Thesis Aminur

Analysis and Re-solution of the Transportation
Problem: A Linear Programming Approach
A dissertation submitted in partial fulfillment of the requirements for the degree of

Master of Philosophy in Mathematics
Submitted by:
AMINUR RAHMAN KHAN
Examination Roll No: 08073
Registration No: 1458
Session: 2007-2008
Supervisor:
DR. MD. SHARIF UDDIN
Associate Professor
Department of Mathematics
Department of Mathematics
Faculty of Mathematical and Physical Sciences
Jahangirnagar University
Savar, Dhaka-1342
Bangladesh
April, 2012
Acknowledgement
All great abundant thanks to the Almighty Allah, the Omniscient and the sole nourish the
sustainer of the universe, who taught me and guided my step to this humble effort.
I would like to express my profound sense of gratitude and sincere thanks to my

supervisor, Dr. Md. Sharif Uddin, Associate Professor, Department of Mathematics,
Jahangirnagar University for his inspiration, constant assistance, valuable suggestions,
sympathetic advice, fruitful conversation and unparalleled encouragement made
throughout the course of study, without which this piece of work would not have taken
its present shape.
I would like to express my deep reverence to Professor Mohammad Abdul Malek, Abdur
Rashid, Abu Sayed Md. Mustafizur Rahaman, Md. Ajijur Rahman, Md. Main Uddin,
Md. Amirul Islam, Sushanta Kumer Roy and Md. Ashraful Babu for their co-operation,
valuable remarks and encouragement during the course of this work.
At the last but not least, I would like to express my gratitude to my family. I am indebted
to my parents and other family members for their continuous correspondence and
making an inspirational environment for me which help me to continue my thesis work
more vigorously.
Aminur Rahman Khan

Declaration
I hereby announce that the work which is being presented in this thesis entitled ‘Analysis
and Re-solution of the Transportation Problem: A Linear Programming Approach’
submitted in partial fulfillment of the requirements for the decoration of the degree of
Master of Philosophy in Mathematics, Department of Mathematics, Jahangirnagar
University, Savar, Dhaka is my own work with proper citation and acknowledgement.
The work of this thesis has not been submitted by me for the award of any other degree
in this University or any other Universities.
Place: Jahangirnagar University, Savar, Dhaka. Aminur Rahman Khan
Dated: April 1, 2012 M. Phil Student
Dept. of Mathematics
Certification
This is to certify that the research work entitled ‘Analysis and Re-solution of the
Transportation Problem: A Linear Programming Approach’ has been carried out by
Aminur Rahman Khan bearing Examination Roll No.: 08073, Registration No.: 1458,
Session: 2007-2008 of Department of Mathematics, Jahangirnagar University, Savar,
Dhaka-1342, Bangladesh in partial fulfillment of the requirements for the degree of
Master of Philosophy in Mathematics of the aforementioned University, under my direct
supervision.
It is to be mentioned that he presented his first seminar on 24 September, 2011 and

second seminar on 25 February, 2012 successfully in the presence of Chairman of the
Department, President of Higher Studies Committee of the Department, Supervisor and
the available teachers and students.
To the best of my knowledge no part of this thesis has been submitted for any other
degree in this University or any other Universities.
Place: Jahangirnagar University, Savar, Dhaka. Dr. Md. Sharif Uddin
Dated: April 1, 2012 Supervisor & Associate Professor
Dept. of Mathematics
Dedicated
to
My Parents
Abstract
The constraint structure of the transportation problem is so important that the literature is
filled with efforts to provide efficient algorithms for solving it. The intent of this
dissertation is to present an efficient approach for finding an initial basic feasible
solution of the transportation problem (TP). We also build up a C program of the
presented algorithm. The proposed method is illustrated with several numerical
examples. An analysis to identify the reason of different solutions by different methods
is also carried out. Finally a comparative study is made, and we observed that the method
presented herein gives a better result.
i
Contents
TOPICS PAGE NO.
Chapter 1 : Introduction 1
1.1 Background of the study 3
1.2 Why we choose Transportation Problem 3
1.3 Structure of the study 4
Chapter 2 : Transportation Problem 6

2.1 Definition 6
2.2 Main Objectives 6
2.3 Basic Steps 6
2.4 Mathematical Formulation 7
2.4.1 The decision variables 7
2.4.2 The objective function 7
2.4.3 The constraints 8
2.5 Network Representation 9
2.6 Balanced Transportation Problem 10
2.7 Unbalanced Transportation Problem 11
2.8 Matrix Terminology 11
2.9 Some Basic Definitions 12
Chapter 3 : Existing Solution Procedure 15

3.1 Row Minimum Method 15
3.2 Column Minimum Method 16
3.3 North West Corner Method 17
3.4 Least Cost Method 18
3.5 Vogel’s Approximation Method 19
3.6 Extremum Difference Method 20
ii
3.7 Optimality Test 22
3.7.1 Stepping Stone Method 22
3.7.2 Modified Distribution Method 24
Chapter 4 : Proposed Optimization Technique 26

4.1 Algorithm 26
4.2 Flowchart 28
4.3 C Program 30
Chapter 5 : Numerical Illustration 44

5.1 Solution of Example 1 44
5.2 Optimality Test of Example 1 50
Chapter 6 : Results and Analysis 73

6.1 Results 73
6.2 Analysis of the solution 75
Chapter 7 : Conclusion 79
References 80
iii
CHapter
Introduction 1
The transportation problem (TP) is a special class of linear programming problems,
which deals with the transportation of a single homogeneous product from several
sources (production or supply centers) to several sinks (destinations or warehouses) [15].
When addressing a TP, the practitioner usually has a given capacity at each supply point
and a given requirement at each demand point. The main objective of TP is to determine
the amounts transported from each source to each sink to minimize the total
transportation cost while satisfying the supply and demand restrictions. The basic steps
for obtaining an optimum solution to a TP are [16]:
Step 1: Mathematical formulation of the transportation problem;
Step 2: Determining an initial basic feasible solution
Step 3: To test whether the solution is an optimal one. If not, to improve it further
till the optimality is achieved.
In this study, we have focused on Step 2 above in order to obtain efficient initial basic
feasible solutions for the transportation problem. Several heuristics viz. Row Minimum
Method (RMM), Column Minimum Method (CMM), Northwest Corner Method
(NWCM), Least Cost Method (LCM), Extremum Difference Method (EDM) and
Vogel’s Approximation Method (VAM) have been developed to find an initial basic
feasible solution to a transportation problem. Although some heuristics can find an initial
basic feasible solution quickly, often the solution they find is not especially good in
terms of minimizing total cost. Other heuristics may not find an initial basic feasible
solution as quickly, but the solution they find is often good in terms of minimizing total
costs [19].
Analysis and Re-solution of the Transportation Problem: A Linear Programming Approach. Page 9 of 91
INTRODUCTION CHAPTER ONE
VAM [3] is based on the concept of penalty cost. A penalty cost is the difference
between the lowest cell cost in the row or column and next to lowest cell cost in the same
row or column. It then makes allocation as much as possible to the minimum cost cell in
the row or column with the largest penalty cost.
In case of EDM [13] a penalty cost is the difference between the highest cell cost in the
row or column and the lowest cell cost in the same row or column. It then makes
allocation as much as possible to the minimum cost cell in the row or column with the
largest penalty cost.
In [3] Serdar Korukoğlu and Serkan Balli proposed Improved VAM by introducing total
opportunity cost (TOC) matrix. The TOC matrix is obtained by adding the ‘row
opportunity cost matrix’ (for each row, the smallest cost of that row is subtracted from
each element of the same row) and the ‘column opportunity cost matrix’ (for each
column, the smallest cost of that column is subtracted from each element of the same
column). It then calculates penalty costs (difference between the lowest cell cost in the
row or column and next to lowest cell cost in the same row or column) and makes
allocation as much as possible to the cell having lowest unit transportation cost among
the row or column containing highest three penalty costs.
In this thesis, we calculate the pointer costs by taking difference of highest cost and next
smaller to the highest cost for each row and each column and consider highest three
pointer costs. We then make allocation to the cell having lowest unit transportation cost
among the row and column containing highest three pointer costs. If more than one cell
contain lowest-cost, we allocate to the cell where allocation is maximum. We illustrate
the proposed method numerically and ordeal the optimality of the presented method by
Modified Distribution method and Stepping Stone method [21]. As a final point, an
analysis to identify the grounds of different solutions by different methods is also carried
out.
1.1 Background of the study
One of the earliest and most abundant applications of linear programming techniques has
been the formulation and solution of the transportation problems as a linear
programming problem.
The standard form of the problem was first formulated, along with a constructive
solution, by Frank L. Hitchcock in 1941, when he presented a study entitled ‘The
Distribution of a Product from several sources to Numerous Localities’ [10]. This paper
sketched out the partial theory of a technique foreshadowing the simplex method; it did
not exploit special properties of a transportation problem except in finding starting
solutions.
In 1947, another investigator T.C. Koopmans, as a member of the Combined Shipping

Board during World War II, also presented a study called ‘Optimum Utilization of the
Transportation System’ [12]. This historic paper was based on his wartime experience.
Because of this and the work done earlier by Hitchcock, the classical case is often
referred to as the Hitchcock Koopmans Transportation Problem.
In 1963, linear programming formulation and the associated systematic method of

solution were first developed by G.B. Dantzig [17].
1.2 Why we choose Transportation Problem
At this moment what is the main problem of our country? It is high cost of commodity.
For this purpose we analysis the total production cost of a product and we get two types
of cost [28]:
(i) Fixed cost and
(ii) Variable cost.
In fixed cost we have Salaries of Staff, Rent of building, Depreciation of machines and
equipment, Interest on capital invested and Insurance.
In Variable cost we have Materials Cost, Transportation Cost, Fuel Cost, Packing Cost,
Maintenance Cost, Advertisement Cost and Commission to Salesman.
Table 1.1: Total Production Cost Analysis
Total Production Cost

Fixed Cost Variable Cost
Salaries of Staff Materials Cost
Rent of building Transportation Cost
Depreciation of machines and Commission to Salesman
equipment
Interest on capital invested Packing Cost
Insurance Maintenance Cost
Advertisement Cost
Fuel Cost
Among these transportation cost has a great affect to the production cost. If we can
reduce the total transportation cost while satisfying the demand and supply restriction, it
will be possible to supply product with comparably low cost. As a result the price of
commodity will reduce by maintaining the quality of the product.
1.3 Structure of the study
The overall review of this thesis introduces an efficient algorithm for finding an initial
basic feasible solution of a transportation problem. This dissertation includes Seven
Chapters. Each Chapter provides the sequential ideas from beginning to end are as
follows:
 In Chapter two- Transportation problem which describes, also formulates the

transportation problem and its basic integrants.
 Chapter three presents detailed algorithm of existing solution procedure for

finding an initial basic feasible solution of a transportation problem and also
provides the algorithm to test the optimality of the solution.
 Chapter four introduces the proposed algorithm along with flowchart and C
program to solve transportation problem.
 In Chapter five-The proposed algorithm and the other existing algorithms are
illustrated by means of three numerical examples.
 Chapter Six includes the Results and Analysis where we present the results and
make out the explanation of different solutions by different methods.
 Chapter Seven which includes the conclusion presents the summary of our
achievement in this dissertation.
Finally this dissertation includes the References of different authors and links from
where we collect the important notes and technique to discuss in this dissertation to
achieve our maximum goal of minimizing total transportation cost of a transportation
problem.
CHapter
Transportation Problem 2
2.1 Definition
The transportation problem is a special class of linear programming problems, which

deals with the transportation of a single homogeneous product from several sources
(production or supply centers) to several sinks (destinations or warehouses). When
addressing a Transportation problem, the practitioner usually has a given capacity at each
supply point and a given requirement at each demand point. Many decision problems,
such as production inventory, job scheduling, production distribution and investment
analysis can be formulated as Transportation Problems. Good financial decisions
concerning facility location also attempt to minimize total transportation and production
costs for the entire system. The transportation problem can be represented as a single
objective transportation problem or as a multi-objective transportation problem.
2.2 Main Objectives
The main objective of Transportation Problem is to determine the amounts transported

from each source to each sink to minimize the total transportation cost while satisfying
the supply and demand restrictions.
2.3 Basic Steps
The basic steps for obtaining an optimum solution to a Transportation Problem are:
Step 1: Mathematical formulation of the transportation problem;
Step 2: Determining an initial basic feasible solution;
TRANSPORTATION PROBLEM CHAPTER TWO
Step 3: To test whether the solution is an optimal one. If not, to improve it further
till the optimality is achieved.
2.4 Mathematical Formulation
The transportation problem can be stated as an allocation problem in which there are m
sources (suppliers) and n destinations (customers). Each of the m sources can allocate to
any of the n destinations at a per unit carrying cost c ij (unit transportation cost from
source i to destination j). Each sources has a supply of S i units, and each
destination has a demand of d j units, . The objective is to determine which
routes are to be opened and the size of the shipment on those routes, so that the total
transportation cost of meeting demand, given the supply constraints, is minimized.
2.4.1 The decision variables
A transportation problem is a complete specification of how many units of the product

should be transported from each source to each destination. Therefore, the decision
variables are:
The amount of the shipment from source to destination ,
where and
2.4.2 The objective function
Consider the shipment from warehouse to destination. For any and any, the
transportation cost per unit is ; and the size of the shipment is . Since we assume the cost
function is linear, the total cost of this shipment is given by . Summing

over all and all now yields the overall transportation cost for all source-destination
combinations.
That is, our objective function is:
Minimize:
2.4.3 The constraints
Considering warehouse . The total outgoing shipment from this warehouse is the sum
. In summation notation, this is written as . Since the
total supply from warehouse is , the total outgoing shipment cannot exceed . That is, we
must require
Again, considering destination. The total incoming shipment at this outlet is the sum
. In summation notation, this is written as . Since the
demand at outlet is , the total incoming shipment should not be less than . That is,
we must require
Of course, as physical shipments, the ’s should be non-negative i.e. .

Then the linear programming model representing the transportation problem is generally
given as
Minimize .
subject to ;
. for all and .
In mathematical terms the above problem can be expressed as finding a set of ’s,
; to
Minimize .
subject to ;
. for all and .
2.5 Network Representation

We represent each source and destination by a node, the amount of supply at source by
, the amount of demand at destination by , the unit transportation cost between

source and destination by and the amount transported from source to destination
by . The arcs joining the source and destination represent the route through which
the commodity is transported. Then the general transportation problem can be
represented by the network as in the following figure.
Figure 2.1: Network Representation of Transportation Problem
Source
c11 : x11 Destination
s
s
S1 1
1 d1
Units of supply
S2 2 2 d2 Units of demand
n dn
Sm m
cmn : xmn
2.6 Balanced Transportation Problem
When the total supply amount is equal to the total demand then this transportation
problem is called the Balanced Transportation problem. That is in the balance
transportation problem
where is the total supply amount and is the total demand amount.

2.7 Unbalanced Transportation Problem
An unbalanced transportation problem is one in which the total supply of the sources and
the total demand at destination centre are not equal. This creates the following two
situations:
i. When the total capacity of the origins exceeds the total requirement of
destinations, a dummy destination is introduced in the transportation table
which absorbs the excess capacity. The cost of shipping from each origin to
this dummy destination is assumed to be zero. The insertion of a dummy
destination establishes equality between the total sources capacities and total
destination requirements.
ii. When total capacity of origins is less than the total requirement of
destinations, a dummy origin is introduced in the transportation table to meet
the excess demand. The cost of shifting from the dummy origin to each
destination is assumed to be zero. The introduction of a dummy origin in this
case establishes the equality between the total capacity of sources and the
total requirement of destinations.
2.8 Matrix Terminology
The matrix used in the transportation models consists of squares called “cells”, which
when stacked from ‘columns’ vertically and rows ‘horizontally’. The cell allocated at
the intersection of a row and a column is designated by its row and column heading.
Thus the cell located at the intersection of row A and column 4 is called cell (A, 4). Unit
transportation costs are positioned in each cell. The availabilities of each source are
placed in the last column while the requirements are placed in the last row.

Table 2.1: Matrix Representation of Transportation Problem
Warehouses Supply
Sources
1 2 3 4 5 15
A 2 5 8 1 9 25
B 5 6 8 7 10 20
C 1 2 5 8 2 10
Demand 16 10 18 12 14
2.9 Some Basic Definitions
i. Source
It is a location from which shipment are dispatched.
ii. Destination
It is the location to which shipment are transported.
iii. Unit Transportation Cost

It is the cost of transporting one unit of the consignment from a source to a
destination.
iv. Feasible solution

A set of non-negative values , ; that satisfies the
constraints is called a feasible solution to the transportation problem.
v. Basic feasible solution

A feasible solution that contains no more than non-negative allocation is called a
basic feasible solution to the transportation problem.

vi. Optimal solution
A feasible solution (not necessarily basic) is said to be optimal if it minimizes the
total transportation cost.
vii. Non-degenerate basic feasible solution

A basic feasible solution to a transportation problem that contains exactly
allocations in independent position.
viii. Degenerate basic feasible solution

A basic feasible solution that contains less than non-negative allocations.
ix. Loop
An ordered set of at least four cells in transportation table is said to be a loop
provided
a) Any two adjacent cells of the ordered set lie either in the same row or in
the same column.
b) No three or more adjacent cells in the ordered set lie in the same row or
column.
A loop is formed Do not form a loop

x. Pointer Cost
Difference of highest cell cost and next smaller to the highest cell cost for each
row and column. In case of row, we call it row pointer and for column, it is
column pointer.
xi. Independent and Non-independent allocation

If the occupied cells do not make a closed loop, then the allocations are said to be
independent. Again a loop may not involve all the allocation. Such allocation
with which a loop can be formed are known as non-independent allocation.
xii. Algorithm
A step-by-step problem-solving procedure, especially an established, recursive
computational course of action for solving a problem in a finite number of steps
is called an Algorithm.
CHapter
Existing Solution Procedure 3

The transportation problem has been formulated by various investigators and solved to
various degrees. The systematic method of solution was first given by Dantzig. There are
many methods of determining an initial basic feasible solution. Well-known heuristics
are North West Corner Method, Row Minimum Method, Column Minimum Method,
Least Cost Method, Vogel’s Approximation Method and Extremum Difference Method.
These methods differ in the quality of the starting basic solution they produce and better
starting solution yields a smaller objective value. Detailed algorithms of these methods
are given below:
3.1 Row Minimum Method (RMM)
Row Minimum Method is used when the purpose of completing the warehouse No. 1 and
then the next. Row minimum cost is useful in small number of supply and when the cost
of transportation on supply. The cost of transportation is less than North West Corner
Method. Detailed process is given below:
Step 1 : Consider the first row and identify the lowest cost cell of the
first row.
Step 2 : Allocate as many units as possible equal to the minimum

between available supply and demand so that either the
capacity of the first supply is exhausted or the demand of the
jth distribution center is satisfied or both. Then three cases
arises:
EXISTING SOLUTION PROCEDURE CHAPTER THREE

Case 1 : If the capacity of the first supply is satisfied, cross off the
first row and move below to the second row.
Case 2 : If the demand of the jth distribution center is satisfied,

cross off the jth column and reconsider the first row with
the remaining demand.
Case 3 : If the capacity of the first supply as well as the demand of

the jth distribution center is completely satisfied, make a
zero allocation in the second lowest cost cell of the first
row. Cross off the row as well as the jth column and
move below to the second row.
Step 3 : Continue the process until all rows and column are satisfied.
3.2 Column Minimum Method (CMM)
Column Minimum Method is used when the purpose of completing the demand No. 1
and then the next. Column minimum cost is useful in small number of demand and when
the cost of transportation on demand. The cost of transportation is less than North West
Corner Method. The standard instructions are paraphrased below:
Step 1 : Consider the first column and identify the lowest cost cell of the
first column.
Case 1 : If the demand of the first distribution center is satisfied,
cross off the first column and move right to the second
between available supply and demand so that either the demand
column.
of the first distribution center is satisfied or the capacity of the
Case ith
2 supply is exhausted
: If the capacity oforith
both. Thenisthree
supply casescross
satisfied, arises:off the ith
row and reconsider the first column with the remaining
demand.
Case 3 : If the demand of the first distribution center as well as the

capacity of the ith supply is completely satisfied, make a
Page 16 of 91
Analysis and Re-solution of the Transportation Problem: A Linear Programming Approach.
zero allocation in the second lowest cost cell of the first
EXISTING SOLUTION PROCEDURE
column. Cross off the column as well as theCHAPTER THREE
ith row and
move right to the second column.
Step 3 : Continue the process until all columns are satisfied.

3.3 North West Corner Method (NWCM)
A particularly simple method of determining an initial basic feasible solution of a

transportation problem is the so-called North West Corner Method, introduced by
Charnes and Cooper, where the basic variables are selected from the North West Corner
(i.e. top left corner). The standard North West corner rule instructions are paraphrased
below:
Step 1 : Select the north west (upper left hand) corner cell of the
transportation table and allocate as many units as possible equal
to the minimum between available supply and demand.
Step 2 : Adjust the supply and demand numbers in the respective rows
and columns.

Step 3 : If the demand for the first cell is satisfied, then move
horizontally to the next cell in the second row.
Step 4 : If the supply for the first row is exhausted, then move down to
the first cell in the second row.
Step 5 : If for any cell, supply equals demand, then the next allocation
can be made in either in the next row or column.
Step 6 : Continue the process until all supply and demand values are
exhausted.
North west corner method provides quick solution because computations take short time
but yields a bad solution because it is very far from optimal solution.
3.4 Least Cost Method (LCM)
The heuristics we describe for finding an initial basic feasible solution to a transportation
problem is called the least cost method. It is also known as Matrix minima Method or
Minimum Cost method or Best Cell Method. In this method, the basic variables are
chosen according to the unit transportation cost. This heuristics strikes a compromise
between finding a feasible solution quickly and finding a feasible solution that is optimal
or very close to the optimal solution. Detailed processes are given below:
Step 1 : Identify the box having minimum unit transportation cost in the
cost matrix of the transportation table. If a tie occurs, choose
any one of them randomly.

between available supply and demand.
Step 3 : Adjust the supply and demand numbers in the respective rows
and columns.
Step 4 : No further consideration is required for the row or column

which is satisfied. If both the row and column are satisfied at a
time, delete only one of the two, and the remaining row or
column is assigned zero supply (or demand).
Step 5 : Continue the process for the resulting reduced transportation

table until all the rim conditions are satisfied.
3.5 Vogel’s Approximation Method (VAM)
The Vogel Approximation (Unit penalty) method is an iterative procedure for computing
an initial basic feasible solution of a transportation problem. This method is preferred
over the three methods i.e. North West Corner Rule, Row minimum Method and Column
Minimum Method, because the initial basic feasible solution obtained by this method is
either optimal or very close to the optimal solution. The standard instructions are
paraphrased below:
Step 1 : If either (total supply > total demand) or (total supply < total
demand), balance the transportation problem.
Step 2 : Determine the penalty cost for each row by taking difference of
lowest cell cost in the row and next to lowest cell cost in the

same row and put in front of the row on the right. In a similar
fashion, calculate the penalty cost for each column and write
them in the bottom of the cost matrix below corresponding
columns.
Step 3 : Choose the highest penalty costs and observe the row or
column to which this corresponds. If a tie occurs, choose any
one of them randomly.
Step 4 : Make allocation min (Si,dj) to the cell having lowest unit
transportation cost in the selected row or column.
Step 5 : No further consideration is required for the row or column

time, delete only one of the two, and the remaining row or
column is assigned zero supply (or demand).
Step 6 : Calculate fresh penalty costs for the remaining sub-matrix as in

Step 2 and allocate following the procedure of Steps 3, 4 and 5.
Continue the process until all rows and columns are satisfied.
Step 7 Compute total transportation cost for the feasible allocations

using the original balanced transportation cost matrix.
3.6 Extremum Difference Method (EDM)
This technique is applicable to find an initial basic feasible solution of a transportation

problem which is also very much nearer to optimal solution and in most of the case it
yields directly the optimal solution. The following steps summarize the approach.

Step 1 : Determine the penalty cost for each row by taking difference of
highest cell cost in the row and lowest cell cost in the same row
and put in front of the row on the right. These numbers are
called row penalties. In a similar fashion, calculate the column
penalties for each column and write them in the bottom of the
cost matrix below corresponding columns.
Step 2 : Choose the highest penalty costs and observe the row or
column to which this corresponds. If a tie occurs, choose any
one of them randomly.
Step 3 : Make allocation min (Si,dj) to the cell having lowest unit
transportation cost in the selected row or column.
Step 4 : Ignore the row and column, for further consideration, which is
satisfied. If both the row and column are satisfied at a time,
delete only one of the two, and the remaining row or column is
assigned zero supply (or demand).
Step 5 : Calculate fresh penalty costs for the remaining sub-matrix as in

Step 6 : Continue the process until all rows and columns are satisfied.
Step 7 Compute total transportation cost for the feasible allocations

3.7 Optimality Test
Make an optimality test to find whether the obtained feasible solution is optimal or not.
An optimality test can, of course, be performed only on that feasible solution in which
a. number of allocation is m+n-1, where m is the number of rows and n is the

number of columns.
b. these m+n-1 allocation should be in independent positions.
Now test procedure for optimality involves examination of each vacant cell to find
whether or not making an allocation in it reduces the total transportation cost. The two
methods used for this purpose are the Stepping-Stone Method and The Modified
distribution Method (MODI) or u-v method.
3.7.1 Stepping Stone Method
It is a method for finding the optimum solution of a transportation problem. The detailed
process is given here:
Step 1 : Determine an initial basic feasible solution using any one of the
following method:
a. Row Minimum Method;

b. Column Minimum Method;
c. North West Corner Method;
d. Least Cost Method;
e. Vogel’s Approximation Method;
f. Extremum Difference Method;
g. Proposed Method.
Step 2 : Make sure that the number of occupied cells is exactly equal to
m+n-1, where is the number of rows and is the number of
columns.

Step 3 : Select an unoccupied cell. Beginning at this cell, trace a closed
path, starting from the selected unoccupied cell until finally
returning to that same unoccupied cell.
Step 4 : Assign plus (+) and minus (-) signs alternatively on each corner
cell of the closed path just traced, beginning with the plus sign
at unoccupied cell to be evaluated.
Step 5 Add the unit transportation costs associated with each of the
cell traced in the closed path. This will give net change in terms
of cost.
Step 6 Repeat steps 3 to 5 until all unoccupied cells are evaluated.
Step 7 : Check the sign of each of the net change in the unit
transportation costs. If all the net changes computed are greater
than or equal to zero, stop; an optimal solution has been
reached. Otherwise, go to Step 8.
Step 8 : Select the unoccupied cell having the most negative net cost
change and determine the maximum number of units that can
be assigned to this cell. The smallest value with a negative
position on the closed path indicates the number of unit that can
be shipped to the entering cell. Add this number to the
unoccupied cell and to all other cells on the path marked with a
plus sign. Subtract this number from cells on the closed path
marked with a minus sign.
Step 9 Go to Step 3.

3.7.2 The Modified Distribution Method (MODI) or u-v method
The Modified Distribution Method also known as MODI method or u-v method
provides a minimum cost solution to the transportation problem. In the Stepping Stone
Method, we have to draw as many closed paths as equal to the unoccupied cells for their
evaluation. To the contrary, in MODI method, only closed path for the unoccupied cell
with most negative opportunity cost is drawn. The method, in outline, is:
Step 1 : Determine an initial basic feasible solution using any one of the
following method:
a. Row Minimum Method;

b. Column Minimum Method;
c. North West Corner Method;
d. Least Cost Method;
e. Vogel’s Approximation Method;
f. Extremum Difference Method;
g. Proposed Method.
Step 2 : Make sure that the number of occupied cells is exactly equal to
m+n-1, where is the number of rows and is the number of
columns.
Step 3 Compute the row indices, and column indices .
Step 4 Determine the opportunity cost for non basic variable using
.
Step 5 : If for all unoccupied cells, Stop; the solution is optimal.
Otherwise, Go to Step 6.

Step 6 : Identify the unoccupied cell with the negative opportunity cost
and select it as the incoming cell. If more than one unoccupied
cell contains negative opportunity cost, select the cell with
most negative opportunity cost.
Step 7 : Draw a closed loop or stepping stone path associated with the
incoming cell.
Step 8 : Assign an alternate plus and minus signs on the corner points of
the closed path with a plus sign at the cell selected as the
incoming cell.
Step 9 The smallest value among the cell with a minus sign on the
closed path indicates the number of units that can be shipped to
the incoming cell. Now, add this quantity to all the cells on the
corner points of the closed path marked with plus signs, and
subtract it from those cells marked with minus signs. If there is
a tie, select any one of the tied cell. The tied cells that are not
selected will be artificially occupied with a zero allocation.
Step 10 Go to Step 3.
CHapter
Proposed Optimization Technique 4

4.1 Algorithm
Here, a new efficient algorithm is developed by introducing pointer cost. Pointer cost is
the difference of highest cell cost and next smaller to the highest cell cost for each row
and each column. First we calculate the pointer cost, then we consider highest three
pointer costs and allocate as much as possible to the feasible cell with the lowest
transportation cost. The detailed process is given below:
Step 1 : If either ( total supply > total demand ) or ( total supply <
total demand), balance the given TP.
Step 2 : Determine the pointer cost for each row by taking difference of
highest cell cost in the row and next smaller to the highest cell
cost in the same row and put in front of the row on the right. In
a similar fashion, calculate the pointer cost for each column
and write them in the bottom of the cost matrix below
corresponding columns.
Step 3 : Choose the highest three pointer costs and observe the row or
column to which this corresponds.
Step 4 : Make allocation min to the cell having lowest unit
PROPOSED OPTIMIZATION TECHNIQUE CHAPTER FOUR

transportation cost among the row and column containing
highest three pointer costs. If more than one cell contains
lowest-cost, allocate to the cell where allocation is maximum.
Step 5 No further consideration is required for the row or column

time, delete the row or column which contains comparably
more unit transportation cost between the two.
Step 6 : Calculate fresh pointer costs for the remaining sub-matrix as in

Step 7 : Continue the process until all rows and columns are satisfied.
Step 8 : Compute total transportation cost for the feasible allocations

This technique is applicable to find an initial basic feasible solution of a transportation

problem which is very close to optimal solution and in most of the case it yields directly
the optimal solution.

4.2 Flowchart
Start
Read ci j ; Supply, Si ; Demand, dj

demand
Calculate total supply & total demand
Yes No
Is total supply = total demand
?
Balance the TP
Determine the pointer cost for each row and for each column
?
Choose the highest? three pointer costs
k=1 ; Allocation[0]=0 ; Total Cost = 0
Allocate Min(Si,dj) to the cell having lowest unit transportation cost

among the row and column containing highest three pointer costs.
If more than one cell contains lowest-cost, allocate to the cell where
allocation is maximum.
Allocation[k]= Min(Si , dj)
Total_Allocation= Allocation[k-1]+ Allocation[k]
7 5
t t
a a
r r
t t
7 5
t t
a a
r r cij
Cost[k]=
t t
No Yes
Is allocation = demand
?
?
Delete corresponding row Delete corresponding column
dj = d j - Si Si = Si - dj
Total Cost=Total Cost+ Allocation[k] × Cost[k]
k=k+1
No Is Total_Allocation = Total Supply Yes
Print “Total Cost”
Stop
4.3 C Program
#include <stdio.h>
#include <stdlib.h>
int supply[10], demand[10];
int supply_check[10], demand_check[10] ;
int cost[10][10], diff[20][2];
int min1, row1,col1, min2, row2,col2, min3, row3,col3;
int minX,rX,cX, allocation[20][20], noOfAllocation;
int sortmax (int values[], int size );
int sortmin (int values[], int size );
int comparemax (const void * a, const void * b);
int sortminmax (int values[], int size);
int findRowCol(int num);
int findRowCol(int num);
int calculatemin(int,int, int,int);
int calculateTotalCost();
int dataread();
int row, col, row_copy, col_copy;
int operation();
int min1st, min2nd, min3rd, max1st,max2nd, max3rd;
int main ()
dataread();
row_copy=row;
col_copy=col;
noOfAllocation=0;
operation();
return 0;
int operation()
int i,j, rowcol, val[10], min,r,c, k;
int tcost_supply, tcost_demand,row_supply,col_demand;
for(k=0;k<(row+col-2);k++)
{ /* make difference for row*/
for(i=0;i<row;i++)
for(j=0;j<col;j++)
val[j]=cost[i][j];
sortminmax(val,col);
if(val[col-2]==-999)
diff[i][0]=val[col-1];
else
diff[i][0]=val[col-1]-val[col-2];
diff[i][1]=i;
for(i=0;i<col;i++) /*make difference for column*/
for(j=0;j<row;j++)
val[j]=cost[j][i];
sortminmax(val,row);
if(val[row-2]==-999)
diff[row+i][0]=val[row-1];
else
diff[row+i][0]=val[row-1]-val[row-2];
diff[row+i][1]=i+100;
for(i=0;i<row+col;i++) /*Copy the difference values and sorting*/
val[i]=diff[i][0];
sortminmax(val,row+col);
max1st=val[row+col-1];
max2nd=val[row+col-2];
max3rd=val[row+col-3];
rowcol=findRowCol(max1st);
calculatemin(rowcol,min,r,c);
min1=minX;
row1=rX;
col1=cX;
rowcol=findRowCol(max2nd);
min2=minX;
row2=rX;
col2=cX;
rowcol=findRowCol(max3rd);
min3=minX;
row3=rX;
col3=cX;
if(min1 <= min2 && min1<= min3)
if (supply[row1] < demand[col1]) /*/Allocation*/
allocation[noOfAllocation][0]=supply[row1];
allocation[noOfAllocation][1]=min1;
demand[col1]=demand[col1]-supply[row1];
supply[row1]=0;
noOfAllocation++;
else
allocation[noOfAllocation][0]=demand[col1];
supply[row1]=supply[row1]-demand[col1];
demand[col1]=0;
noOfAllocation++;
else if (min2 <= min1 && min2<= min3)
supply[row2]=0;
noOfAllocation++;
else
demand[col2]=0;
noOfAllocation++;
else if (min3 <= min1 && min3<= min1)
supply[row3]=0;
noOfAllocation++;
else
demand[col3]=0;
noOfAllocation++;
tcost_supply=0; /*forming new matrix ; find Supply*/
row_supply=-999;
col_demand=-999;
for(i=0;i<row;i++)
if(supply[i]==0 )
row_supply=i;
for(j=0;j<col;j++)
if(cost[i][j]!=-999)
tcost_supply+=cost[i][j];
break;
tcost_demand=0; /*column*/
for(i=0;i<col;i++)
if(demand[i]==0 )
col_demand=i;
for(j=0;j<row;j++)
if(cost[j][i]!=-999)
tcost_demand+=cost[j][i];
break;
if(row_supply!=-999 && col_demand!=-999)
demand_check[col_demand]=0;
supply_check[row_supply]=0;
if(tcost_demand > tcost_supply)
for(i=0;i<row;i++)
cost[i][col_demand]=-999;
demand[col_demand]=-999;
else
for(i=0;i<col;i++)
cost[row_supply][i]=-999;
supply[row_supply]=-999;
else if (row_supply!=-999)
for(i=0;i<col;i++)
cost[row_supply][i]=-999;
supply[row_supply]=-999;
supply_check[row_supply]=0;
else if (col_demand!=-999)
for(i=0;i<row;i++)
cost[i][col_demand]=-999;
demand[col_demand]=-999;
demand_check[col_demand]=0;
calculateTotalCost();
return 0;
int calculateTotalCost()
int i,j;
long int total=0;
for(i=0;i<row;i++)
for(j=0;j<col;j++)
if (cost[i][j]!=-999)
total=total+cost[i][j]*supply[i];
for (i=0;i<noOfAllocation;i++)
printf("\n%d %d",allocation[i][0],allocation[i][1]);
total+=allocation[i][0]*allocation[i][1];
printf("\nMinimum Total cost = %ld", total);
return 0;
int findRowCol(int num)
int i, rowcol;
for(i=0;i<row+col;i++)
if(num==diff[i][0])
diff[i][0]=0;
rowcol=diff[i][1];
break;
return rowcol;
int calculatemin(int rowcol,int min,int r,int c)
int i;
if(rowcol<100)
min=abs(cost[rowcol][0]);
r=rowcol;
c=0;
for(i=1;i<col;i++)
if(abs(min)>abs(cost[rowcol][i]) && cost[rowcol][i]!=-999)
min=cost[rowcol][i];
c=i;
else
min=abs(cost[0][rowcol-100]);
r=0;
c=rowcol-100;
for(i=1;i<row;i++)
if(abs(min)>abs(cost[i][rowcol-100]))
min=cost[i][rowcol-100];
r=i;
minX=min;
rX=r;
cX=c;
return 0;
int dataread()
int i=0;
int j=0;
FILE *fp;
fp=fopen("input.txt","r"); /*indata is like cin*/
fscanf(fp,"%d %d",&row,&col);
i=0;
while (i<row ) { /* keep reading until end-of-file*/
j=0;
while(j<col)
fscanf(fp,"%d",&cost[i][j]);
j++;
i++;
i=0; /*Read Supply*/
while(i<row)
fscanf(fp,"%d",&supply[i]);
supply_check[i]=1;
i++;
j=0; /*read Demand*/
while(j<col)
fscanf(fp,"%d",&demand[j]);
demand_check[j]=1;
j++;
fclose(fp);
return 0;
int comparemax (const void * a, const void * b)
return ( *(int*)a - *(int*)b );
int sortminmax (int value[], int size)
qsort (value, size, sizeof(int), comparemax);
/*x1=value[col-1];
x2=value[col-2];
diff=x1-x2;*/
/*for (n=0; n<6; n++)
printf ("%d ",values[n]);*/
return 0;
CHapter
Numerical Illustration 5
The proposed method and the other existing methods for finding an initial basic feasible
solution are illustrated by the following examples.
5.1 Example 1: Consider the following transportation problem:
Table 5.1: Cost Matrix for the Numerical example
Destination
1 2 3 4 Supply
1 6 1 9 3 70
Factory
2 11 5 2 8 55
3 10 12 4 7 90
Demand 85 35 50 45
Applying the algorithm of the Row Minimum Method (RMM), we obtain the following
allocations:
Table 5.2: Initial Basic Feasible Solution using Row Minimum Method
1 2 3 4 Supply
1 35 35 70
6 1 9 3
2 50 5 55
11 5 2 8
3 85 5 90
10 12 4 7
Demand 85 35 50 45
NUMERICAL ILLUSTRATION CHAPTER
FIVE
Therefore, the solution for the given problem is
, , , , and .
for a flow of 215 units with the total transportation cost
Applying the algorithm of the Column Minimum Method (CMM), we obtain the
following allocations:
Table 5.3: Initial Basic Feasible Solution using Column Minimum Method
1 2 3 4 Supply
1 70 70
6 1 9 3
2 35 20 55
11 5 2 8
3 15 30 45 90
10 12 4 7
Demand 85 35 50 45
, , , , and .
FIVE
Applying the algorithm of the North West Corner Method (NWCM), we obtain the
Table 5.4: Initial Basic Feasible Solution using North West Corner Method
1 2 3 4 Supply
1 70 70
6 1 9 3
2 15 35 5 55
11 5 2 8
3 45 45 90
10 12 4 7
Demand 85 35 50 45
, , , , and .
Applying the algorithm of the Least Cost Method (LCM), we obtain the following
allocations:
Table 5.5: Initial Basic Feasible Solution using Least Cost Method
1 2 3 4 Supply
1 35 35 70
6 1 9 3
2 5 50 55
11 5 2 8
3 80 10 90
10 12 4 7
Demand 85 35 50 45
FIVE
, , , , and .
Applying the algorithm of the Vogel’s Approximation Method (VAM), we obtain the
Table 5.6: Initial Basic Feasible Solution using Vogel’s Approximation Method
1 2 3 4 Supply Row Pointers
1 70 70 (2) -- --
6 1 9 3
2 35 20 55 (3) (3) (6)
11 5 2 8
3 15 30 45 90 (3) (3) (3)
10 12 4 7
Demand 85 35 50 45
(4) (4) (2) (4)
Column Pointers
(1) (7) (2) (1)

(1) -- (2) (1)
, , , , and .
FIVE
Applying the algorithm of the Extremum Difference Method (EDM), we obtain the
Table 5.7: Initial Basic Feasible Solution using Extremum Difference Method
1 35 35 70 (8) (6) (3) --

6 1 9 3
2 5 50 55 (9) (9) (3) (3)
11 5 2 8
3 80 10 90 (8) (6) (3) (3)
10 12 4 7
Demand 85 35 50 45
(5) (11) (7) (5)
Column Pointers
(5) -- (7) (5)

(5) -- -- (5)
(1) -- -- (1)
, , , , and .
FIVE
Applying the algorithm of the Proposed Method, we obtain the following allocations:
Table 5.8: Initial Basic Feasible Solution using Proposed Method
1 35 35 70 (3) (3) (3) --

6 1 9 3
2 5 50 55 (3) (3) (3) (3)
11 5 2 8
3 80 10 90 (2) (3) (3) (3)
10 12 4 7
Demand 85 35 50 45
(1) (7) (5) (1)
Column Pointers
(1) -- (5) (1)

(1) -- -- (1)
(1) -- -- (1)
, , , , and .
FIVE
5.2 Optimality Test of Example 1
Taking initial Basic Feasible Solution due to proposed method, we now proceed for
optimality using Modified Distribution Method. Here we calculate and for
occupied basic cells using and assuming :
Calculation of opportunity cost for non basic variable using :
(1 , 1)
(1 , 3)
(2 , 2)
(2 , 4)
(3 , 2)
(3 , 3)
Since the opportunity costs of the unoccupied cells are not all positive, the basic feasible
solution obtained by the proposed method is not optimal. The value indicates
that the total transportation cost can be reduced by shifting the allocation to this cell.
FIVE
Table 5.9: Improved Solution- Non Optimum
1234Supply135-35+70006 1129325-+5055511-15208380+10-
90410712347Demand8535504561-33
Again, calculation of and for basic variables using and assuming :
Table 5.10: Revised Solution- Optimum
1 2 3 4 Supply
1 30 40 70 0
6 1 9 3
2 5 50 55 4
11 5 2 8
3 85 5 90 4
10 12 4 7
Demand 85 35 50 45
6 1 -2 3
FIVE
(1 , 1)
(1 , 3)
(2 , 1)
(2 , 4)
(3 , 2)
(3 , 3)
Since all opportunity costs are positive, the basic feasible solution obtained by the
proposed method is an optimal solution.
Warehouse
1 2 3 4 5 6 Supply
1 7 10 7 4 7 8 5
2 5 1 5 5 3 3 6
Factory
3 4 3 7 9 1 9 2
4 4 6 9 0 0 8 9
Demand 4 4 6 2 4 2

FIVE
allocations:
1 2 3 4 5 6 Supply
1 3 2 5
7 10 7 4 7 8
2 4 0 2 6
5 1 5 5 3 3
3 2 2
4 3 7 9 1 9
4 4 3 2 9
4 6 9 0 0 8
Demand 4 4 6 2 4 2
, , , , ,
, , and .


FIVE
1 2 3 4 5 6 Suppl
1 4 1 5y
7 10 7 4 7 8
2 4 2 6
5 1 5 5 3 3
3 2 2
4 3 7 9 1 9
4 2 2 4 1 9
4 6 9 0 0 8
Demand 4 4 6 2 4 2
, , , , ,
, , and .


FIVE
Applying the algorithm of the North West Corner Method (NWCM), we obtain the
Table 5.14: Initial Basic Feasible Solution using North West Corner Method
1 2 3 4 5 6 Suppl
1 4 1 1 5y
7 10 7 4 7 8
2 3 3 6
5 1 5 5 3 3
3 2 2 2
4 3 7 9 1 9
4 1 2 4 2 9
4 6 9 0 0 8
Demand 4 4 6 2 4 2
, , , , ,
, , and .


FIVE
allocations:
1 2 3 4 5 6 Supply
1 5 5
7 10 7 4 7 8
2 4 0 2 6
5 1 5 5 3 3
3 1 1 2
4 3 7 9 1 9
4 3 2 4 9
4 6 9 0 0 8
Demand 4 4 6 2 4 2
, , , , ,
, , and .


FIVE
1 2 3 4 5 6 Supply Row Pointers
1 5 5 (3) (3) (0) (0) (0) (0)(0)

7 10 7 4 7 8
2 4 2 6 (2) (1) (2) (4) -- -- --

5 1 5 5 3 3
3 1 0 1 2 (2) (2) (2) (1) (1) (3)(3)

4 3 7 9 1 9
4 3 2 4 9 (0) (0) (4) (2) (2) (5) --

4 6 9 0 0 8
Demand 4 4 6 2 4 2
(0) (2) (2) (4) (1) (5)
(0) (2) (2) (4) (1) --
Column Pointers
(0) (2) (2) -- (1) --

(0) (2) (2) -- -- --
(0) (3) (0) -- -- --
(0) -- (0) -- -- --
(3) -- (0) -- -- --
, , , , ,
, , and .


FIVE
1 5 0 5 (6) (6) (3) (1) (1) (1)(1)

7 10 7 4 7 8
2 4 2 6 (4) (4) (4) (2) -- -- --

5 1 5 5 3 3
3 1 1 2 (8) (6) (6) (5) (5) (5)(2)

4 3 7 9 1 9
4 3 2 4 9 (9) (9) (5) (5) (5) -- --

4 6 9 0 0 8
Demand 4 4 6 2 4 2
(3) (9) (4) (9) (7) (6)
(3) (9) (4) (9) -- (6)
Column Pointers
(3) (9) (4) -- -- (6)

(3) -- (4) -- -- (6)
(3) -- (2) -- -- (1)
(3) -- (0) -- -- (1)
-- -- (0) -- -- (1)
, , , , ,
, , and .


FIVE
1 5 5 (2) (2) (2) (1) (0) (0)

7 10 7 4 7 8
2 4 0 2 6 (2) (2) (2) (2) (0) (0)

5 1 5 5 3 3
3 1 1 2 (2) (2) (2) (2) (3) (3)

4 3 7 9 1 9
4 3 2 4 9 (1) (1) (1) (1) (5) --

4 6 9 0 0 8
Demand 4 4 6 2 4 2
(2) (4) (2) (4) (4) (1)
(2) (4) (2) (4) -- (1)
Column Pointers
(2) (4) (2) -- -- (1)

(2) -- (2) -- -- (1)
(2) -- (2) -- -- --
(2) -- (2) -- -- --
, , , , ,
, , and .


FIVE

We now examine the effect on the transportation costs for each of the unoccupied cells
by Stepping Stone method. The closed path for each of these cells is given below:
+5- 710744051551-1 + 4379324690 +5- 710744 -0 +5 15511

4379324690
Cell (1,1) Cell (1,2)
5+ 710 - 744051 551-1 + 43 793+2 5 +710-747405155311-43+791324 -

- 4690 +46900
Cell (1,4) Cell (1,5)

5 -+ 747802 -+ 55331 7919249008 5 71074+4 0 - 5 15511 -
43+79324690
Cell (1,6) Cell (2,1)


FIVE
5 710 - 7440-+51 551-1 + 43 793+2 5 +71074740+51 -55311-43+791324

- 4690 +4690 -0
Cell (2,4) Cell (2,5)
5 710 - 744-0+51 551+1 - 43 7932 5710744051 5511 +- 437 932+ 469 -0

4690
Cell (3,2) Cell (3,4)
5 +7107474051 55311+ -43791324 +4690 5 74 780+2- 55 331 - 79 1+9242

-0 9008
Cell (3,5) Cell (3,6)


FIVE
5 710 - 744-0+51 5511 + 43 - 5710744051 5511- + + 437 93+2 - 469 0

793+2 - 4690
Cell (4,2) Cell (4,3)
5710747840251 + 553 -311 + 43 - 7919324 - 46900 +8
Cell (4,6)
The net changes in costs, together with their associate closed loops, are as follows:
Table 5.19 : Net Changes in Costs

Unoccupied
Closed path Net change in cost
Cell

FIVE
Unoccupied Closed path Net change in cost

Cell
Since all the unoccupied cells in the above table is associated with a non-negative value,
the total transportation costs will either remain the same or increase if we made a shift of
the unit to any one of the unoccupied cells. Hence the preceding Table provides an
optimum solution.

Destination
1 2 3 4 Supply
1 10 2 20 11 15
Source
2 12 7 9 20 25
3 4 14 16 18 10
Demand 5 15 15 15

FIVE
allocations:
1 2 3 4 Supply
1 0 15 15
10 2 20 11
2 5 15 5 25
12 7 9 20
3 10 10
4 14 16 18
Demand 5 15 15 15
, , , , and .


FIVE
1 2 3 4 Supply
1 15 15
10 2 20 11
2 0 15 10 25
12 7 9 20
3 5 5 10
4 14 16 18
Demand 5 15 15 15
, , , , and .
1 2 3 4 Supply
1 5 10 15
10 2 20 11
Applying the algorithm of the North 5West Corner
15 Method
5 (NWCM), we obtain the
2 25
following allocations: 12 7 9 20
3 10 10
Table 5.23: Initial Basic Feasible
4
Solution
14
using
16
North18
West Corner Method
Demand 5 15 15 15

FIVE
, , , , and .
allocations:
1 2 3 4 Supply
1 15 15
10 2 20 11
2 0 15 10 25
12 7 9 20
3 5 5 10
4 14 16 18
Demand 5 15 15 15
, , , , and .

FIVE

FIVE
1 15 15 (8) (9) -- --
10 2 20 11
2 0 15 10 25 (2) (2) (2) (11)
12 7 9 20
3 5 5 10 (10) (2) (2) (2)
4 14 16 18
Demand 5 15 15 15
(6) (5) (7) (7)
Column Pointers
-- (5) (7) (7)

-- (7) (7) (2)
-- -- (7) (2)

, , , , and .

FIVE
1 15 0 15 (18) (10) (9)

10 2 20 11
2 15 10 25 (13) (11) (11)
12 7 9 20
3 5 5 10 (14) (14) (2)
4 14 16 18
Demand 5 15 15 15
(8) (12) (11) (9)
Column Pointers
(8) -- (11) (9)

-- -- (11) (9)
, , , , and .

FIVE
1 15 15 (9) -- -- --
10 2 20 11
2 0 15 10 25 (8) (8) (11) (11)
12 7 9 20
3 5 5 10 (2) (2) (2) (2)
4 14 16 18
Demand 5 15 15 15
(2) (7) (4) (2)
Column Pointers
(8) (7) (7) (2)

-- (7) (7) (2)
-- -- (7) (2)
, , , , and .

FIVE
Taking initial Basic Feasible Solution due to proposed method, we now proceed for
optimality using Modified Distribution Method. Here we calculate and for
occupied basic cells using and assuming :
(1 , 1)
Calculation of opportunity
(1 , 3) cost for non basic variable using :
(1 , 4)
(2 , 1)
(3 , 2)
(3 , 3)
Since the opportunity costs of the unoccupied cells are not all positive, the basic feasible
solution obtained by the proposed method is not optimal. The value indicates
that the total transportation cost can be reduced by shifting the allocation to this cell.

FIVE
Table 5.28: Improved Solution- Non Optimum
1234Supply115- + 15 091021620-4112 0+1510-255612

7920355103491491618Demand515151512415
Again, calculation of and for basic variables using and assuming :

Table 5.29: Revised Solution- Optimum
1 2 3 4 Supply
1 5 10 15 0
10 2 20 11
2 10 15 25 5
12 7 9 20
3 5 5 10 7
4 14 16 18
Demand 5 15 15 15
-3 2 4 11

FIVE
(1 , 1)
(1 , 3)
(2 , 1)
(2 , 4)
(3 , 2)
(3 , 3)
Since all opportunity costs are positive, the basic feasible solution obtained by the
Proposed method is an optimal solution.

CHapter
Results and Analysis 6

6.1 Results
The following table and graphical representation shows a comparison among the
solutions obtained by our proposed method and by Total
the other existing methods
Transportation Cost and also
Solution
with the optimal obtained
solution by of three sample examples.
by means Example 1 Example 2 Example 3
Row Minimum Method 1165 84 505
Table
Column 6.1: A Method
Minimum Comparative study of different solutions
1220 74 475
Northwest Corner Method 1265 95 520
Least Cost Method 1165 68 475
Vogel’s Approximation Method 1220 68 475
Extremum Difference Method 1165 68 475
Proposed Method 1165 68 475
Optimum Solution 1160 68 435
RESULTS AND ANALYSIS CHAPTER SIX
Figure 6.1: Graphical Representation of Different Solution of Example 1

6.2 Analysis of the Solution
In Table 6.1 we see that (i) Initial basic feasible solution varies for different solution
procedure and also for different example. (ii) Our proposed method, LCM and EDM
gives better result in comparison to other methods.
So we are trying to find out:
 Why the initial basic feasible solution varies in different methods of solution?
 Is there any unique method that will calculate the optimal solution of all TP?
For this purpose, we write down the details solution of each example. We arrange the
unit transportation cost in ascending order and inscribe the corresponding allocation in
the next three tables.
Table 6.2: Analysis of the Solution (Example 1)
PROPOS
RMM CMM NWCM LCM VAM EDM
ED
Allo. Cost Allo. Cost Allo. Cost Allo. Cost Allo. Cost Allo. Cost Allo. Cost
35 1 20 2 5 2 35 1 20 2 35 1 35 1
50 2 30 4 45 4 50 2 30 4 50 2 50 2
35 3 35 5 35 5 35 3 35 5 35 3 35 3
5 7 70 6 70 6 10 7 70 6 10 7 10 7
5 8 45 7 45 7 80 10 45 7 80 10 80 10
85 10 15 10 15 11 5 11 15 10 5 11 5 PROPOS
11
RMM CMM NWCM LCM VAM EDM ED
2 0 4 0 4 0 4 0 4 0 4 0 4 0
4 1 2 0 2 0 2 0 2 0 2 0 2 0
2 1 4 Table
1 6.3:
3 Analysis
1 4of the1Solution
4 1 4
(Example 2)1 4 1
2 3 2 4 3 5 2 3 2 3 2 3 2 3
0 3 2 4 4 7 3 4 0 3 3 4 3 4
4 4 2 5 2 7 1 4 3 4 1 4 1 4
2 4 4 7 2 8 0 5 1 4 5 7 0 5
3 7 1 8 1 9 5 7 5 7 1 7 5 7
3 9 1 8 1 10 1 7 1 7 0 8 1 7
Table 6.4: Analysis of the Solution (Example 3)
PROPOS
RMM CMM NWCM LCM VAM EDM ED
15 2 15 2 10 2 15 2 15 2 15 2 15 2
15 9 5 4 5 7 5 4 5 4 5 4 5 4
0 10 0 7 15 9 0 7 0 7 15 9 0 7
5 12 15 9 5 10 15 9 15 9 0 11 15 9
10 18 5 18 10 18 5 18 5 18 5 18 5 18
5 20 10 20 5 20 10 20 10 20 10 20 10 20
From the above table, we see that North West Corner Method allocate to the cell without
considering the unit transportation cost and therefore yields a bad solution because it is
very far from optimal solution.
Row Minimum Method and Column Minimum Method consider unit transportation cost
row and column wise respectively and so sometimes they yields best starting solution
and sometimes not. The cost of transportation using these two methods is less than North
West Corner Method.
Least Cost Method, Vogel’s Approximation Method, Extremum Difference Method and
Our proposed Method consider unit transportation cost and so they yields better starting
solution and most often they gives directly the optimal solution.
Analyzing the solution we make out the following case that will endow with better
starting solution.
Case 1: Allocation to the cell with least transportation cost.
Case 2: Sometimes allocation to the cell next to least transportation cost.
We present a dummy solution here two realize the second case:
Table 6.5: Dummy Solution
Solution 1 Solution 2
Allo. Cost Allo. Cost
5 1 5 1
3 2 3 2
4 3 3
4 8 4
5 5 4 5
8 6 5 6
Total Cost 96 Total Cost 92
In the above table we see that solution 1 allocate 4 units corresponding to unit
transportation cost 3 while solution 2 avoid transportation cost 3 and it allocate next to
least cost i.e. 4 and in this case the allocation is 8 units. This makes a rearrange in the
allocation to the higher unit transportation cost. As a result we see that the total
transportation cost for solution 1 is 96 while for solution 2 is 92.
At last, we conclude that Least Cost Method and our Proposed Method follow the case 1
while the VAM and EDM follow the case 2 and provide better initial basic feasible
solution. Still there is no unique method that will calculate the optimal solution of all
transportation problems.
CHapter
Conclusion 7
In this dissertation, a new algorithm along with a C program for finding an initial basic
feasible solution of a transportation problem is introduced. We also illustrate this method
numerically and test the efficiency of the proposed method. The optimal solution is
calculated by using ‘TORA Optimization System’ software. A comparative study among
the solution obtained by proposed method and the other existing methods is also
presented by means of sample problems and it is seen that in some of the cases, the
initial transportation cost is the same as the cost obtained by LCM and EDM, sometimes,
it is better than VAM and most often the presented method gives directly the optimal
solution. An analysis to identify the reason of different solutions by different methods is

as well carried out. Finally we claim that our proposed method is equivalent to well
known heuristics LCM and EDM but better than VAM, RMM, CMM and NWCM.
References
[1] Aminur Rahman Khan, 2011 ‘A Re-solution of the Transportation Problem:

An Algorithmic Approach’ accepted for publication in Jahangirnagar
University Journal of Science.
[2] Dr. M Sharif Uddin, Sayedul Anam, Abdur Rashid and Aminur R. Khan,
2011 ‘Minimization of Transportation Cost by Developing an Efficient
Network Model’ accepted for publication in Journal of Mathematics and
Mathematical Sciences.
[3] Serdar Korukoğlu and Serkan Balli, 2011 ‘An Improved Vogel’s
Approximation Method for the transportation problem’ Mathematical and
Computational Applications, Vol. 16, No. 2, pp. 370-381.
[4] P. Pandian, G. Natarajan, 2010 ‘A New Approach for solving Transportation

Problems with Mixed Constraints’ Journal of Physical Sciences, Vol. 14, pp.
53-61.
[5] Abdur Rashid, 2011 ‘An Effective Approach for Solving Transportation
Problems’ accepted for publication in Journal of Mathematics and
Mathematical Sciences.
[6] Md. Ashraful Babu: M.S. Thesis 2009, ‘Numerical Simulation of

Transshipment Model: An Approach to Minimize the Total Transshipment
Cost’ (Dept. of Mathematics, Jahangirnagar University).
[7] Sayedul Anam: M.S. Thesis 2007, ‘On Development of Network Model: An
Approach to Minimize Transportation Cost’ (Dept. of Mathematics,
Jahangirnagar University).
[8] M. Mathirajan and B. Meenakshi, 2004 ‘Experimental Analysis of Some
Variants of Vogel’s Approximation Method’ Asia-Pacific Journal of
Operational Research 21(4), pp. 447-462.
REFERENCES
[9] J. E. Reeb and S. Leavengood, 2002 ‘Transportation Problem: A Special case

for Linear Programming’ Oregon: Oregon State Universit y
Extension Service Publications EM 8779.
[10] Frank L. Hitchcock, 1941 ‘The distribution of a Product from Several

Sources to Numerous Localities’, J. Math. Phys. pp. 224-230.
[11] W.L. Winston, 1991, Introduction to Mathematical Programming-

Applications and Algorithms, Duxbury Press, California.
[12] T.C. Koopmans, 1947 ‘Optimum Utilization of the Transportation System’,

Econometrica, Vol XVII.
[13] H.S. Kasana and K.D. Kumar, 2005, Introductory Operations Research:
Theory and Applications, Springer International Edition, New Delhi.
[14] David G. Luenberger and Yinyu Ye, ‘Linear and Nonlinear Programming’, 3 rd
Edition, Springer, New York, pp. 145-159.
[15] Hamdy A. Taha, 2004 “Operations Research: An Introduction” Prentice
Hall of India, New Delhi, pp.165,181-183.
[16] P.K. Gupta and Man Mohan, 2003, Problems in Operation Research, Sultan
Chand & Sons, New Delhi, pp. 275-276.
[17] G.B. Dantzig, 1963, Linear Programming and Extentions, Princeton

University Press, Princeton, N J.
[18] Frederick S. Hiller and Gerald J. Lieberman, ‘Introduction to Operations

Research’, Seventh Edition, Tata McGraw-Hill Publishing Company Limited,
New Delhi, pp. 350-381.
[19] David R. Anderson, Dennis J. Sweeney and Thomas A. Williams, ‘An

Introduction to Management Science, 11th Edition, pp. 307-379.
[20] Robert J. Childress, Robin D. Gorsky and Richard M. Witt, Mathematics For
Managerial decisions’, 2nd Edition, pp. 498-527.
REFERENCES
[21] Charnes, A. and W.W. Cooper, 1954 ‘The Steeping Stone Method of
Explaining Linear Programming Calculations in Transportation Problems’,
Management Science, 1,1.
[22] Taghrid Imam, Gaber Elsharawy, Mohamed Gomah and Iman Samy, 2009
‘Solving Transportation Problem Using Object-Oriented Model’ International
Journal of Computer Science and Network Security, Vol. 9, Number 2, pp.
353-361.
[23] Nabendu Sen, Tanmoy Som and Banashri Sinha, 2010 ‘A Study of
Transportation Problem for an Essential Item of Southern Part of North
Eastern Region of India as an OR Model and Use of Object-Oriented
Programming’ International Journal of Computer Science and Network
Security, Vol. 10, Number 4, pp. 78-86.
[24] Shiv Kant Kumar, Indu Bhushan Lal and S. P. Varma, 2011 ‘ An Alternative
Method For Obtaining Initial feasible Solution to a Transportation Problem
and Test For Optimality’ International Journal of Computer Science and
Communication, Vol. 2, Number 2, pp. 455-457.
[25] Ramesh Gupta: Master of Engineering Report, 1972 ‘A Fast Algorithm For
Solution of Transportation Problems’ (The Faculty of Engineering, Sir George
Williams University, Montreal, Canada).
[26] Md. Amirul Islam, Aminur Rahman Khan, M. Sharif Uddin and M. Abdul
Malek, 2012 ‘Determination Of Basic Feasible Solution Of Transportation
Problem: A New Approach’ submitted for publication in Jahangirnagar
University Journal of Sciences.
[27] Veena Adlakha and Krzysztof Kowalski, 2009 ‘Alternate Solutions

Analysis For Transportation Problems’ Journal of Business &
Economics Research, Volume 7, Number 11 , pp. 41-49.
[28] S.C Sharma,1999, Production Management, Khanna Pub., Delhi, pp.344-347.

Thesis Aminur

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

Thesis Aminur

Uploaded by

Copyright:

Available Formats

Analysis and Re-solution of the Transportation

Problem: A Linear Programming Approach

A dissertation submitted in partial fulfillment of the requirements for the degree of

I would like to express my profound sense of gratitude and sincere thanks to my

valuable remarks and encouragement during the course of this work.

making an inspirational environment for me which help me to continue my thesis work

Aminur Rahman Khan

Place: Jahangirnagar University, Savar, Dhaka. Aminur Rahman Khan

Dated: April 1, 2012 M. Phil Student

It is to be mentioned that he presented his first seminar on 24 September, 2011 and

Place: Jahangirnagar University, Savar, Dhaka. Dr. Md. Sharif Uddin

Dated: April 1, 2012 Supervisor & Associate Professor

dissertation is to present an efficient approach for finding an initial basic feasible

solution of the transportation problem (TP). We also build up a C program of the

presented algorithm. The proposed method is illustrated with several numerical

examples. An analysis to identify the reason of different solutions by different methods

presented herein gives a better result.

TOPICS PAGE NO.

Chapter 2 : Transportation Problem 6

Chapter 3 : Existing Solution Procedure 15

Chapter 4 : Proposed Optimization Technique 26

Chapter 5 : Numerical Illustration 44

Chapter 6 : Results and Analysis 73

Step 1: Mathematical formulation of the transportation problem;

Step 2: Determining an initial basic feasible solution

till the optimality is achieved.

1.1 Background of the study

In 1947, another investigator T.C. Koopmans, as a member of the Combined Shipping

In 1963, linear programming formulation and the associated systematic method of

1.2 Why we choose Transportation Problem

(i) Fixed cost and

(ii) Variable cost.

Table 1.1: Total Production Cost Analysis

Total Production Cost

1.3 Structure of the study

 In Chapter two- Transportation problem which describes, also formulates the

 Chapter three presents detailed algorithm of existing solution procedure for

The transportation problem is a special class of linear programming problems, which

2.2 Main Objectives

The main objective of Transportation Problem is to determine the amounts transported

2.3 Basic Steps

Step 1: Mathematical formulation of the transportation problem;

Step 2: Determining an initial basic feasible solution;

till the optimality is achieved.

2.4 Mathematical Formulation

2.4.1 The decision variables

A transportation problem is a complete specification of how many units of the product

The amount of the shipment from source to destination ,

2.4.2 The objective function

TRANSPORTATION PROBLEM CHAPTER TWO

That is, our objective function is:

2.4.3 The constraints

. In summation notation, this is written as . Since the

. In summation notation, this is written as . Since the

Of course, as physical shipments, the ’s should be non-negative i.e. .

TRANSPORTATION PROBLEM CHAPTER TWO

. for all and .

. for all and .

2.5 Network Representation

TRANSPORTATION PROBLEM CHAPTER TWO

Figure 2.1: Network Representation of Transportation Problem

2.6 Balanced Transportation Problem