OPERATIONS RESEARCH II

NETWORK MODELS: The Transportation Model contd.

In our last class, we

In this class, we shall

 Finished up the example of computation we were treating in before the class

 Took examples of unbalanced transportation problem and solved

In this class, we shall

 Introduce the Transshipment variant of the Transportation problem

 Raised some exercises
 Discuss and give examples of the Transshipment problem
 Discuss the solution technique of the transshipment model
 Solve an example of the transshipment model
 Discuss the characteristics of the assignment problem
 Treat the solution procedure of the assignment problem
 Solve simple assignment problems

So, at the end of this class, each student will be able to:

 Fully characterize a transshipment problem/model

 Explain the procedure for solving the transshipment problem
 Solve any given transshipment problem
 Fully characterize the assignment problem
 Explain the procedure for solving assignment problem
 Solve any given assignment problem

9.0 The Transshipment Problem

The transshipment model is a variant of the transportation model. The essential difference between the
Transportation model and the Transshipment model is the in the transshipment problem, some or all of
transportation from the sources are not transported directly to the destinations but through fellow sources
or other independent locations and or through one destination location to another.

The following are examples of graphically illustrated transshipment models.

Page 1 of 21
CASE 1

Pure Source Nodes

Pure Destination Nodes
Transshipment Nodes
70 1 Nodes
12
170 140
12 6
7 4

8
100 2 8 14
3
11 17
5 7 160
130
130 3 15

CASE 2

Pure Source Nodes

Transshipment Nodes NodesPure Destination Nodes

70 1
13
1704 140
14 6
10 4

10 8
100 2 14 18 17
7
10
5 7 160
12 130
130 3

Page 2 of 21
 Pure Source Nodes

Transshipment Nodes Nodes

1 Pure Destination Nodes
70 18

170 7 6 120
15 4

6
10 10
100 2 8
8

11 14
3 5 7 180
130
130 3

Pure Source Nodes

Transshipment Nodes Nodes

1 Pure Destination Nodes
70 18

170 7 6 120
15 4
9
6
10 10
100 2 8
8

11 14
3 5 7 180
130
130 3

Transshipment problems are solved by first converting them to equivalent transportation problem and
then solving them with transportation technique.

Page 3 of 21
In the conversion every Pure Source node and Pure destination are treated as before while every
transshipment node is seen as both a source and destination (supply and demand)

From example in Case 1,

Pure Source Nodes

Pure Destination Nodes
Transshipment Nodes Nodes
70 1
12
170 140
12 6
7 4

8
100 2 8 14
3
11 17
5 7 160
130
130 3 15

We have,

3 Pure Source Nodes 1, 2, 3

2 Transshipment Nodes 4, 5

2 Pure Destination Nodes 6, 7

Although there are 7 nodes, every transshipment node will constitute 2 nodes.

So, in converting to Transportation equivalent there will be: 3 + 2 = 5 Source nodes and 2 + 2 = 4
destination nodes.

When set on the transportation table connections that are not possible are given very prohibitively high
transportation cost value of M.

Let us set up the table for the Case 1:

Page 4 of 21
 Destinations
Transshipment Pure Supply
4 5 6 7
1 12 7 M M 70
Pure

2 8 3 M M 100
Sources

3 11 15 M M 130
4 M M 12 14 170
Trans

5 M M 8 17 130
Demand 170 130 140 160
The transportation costs are indicated on the cells.

In this problem, we can adopt M = 100 because it is very much higher than any of the given costs.

Solution

Using Least Cost Method,

Destinations
Transshipment Pure Supply
4 5 6 7
12 7 100 100
1 70, 40, 0
40 30
2 8 3 100 100
100, 0
100
3 11 15 100 100
130, 0
130
100 100 12 14
4 170, 160, 0
10 160
5 100 100 8 17
130, 0
130
Deman 140, 10,
170, 40, 0 130, 30, 0 160, 0
d 0

Page 5 of 21
The resulting starting solution is:

Destinations
Transshipment Pure Supply
4 5 6 7
12 7 100 100 70, 40,0
1
40 30
8 3 100 100
2 100,0
100
11 15 100 100
3 130,0
130
100 100 12 14
4 170,160,0
10 160
100 100 8 17
5 130,0
130

Deman 160,
170,40,0 130,30,0 140,10,0
d 0

The Total cost is: 12×40 + 7×30 + 3×100 + 11×130 +100×0 + 12×10 + 14×160 + 8×130 = N5,820

Iteration 1

Use the Multiplier Method

The

The current basic variables are: x 11 , x 12 , x 22 , x 31 , x 41 , x 43 , x 44 , x53

Use ui + v j=cij ; u1=0

Basic Variable (u, v) Equation Solution

x 11 u1 + v 1=c 11 0+ v 1=12⟹ v 1=12
x 12 u1 + v 2=c 12 0+ v 2=7 ⟹ v 2=7
x 22 u2 + v 2=c 22 u2 +7=3⟹u2=−4
x 31 u3 + v 1=c 31 u3 +12=11⟹u3=−1

Page 6 of 21
 x 41 u 4+ v 1=c 41 u 4+ 12=100⟹u 4=88
x 43 u 4+ v 3=c 43 88+ v 3=12⟹ v 3=−76
x 44 u 4+ v 4 =c 44 88+ v 4 =14⟹ v 4=−74
x 53 u5 + v 3=c 53 u5 +−76=8⟹u5=84

Destinations
Transshipment Pure
Supply
4 5 6 7
v 1=12 v 2=7 v 3=−76 v 4=90
1 12 7 100 100
70
u1=0 40 30
2 8 3 100 100
100
u2=−4 100
3 11 15 100 100
130
u3=−1 130
4 100 100 12 14
170
u 4=88 0 10 160

5 100 100 8 17
130
u5=84 130
Demand 170 130 140 160

u1=0 ,u2 =−4 ,u3 =−1 ,u 4=88 ,u5 =84

v 1=12, v 2=7 , v 3=−76 , v 4=−74

The Non-Basic Variable positions are: x 13 , x 14 , x 21 , x 23 , x 24 , x 32 , x 33 , x 34 , x 42 , x 51 , x 52, x 54

Non-Basic Variable
x 13 u1 + v 3−c 13=0−76−100=−176
x 14 u1 + v 4−c14 =0+90−100=−10
x 21 u2 + v 1−c 21=−4 +12−8=0

Page 7 of 21
 x 23 u2 + v 3−c 23=−4−76−100=−180
x 24 u2 + v 4−c24 =−4 +90−100=−14
x 32 u3 + v 2−c 32=−1+7−15=−9
x 33 u3 + v 3−c 33=−1−76−100=−177
x 34 u3 + v 4−c34 =−1+90−100=−11
x 42 u 4+ v 2−c 42=88+7−100=−5
x 51 u5 + v 1−c 51=84+12−100=−4
x 52 u5 + v 2−c 52=84+7−100=−9
x 54 u5 + v 4−c54 =84−74−17=−7
NO VALUE IS POSITIVE.

Hence, Optimality reached.

Optimal Transportation Problem solution is

Destinations
Transshipmen
Pure Supply
t
4 5 6 7
12 7 100 100
1 70
40 30
8 3 100 100
2 100
100
11 15 100 100
3 130
130
100 100 12 14
4 170
0 10 160
100 100 8 17
5 130
130
Demand 170 130 140 160

Page 8 of 21
The solution can be displayed on the original transshipment diagram

Pure Source Nodes

Transshipment Nodes Nodes

70 1 Pure Destination Nodes
12
40 170
4
12 6 140
30 7 10
160
8
100 2 130 8 14
100 3 130
11 17
5 7 160
130
130 3 15

Exercise

Convert the Transshipment Cases 2 and 3 above to equivalent transportation models and solve them.

9.2 Assignment Problems

The assignment problem is a problem in which a number of workers or assignees, say n are assigned to n
assignments, tasks, jobs or duties based on pre-assessed competence or skill levels of the workers or
assignees at doing different assignments.

Job
1 2 … n

Page 9 of 21
 Worke 1 c 11 c 11 … c 11 1
r 2 c 11 c 11 … c 11 1
⋮ ⋮ ⋮ ⋮… ⋮ ⋮
n c 11 c 11 … c 11 1
1 1 … 1
An assignment problem, just like the regular transportation problem can be represented as a Linear
Programming problem.

Let c ij =cost of assigning a worker i¿ job j; i=1 , 2 ,. . . ,n ; j=1 ,2 , . . ., n

Also, let

x ij ={ ¿ job j¿0 , otherwise ¿

The Linear Programme for the Assignment problem is:

n n
Minimize ∑ ∑ c ij x ij =c 11 x11 + c 12 x 12+. . .+c 1 n x 1 n+ c 21 x 21+ c22 x 22 +. ..+ c 2n x 2 n+ .. .+c n 1 x 1 n +c n 2 x n 2 +. .. ,+ cnn x nn
i=1 j=1

subject to:

∑ x ij=1 ; for all i=1 ,2 , . . ., n

j=1

∑ x ij=1 ; for all j=1 , 2 , .. . , n

i=1

x ij =0∨1 ; i=1 , 2 ,. . . , n; j=1 , 2, . . ., n

The Assignment model, though a linear programming problem which can be solved by the usual Simplex
Method, has its own unique characteristics which make it amenable to use of easier variant or modified
form of Simplex method to solve. One good characteristics of the assignment model is the fact that:

The optimal solution to the Assignment Linear Programme remains unchanged if a constant is added or
subtracted from any row or column of the cost matrix c ij

Suppose we want to obtain a new cost coefficient c 'ij by subtracting a constant pi from row I and a

constant q j from column j such that

'
c ij =c ij − pi−q j

Page 10 of 21
 ( ) ( )
n n n n n n n n n n n n n n

∑ ∑ c 'ij x ij= ( c ij −p i−q j ) x ij =∑ ∑ c ij x ij −∑ pi ∑ x ij −∑ q j ∑ x ij =∑ ∑ c ij x ij −∑ pi (1 )−∑ q j ( 1 )=∑ ∑

i=1 j=1 i=1 j=1 i=1 j=1 j =1 i=1 i=1 j=1 i=1 j=1 i=1 j=1

Because the objective function differs from the original one by a constant, the optimum values of x ij must
be the same in both cases.

Coupled with this, all sum of supplies on each row and all sum of demands on each column each equals 1.

This are the key bases of the method of solution of Assignment problems called the Hungarian Method

The Hungarian Method

Given the Assignment Model,

Job
1 2 … n
1 c 11 c 11 … c 11 1
Worke 2 c 11 c 11 … c 11 1
r ⋮ ⋮ ⋮ ⋮… ⋮ ⋮
n c 11 c 11 … c 11 1
1 1 … 1
Step 1: From the original cost matrix, identify each row’s minimum and subtract it from all the entries of
the row.

Step 2: Using the matrix resulting from Step 1, identify each columns minimum entry and subtract it
from each entry of the column.

Step 3: Identify the optimal solution as feasible assignment associated with the locations of zero elements
of the last matrix obtained in Step 2 or Step 4. If a unique optimal solution (with ties broken arbitrarily)
exists, STOP. Otherwise go to Step 4 or go back to Step 4.

Step 4: If no feasible assignment (with zero entries) can be obtained in Step 3, do the following

i. Draw the minimum number of horizontal and vertical lines in the last reduced matrix that will
cover all zero entries.
ii. Select the smallest uncovered entry and subtract it from all uncovered entries and adding to
every entry at all the intersections of the covering lines.

We shall illustrate the use of the Hungarian Method with examples.

Page 11 of 21
Example 1:

Five jobs are to be assigned to five workers. If the costs (in thousands of Naira) of assignment are as in
the cost/assignment model below:

Job
Job Job5
Job 1 2 Job 3 Job 4
Worker
3 8 2 10 3 1
1
Worker
8 7 2 9 7 1
Worker 2
Worker
6 4 2 7 5 1
3
Worker
8 4 2 3 5 1
4
Worker
7 10 6 9 10 1
5
1 1 1 1 1
Step 1: From the original cost matrix, identify each row’s minimum and subtract it from all the entries of
the row.

The minimum entries are 2, 2, 2, 2, 6 on rows 1, 2,3,4,5 respectively. Perform the operation

Job
Job Job5
Job 1 2 Job 3 Job 4
Worker
1 6 0 8 1 1
1
Worker
6 5 0 7 5 1
Worker 2
Worker
4 2 0 5 3 1
3
Worker
6 2 0 1 3 1
4
Worker 1 4 0 3 4 1

Page 12 of 21
 5
1 1 1 1 1

Step 2: Using the matrix resulting from Step 1, identify each columns minimum entry and subtract it
from each entry of the column.

The minimum column entries are 1, 2, 0, 1, 1 for columns 1, 2, 3, 4, 5 columns respectively. Perform the
operation

Job
Job Job5
Job 1 2 Job 3 Job 4
Worker
0 4 0 7 0 1
1
Worker
5 3 0 6 4 1
Worker 2
Worker
3 0 0 4 2 1
3
Worker
5 0 0 0 2 1
4
Worker
0 2 0 2 3 1
5
1 1 1 1 1

Assignment

Assign Job 5 to Worker 1

Assign Job 3 to Worker 2

Assign Job 2 to Worker 3

Assign Job 4 to Worker 4

Assign Job 1 to Worker 5

Cost of assignment Use the original cost matrix.

Page 13 of 21
 Job
Job Job5
Job 1 2 Job 3 Job 4
Worker
3 8 2 10 3 1
1
Worker
8 7 2 9 7 1
Worker 2
Worker
6 4 2 7 5 1
3
Worker
8 4 2 3 5 1
4
Worker
7 10 6 9 10 1
5
1 1 1 1 1

Cost of Assignment = 3+2+4+3+7 thousands of Naira = N19,000

Example 2:

Four multi-skilled technical personnel are to be assigned the task of operating four different machines for
plainning, grinding, milling and welding. The Industrial team in the Company has assessed the possible
duration (tens of minutes) it takes each proposed operator working with each machine to the company on
the same piece of work and under same working conditions. The result is as displayed on the
cost/assignment matrix below:

Job
Weldin
Worker Plaining Grinding Milling g
Technician
1 4 6 3 1
1
Technician 9 7 10 9 1

Page 14 of 21
 2
Technician
4 5 11 7 1
3
Technician
8 7 8 5 1
4
1 1 1 1
Determine the optimal assignment of job to the workers.

Solution

Note: No worker can be assigned more than one job and no job can be assigned to more than one worker.

Using the Hungarian Method,

Step 1: From the original cost matrix, identify each row’s minimum and subtract it from all the entries of
the row.

The minimums are 1, 7, 4 and 5 for Rows 1, 2, 3 and 4 respectively. Perform the operation to obtain:

Job
Weldin
Plaining Grinding Milling g
Technician
Worker 0 3 5 2 1
1
Technician
2 0 3 2 1
2
Technician
0 1 7 3 1
3
Technician
3 2 3 0 1
4
1 1 1 1

Page 15 of 21
Step 2:Using the matrix resulting from Step 1, identify each columns minimum entry and subtract it from
each entry of the column.

For the columns the minimums are 0, 0, 3, 0 for columns 1, 2, 3 and 4 respectively. Perform the
operations on the last matrix obtained in Step 1.

Job
Weldin
Plaining Grinding Milling g
Technician
0 3 2 2 1
1
Technician
2 0 0 2 1
Worker 2
Technician
0 1 4 3 1
3
Technician
3 2 0 0 1
4
1 1 1 1
Step 3: Judge the optimal assignment (if it exists now) using the zero entries. Rule: Only one zero can be
selected for a worker and a job: i.e. only one zero can be selected on a row and only one zero can be
selected on a column.

Job
Weldin
Plaining Grinding Milling g
Technician
0 3 2 2 1
1
Technician
2 0 0 2 1
Worker 2
Technician
0 1 4 3 1
3
Technician
3 2 0 0 1
4
1 1 1 1

Page 16 of 21
 Job
Weldin
Plaining Grinding Milling g
Technician
0 3 2 2 1
1
Technician
2 0 0 2 1
Worker 2
Technician
0 1 4 3 1
3
Technician
3 2 0 0 1
4
1 1 1 1

Job
Weldin
Plaining Grinding Milling g
Technician
0 3 2 2 1
1
Technician
2 0 0 2 1
Worker 2
Technician
0 1 4 3 1
3
Technician
3 2 0 0 1
4
1 1 1 1

No possible assignment. So, we go to Step 4.

Step 4: If no feasible assignment (with zero entries) can be obtained in Step 3, do the following

i. Draw the minimum number of horizontal and vertical lines in the last reduced matrix that will
cover all zero entries.

Page 17 of 21
 ii. Select the smallest uncovered entry and subtract it from all uncovered entries and adding to
every entry at all the intersections of the covering lines.

Job
Weldin
Plaining Grinding Milling g
Technician
0 3 2 2 1
1
Technician 0
2 0 2 1
Worker 2
Technician 0
1 4 3 1
3
Technician 0
3 2 0 1
4
1 1 1 1

Job
Weldin
Plaining Grinding Milling g
Technician 0
3 2 2 1
1
Technician 2
0 0 2 1
Worker 2
Technician
0 1 4 3 1
3
Technician 3
2 0 0 1
4
1 1 1 1
The minimum uncovered cell value is 1. So, subtract 1 from all uncovered cells and add to the
intersections of lines

We obtain

Job
Plaining Grinding Milling Weldin

Page 18 of 21
 Worker g
Technician
0 2 1 1 1
1
Technician
3 0 0 2 1
2
Technician
0 0 3 2 1
3
Technician
4 2 0 0 1
4
1 1 1 1
The optimal assignment is

Technician 1 to be assigned to Plaining

Technician 2 to be assigned to Milling

Technician 3 to be assigned to Grinding

Technician 4 to be assigned to Welding

Cost of assignment

Job
Weldin
Worker Plaining Grinding Milling g
Technician 1 4 6 3 1
1

Page 19 of 21
 Technician
9 7 10 9 1
2
Technician
4 5 11 7 1
3
Technician
8 7 8 5 1
4
1 1 1 1
Cost = 1 + 5 + 10 + 5 = 21 tens of minutes = 210 minutes

Exercise

1. Three children are to be assigned 3 household shores. Since the children had expressed desire to
earn some money which they would need during a recreational trip. So, the parents handed paper
to each of them to indicate how much he/she would earn doing any shore. The children gave the
charges as in the cost matrix/assignment model below. In what optimal way should the shores be
assigned to the children and at what cost?

Dish- Car-
Washing Sweeping Washing
Amara 15 10 9 1
Shehu 9 15 10 1
Dami 10 12 8 1
1 1 1

2. Given the following cost matrices and assignment model, assign the Workers to jobs optimally.
i.

Job
5
1 2 3 4
1 3 8 2 10 3 1
Worker 2 8 7 2 9 7 1
3 6 4 2 7 5 1
4 8 4 2 3 5 1

Page 20 of 21
 5 9 10 6 9 10 1
1 1 1 1 1
ii.

Job
1 2 3 4 5
1 3 9 2 3 7 1
2 6 1 5 6 6 1
Worker 1
3 9 4 7 3 1
0
4 2 5 4 2 1 1
5 9 6 2 4 5 1
1 1 1 1 1

Page 21 of 21