝟐𝟎

To find row difference using Standard Deviation

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

6 0
SD (𝜎𝑆1 ) ⟹
9 9
3 9

∑ 𝐶𝑖𝑗 = 18 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 18

∑𝐶 18
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 3 = 6 , where 𝑛 = 3
̅̅̅̅

1 1
𝜎𝑆1 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (18) = 2.45
𝑛 3

𝝈𝑺𝟏 = 𝟐. 𝟒𝟓

SD (𝜎𝑆2 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

11 16
2 25
8 1

∑ 𝐶𝑖𝑗 = 21 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 42

∑𝐶 21
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 3 = 7 , where 𝑛 = 3
̅̅̅̅
 1 1
𝜎𝑆2 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (42) = 3.74
𝑛 3

𝝈𝑺𝟐 = 𝟑. 𝟕𝟒

SD (𝜎𝑆3 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

10 9
4 9
7 0

∑ 𝐶𝑖𝑗 = 21 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 18

∑ 𝐶𝑖𝑗 21
̅̅̅̅
𝐶𝑖𝑗 = = = 7 , where 𝑛 = 3
𝑛 3

1 1
𝜎𝑆3 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (18) = 2.45
𝑛 3

𝝈𝑺𝟑 = 𝟐. 𝟒𝟓

SD (𝜎𝑆4 ) ⟹

𝝈𝑺𝟒 = 𝟎
 Source 𝐷1 𝐷3 𝐷4 supply SD row
difference

𝑆1 6 9 3 35 2.45

𝑆2 11 2 8 55 3.74

𝑆3 10 4 7 70 2.45

20
𝑆4 0 0 0 20 (0) 0
Demand 85 (65) 50 45

SD column 4.32 ↑ 3.34 3.20

difference
Table 3.1.12
The highest SD occurs in column 𝐷1 . Allocate maximum possibilities (i.e.) 20
eliminates row 𝑆4 .

ITERATION 3
Source 𝐷1 𝐷3 𝐷4 supply

𝑆1 6 9 3 35

𝑆2 11 2 8 55

𝑆3 10 4 7 70
Demand 65 50 45

Table 3.1.13

Row difference using Standard Deviation

𝝈𝑺𝟏 = 𝟐. 𝟒𝟓 𝝈𝑺𝟐 = 𝟑. 𝟕𝟒 𝝈𝑺𝟑 = 𝟐. 𝟒𝟓
To find column difference using Standard Deviation

SD (𝜎𝐷1 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

6 9
11 4
10 1

∑ 𝐶𝑖𝑗 = 27 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 14

∑𝐶 27
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 3 = 9 , where 𝑚 = 3
̅̅̅̅

1 1
𝜎𝐷1 = √ 𝑐𝑖𝑗 2 = √ (14) = 2.16
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 3

𝝈𝑫𝟏 = 𝟐. 𝟏𝟔
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

9 16

SD (𝜎𝐷3 ) ⟹ 2 9
4 1

∑ 𝐶𝑖𝑗 = 15 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 26

∑𝐶 15
𝐶𝑖𝑗 = 𝑖𝑗 = = 5 , where 𝑚 = 3
̅̅̅̅
𝑚 3

1 1
𝜎𝐷3 = √ 𝑐𝑖𝑗 2 = √ (26) = 2.94
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 3

𝝈𝑫𝟑 = 𝟐. 𝟗𝟒
SD (𝜎𝐷4 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

3 9
8 4
7 1

∑ 𝐶𝑖𝑗 = 18 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 14

∑𝐶 18
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 3 = 6 , where 𝑚 = 3
̅̅̅̅

1 1
𝜎𝐷4 = √ 𝑐𝑖𝑗 2 = √ (14) = 2.16
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 3

𝝈𝑫𝟒 = 𝟐. 𝟏𝟔

Source 𝐷1 𝐷3 𝐷4 supply SD row

difference

𝑆1 6 9 3 35 2.45
50
𝑆2 11 2 8 55 (5) 3.74 ⟵

𝑆3 10 4 7 70 2.45

Demand 65 50 (0) 45

SD column 2.16 2.94 2.16

difference
Table 3.1.14
The highest SD occurs in row 𝑆2 . Allocate maximum possibilities (i.e.) 50 eliminates column
𝐷3 .
ITERATION 4

Source 𝐷1 𝐷4 supply

𝑆1 6 3 35

𝑆2 11 8 5

𝑆3 10 7 70
Demand 65 45

Table 3.1.15

Column difference using Standard Deviation

𝝈𝑫𝟏 = 𝟐. 𝟏𝟔 𝝈𝑫𝟒 = 𝟐. 𝟏𝟔

Row difference using Standard Deviation

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝑆1 ) ⟹
6 2.25
3 2.25

∑ 𝐶𝑖𝑗 = 9 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 4.5

∑𝐶 9
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 2 = 4.5 , where 𝑛 = 2
̅̅̅̅

1 1
𝜎𝑆1 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (4.5) = 1.5
𝑛 2

𝝈𝑺𝟏 = 𝟏. 𝟓
SD (𝜎𝑆2 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

11 2.25
8 2.25

∑ 𝐶𝑖𝑗 = 9 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 4.5

∑𝐶 19
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 2 = 9.5 , where 𝑛 = 2
̅̅̅̅

1 1
𝜎𝑆2 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (4.5) = 1.5
𝑛 2

𝝈𝑺𝟐 = 𝟏. 𝟓

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

SD (𝜎𝑆3 ) ⟹ 10 2.25
7 2.25

∑ 𝐶𝑖𝑗 = 9 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 4.5

∑𝐶 17
𝐶𝑖𝑗 = 𝑖𝑗 = = 8.5 , where 𝑛 = 2
̅̅̅̅
𝑛 2

1 1
𝜎𝑆3 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (4.5) = 1.5
𝑛 2

𝝈𝑺𝟑 = 𝟏. 𝟓
 Source 𝐷1 𝐷4 supply SD row
difference
35
𝑆1 6 3 35 (0) 1.5

𝑆2 11 8 5 1.5

𝑆3 10 7 70 1.5
Demand 65 (30) 45

SD column 2.16 ↑ 2.16

difference
Table 3.1.16
The highest SD occurs in column 𝐷1 . Allocate maximum possibilities (i.e.) 35 eliminates
row 𝑆1.

ITERATION 5

Source 𝐷1 𝐷4 supply

𝑆2 11 8 5

𝑆3 10 7 70
Demand 30 45

SD column 2.16 2.16

difference
Table 3.1.17

Row difference using standard deviation

𝝈𝑺𝟐 = 𝟏. 𝟓 𝝈𝑺𝟑 = 𝟏. 𝟓
Column difference using standard deviation

SD (𝜎𝐷1 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

11 0.25
10 0.25

∑ 𝐶𝑖𝑗 = 21 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 0.5

∑𝐶 21
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 2 = 10.5 , where 𝑚 = 2
̅̅̅̅

1 1
𝜎𝐷1 = √ 𝑐𝑖𝑗 2 = √ (0.5) = 0.5
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 2

𝝈𝑫𝟏 = 𝟎. 𝟓

SD (𝜎𝐷4 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

8 0.25
7 0.25

∑ 𝐶𝑖𝑗 = 15 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 0.5

∑𝐶 15
𝐶𝑖𝑗 = 𝑖𝑗 = = 7.5 , where 𝑚 = 2
̅̅̅̅
𝑚 2

1 1
𝜎𝐷4 = √ 𝑐𝑖𝑗 2 = √ (0.5) = 0.5
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 2

𝝈𝑫𝟒 = 𝟎. 𝟓
 Source 𝐷1 𝐷4 supply SD row
difference

𝑆2 11 8 5 1.5
45
𝑆3 10 7 70 (25) 1.5 ←
Demand 30 45 (0)

SD column 0.5 0.5

difference
Table 3.1.18

The highest SD occurs in row 𝑆3 . Allocate maximum possibilities (i.e.) 45 eliminates

column 𝐷4 .

ITERATION 6

Source 𝐷1 supply SD row

difference

𝑆2 11 5 −
25
𝑆3 10 25 (0) −

Demand 30 (5)

SD column ↑ 0.5
difference
Table 3.1.18
The highest SD occurs in column 𝐷1 . Allocate maximum possibilities (i.e.) 25 eliminates row
𝑆3 .

ITERATION 7

Source 𝐷1 supply SD row

difference
5
𝑆2 11 5 (0) −

Demand 5 (0)

SD column 0.5
difference
Hence, we have the following allocations.

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

35 3
𝑆1 6 5 1 9 3 70

5 50
𝑆2 11 5 2 8 55

25 45
𝑆3 10 12 4 7 70

20
𝑆4 0 0 0 0 20

Demand 85 35 50 45 215

The total transportation cost (IBFS) is

= (6 × 35) + (1 × 35) + (5 × 11) + (2 × 50) + (10 × 25) + (7 × 45) + (0 × 20)
= 965

Problem 3.2.2
Consider the following transportation problem involving four sources and three
destinations.

Source 𝐷1 𝐷2 𝐷3 supply

𝑆1 5 6 9 100

𝑆2 3 5 10 75

𝑆3 6 7 6 50

𝑆4 6 4 10 75

Demand 70 80 120
Solution:
To find the given problem is balanced or unbalanced transportation problem.

Now, Total no. of supply = ∑𝑖 𝑎𝑖 = 300

Total no. of demand = ∑𝑗 𝑏𝑗 = 270

It seems that total no. of supply and total no. of demand is not equal. Hence, the given
transportation problem is Unbalanced TP.

To change the Unbalanced TP into Balanced TP. Then add a dummy column (𝐷4 ) in
the initial transportation table.

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

𝑆1 5 6 9 0 100

𝑆2 3 5 10 0 75

𝑆3 6 7 6 0 50

𝑆4 6 4 10 0 75

Demand 70 80 120 30 300

(Table 3.2.1)

Now, No. of supply = No. of demand. Hence the transportation problem is Balanced TP.
METHOD 1
To find initial basis feasible solution using Vogel’s approximation method or
penalty method as follows:

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply Row

difference

𝑆1 5 6 9 0 100 5

𝑆2 3 5 10 0 75 3
30
𝑆3 6 7 6 0 50 (20) 6 ←

𝑆4 6 4 10 0 75 4

Demand 70 80 120 30 (0)

Column 2 1 3 0
difference
(Table 3.2.2)

Source 𝐷1 𝐷2 𝐷3 supply Row

difference

𝑆1 5 6 9 100 1

𝑆2 3 5 10 75 2
20
𝑆3 6 7 6 20(0) 1

𝑆4 6 4 10 75 2

Demand 70 80 120 (100)

Column 2 1 3↑
difference
(Table 3.2.3)
 Source 𝐷1 𝐷2 𝐷3 supply Row
difference

𝑆1 5 6 9 100 1
70
𝑆2 3 5 10 75 (5) ← 2

𝑆4 6 4 10 75 2
Demand
70 (0) 80 100
Column
difference 2 1 1
(Table 3.2.4)

Source 𝐷2 𝐷3 supply Row

difference

𝑆1 6 9 100 3

𝑆2 5 10 5 5
75
𝑆4 4 10 75 (0) ← 6
Demand 80 (5) 100

Column 1 1
difference
(Table 3.2.5)
 Source 𝐷2 𝐷3 supply Row
difference

𝑆1 6 9 100 3
5
𝑆2 5 10 5 (0) ← 5
Demand 5 (0) 100

Column 1 1
difference
(Table 3.2.6)

Source 𝐷3 supply Row

difference
100
𝑆1 9 100 (0) −
Demand 100 (0)

Column −
difference
(Table 3.2.7)

To find the total cost (IBFS) in the given transportation problem is as follows:

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

100
𝑆1 5 6 9 0 100
70 5
𝑆2 3 5 10 0 75
20 30
𝑆3 6 7 6 0 50
75
𝑆4 6 4 10 0 75

Demand 70 80 120 30 300

(Table 3.2.8)
The transportation cost (IBFS) is
= (3 × 70) + (5 × 5) + (9 × 100) + (4 × 75) + (6 × 20) + (0 × 30)
= 1,555
Hence, the total transportation cost (IBFS) is 1,555

METHOD 2
To find initial basis feasible solution using Standard Deviation method as follows:

ITERATION 1

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

𝑆1 5 6 9 0 100

𝑆2 3 5 10 0 75

𝑆3 6 7 6 0 50

𝑆4 6 4 10 0 75

Demand 70 80 120 30 300

(Table 3.2.9)

𝑭𝒐𝒓𝒎𝒖𝒍𝒂:

1
𝒔𝒕𝒂𝒏𝒅𝒂𝒓𝒅 𝒅𝒆𝒗𝒊𝒂𝒕𝒊𝒐𝒏 (𝒓𝒐𝒘𝒔) = 𝜎𝑠𝑖 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2
𝑛

1
𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑖𝑡𝑜𝑛 (𝑐𝑜𝑙𝑢𝑚𝑛𝑠) = 𝜎𝐷𝑖 = √ 𝑐𝑖𝑗 2
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚
To find column difference using Standard Deviation:

SD (𝜎𝐷1 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

5 0
3 4
6 1
6 1

∑ 𝐶𝑖𝑗 = 20 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 6

∑𝐶 20
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 4 = 5 , where 𝑚 = 4
̅̅̅̅

1 1
𝜎𝐷1 = √ 𝑐𝑖𝑗 2 = √ (6) = 1.22
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟏 = 𝟏. 𝟐𝟐

SD (𝜎𝐷2 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
6 0.25
5 0.25
7 2.25
4 2.25

∑ 𝐶𝑖𝑗 = 22 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 5

∑𝐶 22
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 4 = 5.5 , where 𝑚 = 4
̅̅̅̅
 1 1
𝜎𝐷2 = √ 𝑐𝑖𝑗 2 = √ (5) = 1.12
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟐 = 𝟏. 𝟏𝟐

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝐷3 ) ⟹
9 0.0625
10 1.5625
6 7.5625
10 1.5625

∑ 𝐶𝑖𝑗 = 35 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 10.75

∑𝐶 35
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 4 = 8.75 , where 𝑚 = 4
̅̅̅̅

1 1
𝜎𝐷3 = √ 𝑐𝑖𝑗 2 = √ (10.75) = 1.64
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 4

𝝈𝑫𝟑 = 𝟏. 𝟔𝟒

SD (𝜎𝐷4 ) ⟹

𝝈𝑫𝟒 = 𝟎
To find Row Difference using Standard Deviation:

SD (𝜎𝑆1 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

5 0
6 1
9 16
0 25

∑ 𝐶𝑖𝑗 = 20 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 42

∑𝐶 20
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 4 = 5 , where 𝑛 = 4
̅̅̅̅

1 1
𝜎𝑆1 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (42) = 3.24
𝑛 4

𝝈𝑺𝟏 = 𝟑. 𝟐𝟒

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝑆2 ) ⟹ 3 0.25
5 0.25
10 30.25
0 20.25

∑ 𝐶𝑖𝑗 = 18 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 51

∑𝐶 18
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 4 = 4.5 , where 𝑛 = 4
̅̅̅̅
 1 1
𝜎𝑆2 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (51) = 3.57
𝑛 4

𝝈𝑺𝟐 = 𝟑. 𝟓𝟕

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝑆3 ) ⟹
6 1.5625
7 5.0625
6 1.5625
0 22.5625

∑ 𝐶𝑖𝑗 = 19 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 30.75

∑ 𝐶𝑖𝑗 19
𝐶𝑖𝑗 = 𝑛 = 4 = 4.75 , where 𝑛 = 4
̅̅̅̅

1 1
𝜎𝑆3 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (30.75) = 2.77
𝑛 4

𝝈𝑺𝟑 = 𝟐. 𝟕𝟕

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝑆4 ) ⟹
6 1
4 1
10 25
0 25

∑ 𝐶𝑖𝑗 = 20 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 52
 ∑𝐶 20
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 4 = 5 , where 𝑛 = 4
̅̅̅̅

1 1
𝜎𝑆4 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (52) = 3.61
𝑛 4

𝝈𝑺𝟒 = 𝟑. 𝟔𝟏

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply SD row

difference

𝑆1 5 6 9 0 100 3.24

𝑆2 3 5 10 0 75 3.57

𝑆3 6 7 6 0 50 2.77

75
𝑆4 6 4 10 0 75 (0) 3.61 ⟵

Demand 70 80 (5) 120 30

SD column 1.22 1.12 1.64 0

difference
Table 3.2.10

The highest SD occurs in row 𝑆4 . Allocate maximum possibilities (i.e.) 75 eliminates 𝑆4 .

ITERATION 2

Source 𝐷1 𝐷2 𝐷3 𝐷4 Supply

𝑆1 5 6 9 0 100

𝑆2 3 5 10 0 75

𝑆3 6 7 6 0 50

Demand 70 5 120 30
Table 3.2.11
Row difference using Standard Deviation

𝝈𝑺𝟏 = 𝟑. 𝟐𝟒 𝝈𝑺𝟐 = 𝟑. 𝟓𝟕 𝝈𝑺𝟑 = 𝟐. 𝟕𝟕

Column difference using Standard Deviation

SD (𝜎𝐷1 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

5 0.1089
3 2.7889
6 1.7689

∑ 𝐶𝑖𝑗 = 14 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 4.6667

∑𝐶 14
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 3 = 4.67 , where 𝑚 = 3
̅̅̅̅

1 1
𝜎𝐷1 = √ 𝑐𝑖𝑗 2 = √ (4.6667) = 1.25
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 3

𝝈𝑫𝟏 = 𝟏. 𝟐𝟓
 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2
SD (𝜎𝐷2 ) ⟹
6 0
5 1
7 1

∑ 𝐶𝑖𝑗 = 18 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 2

∑𝐶 18
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 3 = 6 , where 𝑚 = 3
̅̅̅̅

1 1
𝜎𝐷2 = √ 𝑐𝑖𝑗 2 = √ (2) = 0.82
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 3

𝝈𝑫𝟐 = 𝟎. 𝟖𝟐

SD (𝜎𝐷3 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

9 0.4489
10 2.7889
6 5.4289

∑ 𝐶𝑖𝑗 = 25 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 8.6667

∑𝐶 25
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 3 = 8.33 , where 𝑚 = 3
̅̅̅̅

1 1
𝜎𝐷3 = √ 𝑐𝑖𝑗 2 = √ (8.6667) = 1.70
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 3

𝝈𝑫𝟑 = 𝟏. 𝟕𝟎
 Source 𝐷1 𝐷2 𝐷3 𝐷4 Supply SD row
difference

𝑆1 5 6 9 0 100 3.24

70
𝑆2 3 5 10 0 75 (5) 3.57 ⟵

𝑆3 6 7 6 0 50 2.77

Demand 70 (0) 5 120 30

SD column 1.25 0.82 1.70 0

difference
Table 3.2.12
The highest SD occurs in row 𝑆2 . Allocate maximum possibilities (i.e.) 70 eliminates 𝐷1 .

ITERATION 3

Source 𝐷2 𝐷3 𝐷4 Supply

𝑆1 6 9 0 100

𝑆2 5 10 0 5

𝑆3 7 6 0 50

Demand 5 120 30
Table 3.2.13

Column difference using Standard Deviation

𝝈𝑫𝟐 = 𝟎. 𝟖𝟐 𝝈𝑫𝟑 = 𝟏. 𝟕𝟎 𝝈𝑫𝟒 = 𝟎

Row difference using Standard Deviation

SD (𝜎𝑆1 ) ⟹
𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2

6 1
9 16
0 25

∑ 𝐶𝑖𝑗 = 15 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 =42

∑𝐶 15
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 3 = 5 , where 𝑛 = 3
̅̅̅̅

1 1
𝜎𝑆1 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (42) = 3.74
𝑛 3

𝝈𝑺𝟏 = 𝟑. 𝟕𝟒

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝑆2 ) ⟹
5 0
10 25
0 25

∑ 𝐶𝑖𝑗 = 15 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 50

∑𝐶 15
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 3 = 5 , where 𝑛 = 3
̅̅̅̅

1 1
𝜎𝑆2 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (50) = 4.08
𝑛 3

𝝈𝑺𝟐 = 𝟒. 𝟎𝟖
 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅
𝐶𝑖𝑗 )2
SD (𝜎𝑆3 ) ⟹ 7 7.1289
6 2.7889
0 18.6667

∑ 𝐶𝑖𝑗 = 13 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 28.6667

∑𝐶 13
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 3 = 4.33 , where 𝑛 = 3
̅̅̅̅

1 1
𝜎𝑆3 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (28.6667) = 3.09
𝑛 3

𝝈𝑺𝟑 = 𝟑. 𝟎𝟗

Source 𝐷2 𝐷3 𝐷4 Supply SD row

difference

𝑆1 6 9 0 100 3.74

5
𝑆2 5 10 0 5 (0) 4.08 ⟵

𝑆3 7 6 0 50 3.09

Demand 5 (0) 120 30

SD column 0.82 1.70 0

difference
Table 3.2.14

The highest SD occurs in row 𝑆2 . Allocating maximum possibilities (i.e.) 5 eliminates 𝑆2

and 𝐷2 .
ITERATION 4

Source 𝐷3 𝐷4 Supply

𝑆1 9 0 100

𝑆3 6 0 50

Demand 120 30
Table 3.2.15

Column difference using Standard Deviation

𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2
SD (𝜎𝐷3 ) ⟹
9 2.25
6 2.25

∑ 𝐶𝑖𝑗 = 15 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 =4.5

∑𝐶 15
𝐶𝑖𝑗 = 𝑚𝑖𝑗 = 2 = 7.5 , where 𝑚 = 2
̅̅̅̅

1 1
𝜎𝐷3 = √ 𝑐𝑖𝑗 2 = √ (4.5) = 1.5
∑(𝑐𝑖𝑗 − ̅̅̅)
𝑚 2

𝝈𝑫𝟑 = 𝟏. 𝟓

SD (𝜎𝐷4 ) ⟹

𝝈𝑫𝟒 = 𝟎
Row difference using Standard Deviation

SD (𝜎𝑆1 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

9 20.25
0 20.25

∑ 𝐶𝑖𝑗 = 9 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 =40.5

∑𝐶 9
𝐶𝑖𝑗 = 𝑛 𝑖𝑗 = 2 = 4.5 , where 𝑛 = 2
̅̅̅̅

1 1
𝜎𝑆1 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (40.5) = 4.5
𝑛 2

𝝈𝑺𝟏 = 𝟒. 𝟓

SD (𝜎𝑆3 ) ⟹ 𝐶𝑖𝑗 (𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2

6 9
0 9

∑ 𝐶𝑖𝑗 = 6 ∑(𝐶𝑖𝑗 − ̅̅̅̅

𝐶𝑖𝑗 )2 = 18

∑𝐶 6
𝐶𝑖𝑗 = 𝑖𝑗 = = 3 , where 𝑛 = 2
̅̅̅̅
𝑛 2

1 1
𝜎𝑆3 = √ ∑(𝑐𝑖𝑗 − ̅̅̅)
𝑐𝑖𝑗 2 = √ (18) = 3
𝑛 2

𝝈𝑺𝟑 = 𝟑
 Source 𝐷3 𝐷4 Supply SD row
difference

30
𝑆1 9 0 100 (70) 4.5 ⟵

𝑆3 6 0 50 3

Demand 120 30 (0)

SD column 1.5 0
difference
Table 3.2.16

The highest SD occurs in row 𝑆1 . Allocating maximum possibilities (i.e.) 30 eliminates

column 𝐷4 .

ITERATION 5

Source 𝐷3 Supply SD row

difference

𝑆1 9 70 −

50
𝑆3 6 50 (0) −

Demand 120 (70)

SD column 1.5 ↑
difference
Table 3.2.17

The highest SD occurs in column 𝐷3 . Allocate maximum possibilities (i.e.) 50 eliminates

row 𝑆3 .
ITERATION 6

Source 𝐷3 Supply SD row

difference
70
𝑆1 9 70 (0) −

Demand 70 (0)

SD column −
difference
Table 3.2.18

Hence, we have the following allocations.

Source 𝐷1 𝐷2 𝐷3 𝐷4 supply

70 30
𝑆1 5 6 9 0 100

70 5
𝑆2 3 5 10 0 75
50
𝑆3 6 7 6 0 50

75
𝑆4 6 4 10 0 75

Demand 70 80 120 30 300

Table 3.2.18

Hence, the total transportation cost is

= (9 × 70) + (0 × 30) + (3 × 70) + (5 × 5) + (6 × 50) + (4 × 75)

= 1465
3.3 Results and discussions
After obtaining an optimum solution by the proposed “Standard Deviation Method
(SDM)”, the obtained result is compared with the results obtained by Vogel’s Approximation
Methods (VAM) is shown in Table 3.3.1

Problem Name of the method Allocations Optimal solution

3.2.1 VAM 𝑥11 = 65; 𝑥12 = 5; 𝑥22 = 30 1010
𝑥23 = 25; 𝑥33 = 25; 𝑥34 = 45
𝑥41 = 20
SDM 𝑥11 = 35; 𝑥12 = 3 5; 𝑥21 = 5 965
𝑥23 = 50; 𝑥31 = 25; 𝑥34 = 45
𝑥41 = 20
3.2.2 VAM 𝑥13 = 100; 𝑥21 = 70; 𝑥22 = 5 1555
𝑥33 = 20; 𝑥34 = 30; 𝑥42 = 75
SDM 𝑥13 = 70; 𝑥14 = 30; 𝑥21 = 70 1465
𝑥22 = 5; 𝑥33 = 50; 𝑥42 = 75
Table 3.3.1

As can be shown from Table 3.3.1, the SDM offers a better first fundamental feasible solution
in comparison to the outcomes produced by the conventional method VAM. The effectiveness
of SDM has also been examined through the resolution of several unbalanced transportation
issues are cost-minimized, and it is discovered that the SDM method produces the greatest
results overall.
ConClusion
 CONCLUSION

The information supplied in this study and covered above provides an IBFS of the
unbalanced TP. According to the problem used to explain this method, SDM requires fewer
calculations and provides the best possible answer when compared to VAM. SDM gives an
immediate optimal solution or a better optimal solution with the fewest possible iterations to
the unbalanced transportation issue. The technique presented here guarantees a very good
solution to identify the IBFS of an unbalanced TP. Finally, it can be asserted that the SDM, by
assuring minimal transportation costs, may offer a wonderful First Basic Feasible Solution.
This will assist individuals who desire to maximise their profit while reducing their
transportation costs in achieving their objectives.
bibliography
 BIBLIOGRAPHY

1. International Journal of Pure and Applied Mathematics, Volume 119 No. 16

2018, 4971-4989, ISSN: 1314-3395 (on-line version) url:
http://www.acadpubl.eu/hub/