Lecture Notes on Operations Research

LECTURE 1

INTRODUCTION

History of Operations Research

The beginning of Operations Research (O.R) is attributed to military services in World War
II. The British and American military management called on a group of scientist to research
into allocation of scarce resources to (military) operations effectively. It is claimed that
their efforts were instrumental in winning various battles. Due to the apparent success, the
industry became interested in OR and therefore many of the OR techniques were developed
to assist in rebuilding after the war. The development of computers in the early 1960’s with
their ability to perform complex arithmetic calculations very fast further enhanced the
development of OR.

The Nature of Operations Research

Operations research is concerned with optimal decision making in, and modelling of,
deterministic and probabilistic systems that originate from real life. These applications that
occur in government, business, engineering, economics, and natural science and social
science, are characterised largely by the need to allocate limited resources.

Many industries, including aircraft and missile, automobile, communication computer,

electric power, electronics, food, metallurgy, mining, paper, petroleum, transportation,
financial institutions, governmental agencies and hospitals, have made widespread use of
operations research.

Operations Research techniques can broadly be classified into 2 main parts:

1. Mathematical Programming Models

2. Probabilistic Models

Mathematical Programming Comprises:

• Linear programming and extensions
• PERT/CPM.
• Network flow
• Dynamic programming
• Game Theory
• Integer programming
• Non-linear programming

Probabilistic Models Comprise:

• Markov Chains and Markovian Decision Processes
• Queuing Theory and Applications
• Inventory Theory
• Forecasting
• Reliability
• Decision Analysis
• Simulation

Introduction 1
 Lecture Notes on Operations Research

What is a Model?
A model is a representation of a group of objects or ideas in some form other than that of
the entity itself. Models are important because they are used to enable one learn something
about a real world system that cannot be observed or experimented with directly.

A Classification of Models
In terms of their prevalence in practice there are three types of models:
• verbal models
• physical models, and
• mathematical models

Verbal Models
A verbal model expresses all of the functional relationships between the variables in a word
passage. Verbal models are used extensively in the business world (especially in
advertising) and have the advantage of being easy to understand.

The disadvantages are that they cannot be experimented with, they do not indicate how
outcomes or measures of effectiveness change with decision alternatives and it is not easy
to show how the relationships change with decision alternatives.

Verbal models however play an important role in the decision making process. They can be
used to verbalise decision strategies from more sophisticated models.

Physical Models
A physical model takes on the physical appearance of the object of study. It is normally
used to display or test the design of items like new buildings or new products. For instance,
in the aircraft industry a scale model may be used to test the aerodynamics of the design in
a wind tunnel.

Physical models are advantageous for being usable for experimentation and also lucidly
describe the problem or system under study which may help in generating innovative
design alternatives for solving decision problems. Physical models are however limited to
solving a relatively limited class of problems (mainly design problems). In addition
physical models do not contain explicit relationships between decision alternatives,
therefore trial-and-error method is used which may make it time consuming and expensive.

Mathematical Models
A mathematical model expresses its relationship in mathematical symbols. Mathematical
models are therefore abstract as one cannot visualise the system being portrayed.

The main shortcomings of mathematical models are: the degree of abstraction makes it
difficult for managers to accept such models and due to some limitations in mathematical
symbolism some systems (real problems) may be grossly oversimplified.

Mathematical models have the advantage of facilitating experimentation due to explicitly

stated interrelationships between the various variables, the effects of different decision
alternatives can be tested more easily, they can also represent many complex problems
efficiently and concisely and in many cases provide the cheapest way to analyse problems.

Mathematical models are used extensively in operations research.

Introduction 2
 Lecture Notes on Operations Research

Modelling Approach in Operations Research

An operations research study normally involves the following phases:
1. Formulating the problem
2. Constructing a mathematical model
3. Deriving a solution from the model
4. Testing the model and the solution
5. Establishing controls over the solution

Formulating the Problem

• Study of the relevant system to develop a well defined statement of the problem to be
considered.
• Study includes determining objectives, constraints on what can be done
• Study interrelationships between area to be studied and other areas of organisation,
possible alternative courses of action and time limits to make decisions

Constructing a Mathematical Model

• Constructing a mathematical model that represents the essence of the problem.
• The number of quantifiable decisions to be made should be represented by decision
variables (e.g. x1, x2, xn) whose respective values are to be determined
• The appropriate measure of performance should also be expressed as a mathematical
function of the decision variables. This function is called an objective function.
• Any restrictions on the values that can be assigned to the decision variables are also
expressed as equations or inequalities - these expressions for restrictions are called
constraints.

Deriving a Solution
• In many cases, this involves the use of one of the standard algorithms applied on a
computer. This may involve the use of a standard software package of the development
of computer codes on the basis of the algorithm.
• Normally an optimal or a good sub-optimal solution is obtained based on time or cost
involved.
• Postoptimality analysis is conducted. This involves analyses of the sensitivity of the
optimal (sub optimal) solution to various sensitive parameters of model.

Testing the Model and Solution

Involves testing the validity of the model as to whether or not it predicts the effects of
alternative courses of action with sufficient accuracy to permit a sound decision.

Establishing Control over Solution

• Install a well-documented system (including post optimality analysis) for applying an
acceptable model.
• A systematic procedure for detecting change and controlling the solution should also be
installed.

Introduction 3
 Lecture Notes on Operations Research

Course Outline
The course will be presented under the following topics.

Introduction
General overview of operations Research (OR) including history, nature of OR, models and
modelling approach to OR.

Linear Programming (LP)

General structure of LP models, model formulation, solution of LP problems including
graphical, simplex and treatment of special cases.

Duality and Sensitivity Analysis

Duality, symmetric and asymmetric primal - dual relationships, economic interpretation of
the dual and dual simplex method. Sensitivity analysis involving right-hand-side and
objective function coefficients and computer solution of linear programming problems.

Transportation and Assignment Methods

Nature of transportation problem, balanced transportation problem, north west corner and
Vogel’s approximation methods, stepping stone and MODI methods, unbalanced
transportations problem, degeneracy and other special cases, maximisation problem. Nature
of assignment method and the Hungarian method in application to some examples.

Goal Programming
Pre-emptive and non-pre-emptive goal programming models, goal programming model
formulation.

Integer Programming
Problems requiring integer solution, solution of integer programming problem by branch-
and-bound method, computer solution of IP problems and some applications of integer
programming.

Decision Analysis
Basic concepts, decision making under certainty, decision making under risk, decision
making under uncertainty, Hurwicz, Wald, Savage, and Laplace decision criteria, areas of
application and examples.

Markov Analysis
Basic concepts: states, retention and transition probabilities, Markov chains of first, second,
third and higher orders and objectives, calculation of future probable states (Short-run and
Long-run), areas of application and examples.

Simulation
Simulation models, the simulation process, stochastic simulation, Monte Carlo sampling,
random process generator, simulation of a queuing system, generation of random variates,
analysis of simulation input, analysis of simulation output and simulation languages.

Queuing Models
Basic concepts: birth and death process, queues, single-channel and multi-channel queues,
arrival and service rates, single channel queuing models, Poisson arrival with exponential

Introduction 4
 Lecture Notes on Operations Research

service rates, Poisson arrival, minimum cost service rates, multi-channel Poisson arrivals
with exponential service rates, system utilisation factor, areas of application and solution of
typical problems

Inventory Control Models

Inventory-control models, major types of inventory systems, fixed-order quantity, fixed
period, variable-order quantity, variable period and combination of systems. Explanation of
basic concepts: Economic Order Quantity (EOQ), Reorder Point, Safety or Buffer Stock,
Optimal Number of Orders per Year, simultaneous sales and production of two or more
items, inventory methods with variable demand, ABC analysis.

Network Scheduling (CPM/PERT)

Components of CPM/PERT network and precedence relationships, activity scheduling,
uncertainties in activity times, project crashing.

Course Presentation and Student Assimilation

The course is taught through lectures supported with limited hand-outs. The student can
best understand and appreciate the subject by attending all lectures and tutorials, by reading
references and hand-outs, and by completing all assignments and course work on schedule.

Tutorials, Assignments, Course Work and Examination

Tutorials will be in the form of problem solving and will constitute an integral part of each
lecture; in this way, the application of each theory or model in mining can be directly
demonstrated.

Assignments will be given for each topic treated to encourage students to test and
consolidate their understanding of the discipline.

For the course work, students in a group will be required to study a specific problem,
formulate a model and provide useful solution to the problem. The methodology of study,
the approach to model formulation and recommended solution shall be presented in a
technical report to the class for discussion.

The examination will be an open-book examination lasting for 4 hrs or more and will aim at
finding the students’ understanding of the discipline.

Assessment of Students

The final subject mark will comprise the following:

(i) Continuous Assessment 40%, made up of:

• Assignments + Attendance 30%
• Coursework 10%

(ii) Final Examination 60%

Introduction 5
 Lecture Notes on Operations Research

Assessment of Lecturer

At the end of the course, each student will be required to evaluate the course and the
lecturer’s performance by answering a questionnaire specially prepared to obtain the views
and opinion of students about the course and the lecturer. Students are encouraged to be
sincere and frank in their answers so that the answers can be taken as a good basis for
improvement.

References

1. James E. Shambin and Steves, G.T., Jr. (1974). Operations Research – A Fundamental
Approach, McGraw Hill, 404 pp., ISBN 0-07-056378-0

2. Ackoff, R.L. and Sasieni, M.W. (1968). Fundamentals of Operations Research, Wiley
& Sons, 445 pp., ISBN 0-471-003344.

3. Howard Raiffa (1970). Decision Analysis – Introductory Lectures on Choices under

Uncertainty, Addison-Wesley, 309 pp., ISBN 0-201-062909.

4. Hillier and Liberman (1992). Introduction to Operations Research, Holden-Day Inc.,

Oakland, USA, p. 5.

5. Theifrauf, R.J., Klekamp, R.C. and Rume, M.L. (1985). Management Science: A Model
Formulation Approach with Computer Applications, Bell & Howeell Co., Ohio, USA, p.
4.

6. Raifa, H. (1968). Decision Analysis: Introductory Lectures on Choices under

Uncertainty, Addison-Wesley, Massachusetts USA, 309 pp.

7. Battersby, A. (1979). Network Analysis: for Planning and Scheduling, Macmillan,

London, USA, 316 pp.

8. Smith, D. (1973). Linear Programming Models in Business, Polytech, Stockport, UK,

219 pp.

9. Pegden, C. D., Shannon, R. E. and Sadowski, R. P. (1995), Introduction to Simulation

Using Siman, 2nd Ed., McGraw-Hill, New York, 600 pp.

10. Taylor, B. W. (1982), Introduction to Management Science, Wm. C. Brown Company

Publishers, Dubuque, 691 pp.

11. Winston, W. L. (1994), Operations Research – Applications and Algorithms, 3rd Ed.,
Duxbury Press, Belmont, 1318 pp.

12. Krajewski, L. J. and Thompson, H. E. (1981), Management Science: Quantitative

Methods in Context, John Wiley & Sons, New York, 544 pp.

Introduction 6
 Lecture Notes on Operations Research

Course Aims and Objectives

The main aims of the course are:

(i) To prepare the student to become a good engineer by equipping him/her with OR
scientific approach in problem-solving and decision making.
(ii) To equip the student with OR skills by which he/she can make objective
managerial decisions in his/her professional life.

The objectives of the course are:

(i) To introduce students to the OR concepts and techniques.

(ii) To expose students to problems in the mining and mineral industry which can be
solved through the application of OR techniques.
(iii) To encourage students to employ OR techniques to arrive at objective decisions.

Expected Outcomes of Course

At the end of the course, it is expected that the student will:

• Understand OR concepts and methodologies.

• Be able to analyse problems and solve them by OR methods.
• Be able to make objective decisions in their professional life.

Introduction 7