NATIONAL UNIVERSITY OF SCIENCE AND TECHNOLOGY

FACULTY OF ENGINEERING
DEPARTMENT OF INDUSTRIAL AND MANUFACTURING
ENGINEERING

NAME: BRIGHT

SURNAME: HUNGWE

STUDENT NUMBER: N0171020J

COURSE: OPERATION RESEARCH

COURSE CODE: 5208

LECTURER: ENG. B MKETO

ASSIGNMENT ONE
QUESTION 1

(a) Describe the main phases of an Operations Research project. [10]

Definition of Problem

This involves the development of a clear and concise statement of the problem at hand. This
gives direction and meaning to other steps. In defining the problem, it is important that the
whole system be examined critically in order to recognize all the areas that could be affected
by any decision taking. It is essential to examine the symptoms and true causes of the
problem when defining the issue. When the problem is difficult to qualify, it may be
necessary to develop specific and measurable objectives that may not solve the real problem.

Construction of a Model
This step involves the construction of a suitable model usually mathematical, which is a
representation of the problem at hand. It might be of a functional nature as in linear
programming or have a logical structure as in simulation and algorithms.

Data Collection
It involves obtaining quantitative data either from existing records or a new survey that fits
well into the constructed model of the problem.

Developing a Solution
This involves the manipulation of the model to arrive at the optimal solution to the problem.
It may require solving some mathematical equations for optimal decisions as in calculus or
linear programming models. It may also be a logical approach or a functional approach which
does not require solving a mathematical equation, such as in queuing theory. The optimal
solution is then determined by some criteria.

Testing the Model and its Solution

This involves determining the accuracy of the data used in the model because inaccurate data
leads to inaccurate solutions. If the model can adequately predict the effect of the changes in
the system, however simple it may be, it is acceptable.

Interpretation of the results or establishment of controls

This involves determining the implication of the solutions to the system. In most cases, a
solution to a problem will result in a kind of action or change in the organization. The
implication of these actions or changes must be determined and analysed before results are
implemented. This analysis that is sensitivity analysis determines how much the solution
changes if there were changes in model or in the input data. Controls are therefore established
so that changes that make a significant difference in solution are recognized and taken into
account.

Implementation of Model
This is the process of incorporating the solution into the system, which is carried out by the
personnel already working in the area and not the operations research team. The solution to
the problem has to be translated to a set of operating instructions which can be handled by the
personnel involved.

(b) Discuss the application of Operations Research Techniques in military operations. [10]

Operation research techniques are a core discipline in military and defence management.
They provide contributions to logistics, supply chain and strategic simulation, while enabling
superior decision. In any war field two or more parties are involved, each having different
resources manpower and ammunition different strategies for application. Every opponent has
to guess the resources with the enemy and his courses of action and accordingly he has to
attack the enemy. For this he needs scientific, logical analysis of the problem to get fruitful
results. Here one can apply the techniques like linear programming, game theory and
inventory models to win the game.

Operation research techniques have been used to solve many sophisticated and complex
defence related problems not only limited to combat operations but also encompassing
logistics, manpower planning, equipment procurement, training, infrastructure defence and
many other areas. The global military community has fully embrace operation research
techniques for ensuring efficiency, readiness and providing innovative ways of solving
military problems. These techniques has grown to include analytics and many applications
including artificial intelligence, cyber security and strategic simulation.

Resource Allocation

Asset allocation on the investment strategy that attempts to balance risk versus reward by
adjusting the percentage of each asset in the investment portfolio according to the investors
risk tolerance goals and time frame. Resource allocation also includes capital budgeting
which is the process the military use to determine which proposed fixed assets purchase it
should accept and which should be declined. This process is used to create a quantitative
view of each proposed fixed asset investment, thereby giving a rational basis for making
decisions. This helps to balance competitive needs and priorities to determine the most
effective course of action in order to maximise the effective use of limited resources and gain
the best return on investments.
SOLUTIONS 2(a)
Let X be wooden coffee tables
Let Y be bookcases
OBJECTIVE FUNCTION

Z = 9X + 12Y
SUBJECT TO:
STRUCTURAL CONSTRAINTS:

X + Y ≤ 10
X + 2Y ≤ 12
NON NEGATIVE CONSTRAINTS

X AND Y ≥ 0

2(b)
Simplex method use equations and not inequalities, so the equations are as follows (adding
slack variables S1 and S2)

Maximise Z = 9X + 12Y +0S1 + 0S2

Subject to: X + Y + S1 ═ 10
X + 2Y + S2 ═ 12
X, Y, S1 and S2 ≥ 0
Initial Simplex Table:

CBi Cj 9 12 0 0 Solution Ratio

Basic X Y S1 S2
Variable
0 S1 1 1 1 0 10 10
0 S2 1 2 0 1 12 6
Zj 0 0 0 0
Cj – Zj 9 12 0 0

 The condition of optimality for a maximisation problem is Cj – Zj ≤ 0

  In this case we still have positive numbers, which are 12 and 9.
 So we proceed and find the Key Column which is the column with the highest
positive value of Cj – Zj
 This is 12. Hence the key column is under Y.
 To find the Ratio, divide the solution column by the key column.
 To get the Key row, we choose the lowest positive ratio value which in this case is -
12/2.
 S2 row is the Key row.
 The key element is 2. So Y is the entering variable and S2 is our leaving variable

Simplex iteration 2
CBi Cj 9 12 0 0 Solution Ratio
Basic X Y S1 S2
Variable
0 S1 0.5 0 1 -0.5 4 8
12 Y 0.5 1 0 0.5 6 12
Zj 6 12 0 6 72
Cj – Zj 3 0 0 -6

 For new values of entering variable, simply divide old values by the key element.
 For the new values of the 1st and 2nd rows, S1 and S2, we use the following formula;
New value = Old value – (Corr. Key Column Value x Corr. New row value of entering
variable)

For S1: 1 – (1x0.5) = 0.5, 1 – (1x1) = 0, 1 – (1x0) = 1, 0-(0.5x1) = -0.5, 10 –

(1x6) = 4

Values of Zj

0.5 × 0 + 12 × 0.5 = 6

1 × 9 + 12 × 0 = 9

0 × 0 + 12 × 1 = 12

-0.5 × 0 + 12 × 0.5 = 6
0 × 4 + 12 × 6 = 72

Compute cj – zj values as shown on the simplex table 2. Since we still have a positive
number, the condition of optimality has not been reached. Therefore we have to proceed to
find the Key Column, Ratio and Key Row and move to iteration 3.

Simplex iteration 3

CBi Cj 9 12 0 0 Solution Ratio

Basic X Y S1 S2
Variable
9 X 1 0 2 -1 8 10
12 Y 0 1 -1 1 2 6
Zj 9 12 6 3 96
Cj – Zj 0 0 -6 -3

For Y: 0.5 – (1x0.5) = 0, 1 – (0x0.5) = 1, 0 – (0.5x2) = -1, 0.5-(0.5x-1) = -1, 6 –

(0.5x8) = 2

Values of Zj

1 × 9 + 12 × 0 = 9

0 × 9 + 12 × 1 = 12

2 × 9 + 12 × -1= 6

-1 × 9 + 12 × 1 = 3

8 × 9 + 12 × 2 = 96

The optimality has been reached at these values, X = 8, Y =2 and Z (opt) = 96

2(c) Maximum profit = $96

SOLUTION 3(a)
Unbalance occurs when both supply and demand are not equal then it is said to be an
unbalanced transportation problem. Either a dummy row or a dummy column is added
according to the requirement to make it a balanced problem. This can be solved using any of
the following methods:

 North West corner cell method

 Least cost cell method
 Vogel’s Approximation Method

3(b)

PROJECT A PROJECT B PROJECT C SUPPLY

PLANT W 5 4 9 120
PLANT X 4 3 5 160
PLANT Y 7 4 2 140
DEMAND 140 200 80 420

PROJECT A PROJECT B PROJECT C SUPPLY

PLANT W 120(4) 120
PLANT X 160(2) 160
PLANT Y 20(5) 40(3) 80(1) 140
DEMAND 140 200 80 420

Transportation cost = 120 x 5 + 3 x 160 + 4 x 40 + 2 x 80 +7 x 20

= $ 1540

Occupied Cells: WA, XB, YA, YB, YC

D (W) + R (A) = 5

D(X) + R (B) = 3

D(Y) + R (A) = 7
D(Y) + R (B) = 4

D(Y) + R(C) = 2

D (W) = 0

D(X) = 1

D(Y) = 2

Unoccupied cells: WB, WC, XA, and XC

D (W) + R (B) = 0 + 2 = 2

D (W) + R(C) = 0

D (X) + R (A) = 1 + 5 = 6

D (X) + R(C) =1 + 0 = 1

Cell Actual cost Shadow cost Result

XA 4 6 -2
WB 4 2 +2
WC 9 0 +9
XC 5 1 +4

 There is a negative on the comparison of the actual and shadow cost. This means that
the total costs could be reduced for every two units that can be transferred into cell
XA. The next iteration is performed.

PROJECT A PROJECT B PROJECT C SUPPLY

PLANT W 120 120
PLANT X 160 160
PLANT Y 20 40 80 140
DEMAND 140 200 80 420

PROJECT A PROJECT B PROJECT C SUPPLY

 PLANT W 120 120
PLANT X 20 140 160
PLANT Y 60 80 140
DEMAND 140 200 80 420

Transport cost = 120 x 5 + 20 x 4 + 140 x 3 + 60 x 4 + 80 x 2

= $ 1500

Shadow Cost: WB, WC, XA, and XC

D (W) + R (B) = 0 + 2 = 2

D (Y) + R (A) = 2 + 5 = 7

D (X) + R (C) = 1 + 0 = 1

D (W) + R(C) =0 + 0 = 0

CELL ACTUAL COST SHADOW COST RESULTS

YA 7 7 0
WB 4 2 +2
WC 9 0 +9
XC 5 1 +4

The answers are positive, therefore no further cost reduction and hence the optimum solution
has been reached with transportation cost of $1500.
SOLUTION 4

A B C
Grinding 1 2 1
Turning 3 1 4
Assembly 6 3 4
Testing 5 4 6

Let the company manufacture a units of A, b units of B and c units of C. Then the inequalities
of the constraints are:

Maximise; Z = 4a + 6b + 5c

Subject to:

a + 2b + c ≤ 30

3a + b + 4c ≤ 60

6a + 3b + 4c ≤ 200

5a + 4b + 6c ≤ 200

a, b and c ≥ 0

SIMPLEX FOR: INTRODUCING SLACK VARIABLES

Maximise: Z = 4a + 6b + 5c + 0S1 + 0S2 + 0S3 + 0S4

Subject to:

a + 2b + c + S1 = 30

3a + b + 4c + S2 = 60

6a + 3b + 4c + S3 = 200

5a + 4b + 6c + S4 = 200

a, b, c, S1, S2, S3, S4 ≥ 0

Simplex Iteration 1

CBj Cj 4 6 5 0 0 0 0 Solutio Ratio

n
Objective A B C S1 S2 S3 S4
variables
0 S1 1 2 1 1 0 0 0 30 15
0 S2 3 1 4 0 1 0 0 60 60
0 S3 6 3 4 0 0 1 0 200 200/3
0 S4 5 4 6 0 0 0 1 200 50
Zj 0 0 0 0 0 0 0
Cj –Zj 4 6 5 0 0 0 0

 The condition of optimality for a maximisation problem is Cj – Zj ≤ 0

 In this case we still have positive numbers, which are 4, 6 and 5.
 So we proceed and find the Key Column which is the column with the highest
positive value of Cj – Zj
 This is 6. Hence the key column is under b.
 To find the Ratio, divide the solution column by the key column.
 To get the Key row, we choose the lowest positive ratio value which in this case is
15.
 S1 row is the Key row.
 The key element is 2. So b is the entering variable and S1 is the leaving variable
SIMPLEX ITERATION 2

CBj Cj 4 6 5 0 0 0 0 Solutio Ratio

n
Objective A B c S1 S2 S3 S4
variables
6 b 0.5 1 0.5 0.5 0 0 0 15 30
0 S2 2.5 0 3.5 -0.5 1 0 0 45 90/7
0 S3 4.5 0 2.5 -1.5 0 1 0 155 62
0 S4 3 0 4 -2 0 0 1 140 35
Zj 3 6 3 3 0 0 0 90
Cj –Zj 1 0 2 -3 0 0 0

S2 NEW VALUES

3 - (1 x 0.5) = 2.5

1 - (1 x 1) = 0

4 - (1 x 0.5) = 3.5

0 - (1 x 0.5) = -0.5

1 - (1 x 0) = 1

0 - (1 x 0) = 0

0 - (1 x 0) = 0

60 - (1 x 15) = 45

S3 NEW VALUES

6 - (3 x 0.5) = 4.5

3 - (3 x 1) = 0

4 - (3 x 0.5) = 2.5

0 - (3 x 0.5) = -1.5
0 - (3 x 0) = 0

1 - (3 x 0) = 1

0 - (3 x 0) = 0

200 - (3 x 15) = 155

S3 NEW VALUES

5 - (4 x 0.5) = 3

4 - (4 x 1) = 0

6 - (4 x 0.5) = 4

0 - (4 x 0.5) = -2

0 - (4 x 0) = 0

0 - (4 x 0) = 0

1 - (4 x 0) = 1

200 - (40 x 15) = 140

Z VALUES:

6 x 0.5 + 0 x 2.5 + 0 x 4.5 + 0 x 3 = 3

6x1+0x0+0x0+0x0=6

6 x 0.5 + 0 x 2.5 + 0 x 3.5 + 0 x 4 = 3

6 x 0.5 + 0 x -1.5 + 0 x -0.5 + 0 x -2 = 3

6x0+0x0+0x1+0x0=0

6x0+0x1+0x0+0x0=0

6x0+0x0+0x0+0x1=0

6 x 15 + 0 x 45 + 0 x 155 + 0 x 140 = 90
SIMPLEX ITERATION 3

CBj Cj 4 6 5 0 0 0 0 Solutio Ratio

n
Objective A b c S1 S2 S3 S4
variables
6 b 1/7 1 0 4/7 -1/7 0 0 60/7
5 c 5/7 0 1 -1/7 2/7 0 0 90/7
0 S3 19/7 0 0 -8/7 -5/7 1 0 860/7
0 S4 1/7 0 0 -10/7 -8/7 0 1 620/7
Zj 31/7 6 5 19/7 4/7 0 0 810/7
Cj –Zj -3/7 0 0 -19/7 -4/7 0 0

b) New Values:

0.5 – (0.5 x 5/7) = 1/7

1 – (0.5 x 0) = 1

0.5 – (0.5 x 1) = 0

0.5 – (0.5 x -1/7) = 4/7

0 – (0.5 x 2/7) = -1/7

0 – (0.5 x 0) = 0

0 – (0.5 x 0) = 0

15 – (0.5 x 90/7) = 60/7

S3 New Values:

4.5 – (2.5 x 5/7) = 19/7

0 – (0 x 2.5) = 0

2.5 – (1 x 2.5) = 0
-1.5 – (2.5 x -1/7) = -8/7

0 – (2.5 x 2/7) = -5/7

1 – (2.5 x 0) = 1

0 – (2.5 x 0) = 0

155 – (2.5 x 90/7) = 860/7

S4 New Values:

3 – (4 x 5/7) = 1/7

0 – (4 x 0) = 0

4 – (4 x 1) = 0

-2 – (4 x -1/7) = -10/7

0 – (4 x 2/7) = -8/7

0 – (4 x 0) = 0

1 – (4 x 0) = 1

140 – (4 x 90/7) = 620/7

Z VALUES

6 x 1/7 + 5 x 5/7 + 0 x 19/7 + 0 x 1/7 = 31/7

6x1+5x0+0x0+0x0=6

6x0+5x1+0x0+0x0=5

6 x 4/7 + 5 x -1/7 + 0 x -8/7 + 0 x -10/7 =19/7

6 x -1/7 + 5 x 2/7 + 0 x -5/7 + 0 x -8/7 = 4/7

6x0+5x0+0x1+0x0=0

6x0+5x0+0x0+0x1=0

6 x 60/7 + 5 x 90/7 + 0 x 860/7 + 0 x 620/7 = 810/7

The manufacturer has to produce 8.6 units of b and 12.9 units of c. The optimal profit = 6 x
60/7 + 5 x 90/7 = US$25.7

4(b) Linear Programming Assumptions

Proportionality

The basic assumption underlying the linear programming is that any changes in the
constraints inequalities will have the proportional change in the objective function. It is
assumed that the relationship between variables in the problem and the resources available
that is constraints of the problem exhibits linearity. This assumption is very useful as it
simplifies modelling of the problem.

Additivity

This assumption asserts that the total profits of the objective function is determined by the
sum of profit contributed by each product separately. Similarly, the total amount of resources
used is determined by the sum of resources used by each product separately. This implies,
there is no interaction between the decision variables.

Continuity

The assumption states that the decision variables are continuous. This means a combination
of outputs can be used with the fractional values along with the integer value. It is assumed
that the decision variables are continuous. For example, company manufacture 2.5 vehicles
and 3.2 barrels of oil. This is referred to as the assumption of divisibility.

Certainty

It is assumed that the decision maker here is completely certain that is deterministic condition
regarding all aspects of the situation, which are availability of resources, profit contribution
of the products, technology, courses of action and their consequences. The parameter of
objective function coefficients and coefficient of the constraint inequalities are known with
certainty. For example profit per unit product, availability of material, requirement of
material and labour per unit is given in the programming problem.

Finite Choices
This assumption implies that the decision maker has certain choices and the decision
variables assume non negative values. The non-negative assumption is true in the sense that
the output in the production problem cannot be negative.

Linear Programming assume fixed technology. Fixed technology refers to the fact that the
production requirements are fixed during the planning period and will not change in the
period.
SOLUTION 5

TYPE
Manufacture I II III IV Supply
r
A 2.5 4.0 5.0 2.0 150
B 3.0 3.5 5.5 1.5 450
C 2.0 4.5 4.5 2.5 250
Demand 100 200 450 150

 Here total demand is greater than supply hence we have to open a dummy row whose
requirement constraint is 50, so that total of availability will be equal to the total
demand.

TYPE
Manufacture I II III IV Supply
r
A 2.5 4.0 5.0 2.0 150
B 3.0 3.5 5.5 1.5 450
C 2.0 4.5 4.5 2.5 250
D 0 0 0 0 50
Demand 100 200 450 150 900

 Subtract all the elements in the matrix from the highest element in the matrix which is
5.5. Then the problem becomes maximization problem. Then onwards follow all the
steps of maximization problem to get the solution.
 TYPE
Manufacture I II III IV Supply
r
A 3 1.5 0.5 3.5 150
B 2.5 2.0 0 4.0 450
C 3.5 1.0 1.0 3.0 250
D 5.5 5.5 5.5 5.5 50
Demand 100 200 450 150 900

TYPE
Manufacture I II III IV Supply
r
A 100(4) 50(5) 150
B 450(1) 450
C 200(2) 50(3) 250
D 50(6) 50
Demand 100 200 450 150 900

Cost = 100 x 2.5 + 200 x 4.5 + 450 x 5.5 + 50 x 2.5+ 50 x 2.0 + 50 x 0 =$ 3850

OCCUPIED CELLS

D (A) + R (I) = 2.5

D (A) + R (IV) = 2.0

D (B) + R (III) = 5.5

D (C) + R (II) = 4.5

D (C) + R (IV) = 2.5

D (D) + R (IV) = 0

D (A) = 0

R (I) = 2.5

R (IV) = 2.0

R (II) = 4.0

D (C) = 0.5

D (D) = -2.0

UNOCCUPIED CELLS

D (A) + R (II) = 0 + 4.0 = 4.0

D (B) + R (IV) = 1.5

D (C) + R (I) = 2.5 + 0.5 = 3.0

D (D) + R (I) = -2 + 2.5 = 0.5

D (D) + R (II) = -2 + 4 = 2

CELL ACTUAL SHADOW COST RESULT

AII 4.0 4.0 0
CI 2.0 3.0 -1
DI 0 0.5 -0.5
DII 0 2 -2

The results are negative, therefore no further cost reduction and hence the optimum solution
has been reached with transportation cost of $3850.

(b) Cost = 100 x 2.5 + 200 x 4.5 + 450 x 5.5 + 50 x 2.5+ 50 x 2.0 + 50 x 0 =$ 3850
(c) Optimisation

OCCUPIED CELLS

D (A) + R (I) = 2.5

D (A) + R (IV) = 2.0

D (B) + R (III) = 5.5

D (C) + R (II) = 4.5

D (C) + R (IV) = 2.5

D (D) + R (IV) = 0

D (A) = 0

R (I) = 2.5

R (IV) = 2.0

R (II) = 4.0

D (C) = 0.5

D (D) = -2.0

UNOCCUPIED CELLS

D (A) + R (II) = 0 + 4.0 = 4.0

D (B) + R (IV) = 1.5

D (C) + R (I) = 2.5 + 0.5 = 3.0

D (D) + R (I) = -2 + 2.5 = 0.5

D (D) + R (II) = -2 + 4 = 2

CELL ACTUAL SHADOW COST RESULT

AII 4.0 4.0 0
CI 2.0 3.0 -1
DI 0 0.5 -0.5
 DII 0 2 -2

The results are negative, therefore no further cost reduction and hence the optimum solution
has been reached with transportation cost of $3850.

SOLUTION 6(a)

Operations research model is a mathematical or theoretical description of various variables of

a system representing some aspects of a problem on some subject of interest. The model
enables to conduct a number of experiment involving theoretical subjective manipulations to
find some optimum solution to the problem on hand which is to be solved.

Advantages of a good operations research model

 It provides logical and systematic approach to the problem

 It provides the analyst a base for understanding the problem and think of methods of
solving.
 Models saves resources like money and time
 It incorporates useful tools which help eliminating duplication of methods applied to
solve specific situation
 It indicates the nature of measurable quantities in the problem

Limitation of operations research models

 Models are constructed only to understand the problem and attempt to solve the
problem that is they are not to be considered as real problem or system.
 The validity of any model can be verified by conducting the experimental analysis
and with relevant data characteristics
 In operation research models, there are no place the factors that cannot be quantified
6(b) Operations Research Models

Deterministic Models

When the change of one variable has a certain or define change in the outcome, the model is
called a deterministic model. In this model the operations research analyst assumes complete
certainty about the values of the variables, the available resources and expects that they do
not change during the planning horizon. All these are deterministic models and do not contain
the element of uncertainty or probability. For example linear programming, assumes certainty
regarding the values of variables and constraints which makes linear programming model a
deterministic model.

Simulation Model

In simulation model, the behaviour of the system under consideration is initiated over a
period of time. Simulation models do not need mathematical variables to be related in the
form of equations. Normally these models are used for solving such problems that cannot be
solved mathematically. Simulation is a general technique, which helps us in developing
dynamic models, which are similar to the real process. Developing good simulation models is
difficult because creating a real life situation to perfection is extremely difficult.

Mathematical Models

Mathematical models focuses on creating a mathematical representation of management

problems in organizations. All the variables in a particular problem are expressed
mathematically. The model then provides different outcomes which will result from the
different choices the management wishes to use. The best outcome in a particular situation
will help the management in decision-making. These models use set of mathematical symbols
to represent different variables in many business and industry situations by mathematical
equations. For example the demand curve in economics can be seen as symbolic
representation of the customer behaviour at varying price levels.
SOLUTION 7(a) Decision Table

STATE OF NATURE
ALTERNATIVE Good Fair Poor Maximum Minimum Average
S market market market in row in row in row
Small 50,000 20,000 -10,000 50,000 -10,000 20,000
Medium 80,000 30,000 -20,000 80,000 -20,000 30,000
Large 100,000 30,000 -40,000 100,000 -40,000 30,000
Very Large 300,000 25,000 -160,000 300,000 -160,000 55,000

i. Maximax is $30,000
ii. Maximin is $10,000
iii. Average is $55,000
 Good market

Fair market
1

Poor market

7(b)

Good market
Small fair market
2
Medium poor market

Decision node
Large good market

Fair market
3
Very large poor market

Good market

4 Fair market
 Poor-market

NODE ALTERNATIVE EVM CALCULATION EVM

S
1 Small 1/3(50,000) + 1/3(20,000) + 1/3(-10,000) $20,000

2 Medium 1/3(80,000) + 1/3(30,000) + 1/3(-20,000) $30,000

3 Large 1/3(100,000) + 1/3(30,000) + 1/3(-40,000) $30,000

4 Very large 1/3(300,000) + 1/3(250,000) + 1/3(-160,000) $55,000

Expected Monetary Values:

Small = $20,000

Medium = $30,000

Large = $30,000

Very Large = $55,000

 Based on EVM values, the best decision is to construct a very large fuel station.