COS 419 Operations Research 3 units

MODULE ONE: Introduction to Operations Research and Linear Programming

Module Overview

Organizations have grown significantly in size and complexity since the emergence of
industrial revolution. A significant expansion in the division of labour and segmentation of
management tasks in these businesses has played a crucial role in the revolutionary
development. However, in conjunction with its benefits, this growing specialization has also
brought along new issues, issues that continue to exist in lots of businesses. One issue is the
propensity of an organization’s several parts to develop into quite independent empire with
their personal aims and value systems, losing sight of how their actions and goals fit with
those of the general enterprise. The components may wind up acting against one another
because what is beneficial for one component often is harmful for another. A related issue is
that it gets harder to distribute the resources available to the various activities in such a most
productive way for the entire organization as complexity and specialization rise inside an
organization. Operations Research came into emergence as a result of the need to respond to
these issues and find a suitable way to solve it. There is therefore need to having a better
understanding of operations research.
One of the most significant scientific developments of the mid-20th century has been the
creation of linear programming. It is a very popular technique in operations research. By
using linear programming, one can find the optimal (i.e., best) solution to the issue of
dividing few resources among competing activity. In order to be more exact, this challenge
entails deciding how many activities should compete for the limited resources required to
carry out those tasks. How much of each resource will be used by each activity depends on
the chosen activity levels. This description can be used in a wide range of circumstances
(situations) but the requirement to allocate resources to activities by selecting the degrees of
those activities is the one element that each of these circumstances has in common.
This Module provides background introduction to the key terminologies and the concept of
Operations research. It discusses the issues of operations research modelling approach and
provides insight into problem and mathematical model formulation as well as the procedure
for model testing and deriving solutions. Additionally, this module introduces the main
concepts in linear programming, discusses solutions to such problems in terms of the number
of decisions variables and the appropriate technique for each problem scenario. The emphasis
of this section is on maximization or minimization of the objective function subject to the
constraints imposed on resources. The geometrical concepts of linear programming solutions
seen in graphical method and how they apply to simplex method are demonstrated together
with other forms of solution; tabular, algebraic, pivot, duality, etc.
This module introduces you to the following unit:
Study Unit 1: Operation Research and Mathematical Modelling
Study Unit 2: Solving Linear Programming Problems and Graphical Method
Study Unit 3: Simplex Method, Canonical Method and Two Phase Simplex Method
Study Unit 1: Operation Research and Mathematical Modelling

Introduction
Operations Research (OR) is a field in mathematics, specifically applied mathematics, and it
is used to give management a scientific foundation for making quick and efficient decisions
response to their issues. It seeks to steer clear of the risks associated with making decisions
based solely on assumptions or rules of thumb (generalizations). The limits and content of
OR's are still up in the air. Consequently, it is challenging to provide a clear definition of the
phrase "Operations Research." In this unit, we shall discuss the meaning and definitions of
OR, phases of OR, various applications of OR, techniques used in OR and we shall conclude
the unit by discussing the various classifications of OR.

Learning Outcomes from Study Unit 1

At the end of this study unit, you should be able to:

1.1 Define and use correctly keywords printed in bold. (SAQ 1.1)
1.2 State issues and objectives of Operations Research (SAQ 1.2)
1.3 Identify the Stages and tools of Operations Research. (SAQ 1.3)
1.4 State the concept of models and the classification of model. (SAQ 1.4)

1.1 Meaning and Definition of Operations research

Operations Research is also known as Management Science. Operations Research (OR), as its
name suggests, is "research on operations". Therefore, challenges relating to how to carry out
and coordinate the activities (i.e. operations) within an organization are addressed using
operations research. The origins of OR can be seen in early attempts to manage organizations
scientifically, which dates back many years. However, it has generally been agreed that the
military services in the early stages of World War II were responsible for the birth of the
activity known as Operations Research.
The OR begins when quantitative and mathematical methods are applied to support the
choice being made. Below is a definition of OR:
Operations Research is the study and analysis of issues affecting complex systems using
scientific, and particularly mathematical, approaches.

However, as different definitions are emphasized by various specialists and societies, it might
be challenging to provide a specific explanation of the word "Operations Research". People
who need to make organizational decisions can use a collection of methods found in
operations research (OR). The allocation of limited resources in order to maximize the
achievement of a goal (such as profit or level of service) or to reduce some unfavourable
effect of an organization's operation is a common theme in these issues (such as cost or
environmental degradation). A complicated problem that arises from the operation of a
system or organization must be framed in a fashion that allows for effective analysis in order
to be solved.

1.1.1 Developmental Phases of Operations Research

The developmental phases of OR also known as the developmental stages or process of OR

are summarized below:
1. Determine the problem of interest and assemble pertinent data.
2. Prepare a mathematical model that fully represents the problem.
3. Formulate a procedure on how solutions to the problem can be obtained using the
developed model.
4. Test the model and make amendments if needed
5. Implement.

1. Determine the problem of interest and assemble pertinent data

The first phase of OR's creation involves generating a problem statement and having a solid
grasp of the system. To arrive at a clearly defined problem statement, this step involves a
number of actions, including interviews, site visits, observations, conferences, etc. This
allows for the observation and consideration of specific problem objectives or constraints for
the activity. This stage is crucial because it significantly affects how well the study will turn
out.

2. Prepare a mathematical model that fully represents the problem

The next step after receiving the problem description is to reframe the problem under
consideration in a way that will allow for analysis. Creating a mathematical model that
accurately represents the nature of the issue is the conventional OR approach to this. This
phase includes tasks like generating equations and defining variables.

3. Formulate a procedure on how solutions to the problem can be obtained using the
developed model

The following step after creating a mathematical model of the problem under consideration is
to create an action plan, typically using computers, to find solutions. Finding the optimal or
best solution to that problem is a common idea in OR.

4. Test the model and make amendments if needed

This is the model verification stage. Once the computer-based solution (computer program)
has been developed, it is not implemented immediately instead the solution is used to check if
limitations can be found in the program. It is ideal to test the program thoroughly to detect
and rectify bugs (errors) as much as possible. The programmer can only ascertain that a
program is good for implementation (to some extent) if it passes through this stage. This
phase in some cases reduces but does not eradicate the bugs in a program as some bugs might
be detected in the future.

5. Implement.
The final step in the OR study involves putting the system into place as instructed by
management. This is a highly important step because this is when the study's findings are
presented. Here, hidden bugs are fixed and the model solutions are translated precisely to the
operating procedure of the management that requires them. To acquire a fully implemented
solution, it is critical to involve both the management team and the OR team at this stage.

1.1.2 Operations Research Tools and Techniques

Tools used in operations research come from a variety of disciplines. Operations research
uses methods from various fields, including; computing, engineering, economics,
psychology, statistics, and mathematics, to create a new body of knowledge for making
decisions.
 1. Linear programming
This method of restricted or constrained optimization improves some criteria while adhering
to certain restrictions/constraints. The objective function (profit, loss, or return on
investment) and restrictions in linear programming are both linear. The linear programming
problem can be solved using a variety of techniques.

2. Game Theory
This is utilized when there are one or more players or opponents and there are conflicting
situations. The players' motivations are dichotomized in this. They are in conflict because one
player's success frequently comes at the expense of the other players.

3. Queuing Theory
This is utilized in instances where a line forms (such as when customers are waiting to be
served, airplanes are waiting to land, or jobs are waiting to be processed by the computer
system). The goal in this situation is to reduce waiting time expenses without raising service
fees.

4. Decision Theory
Decision-making in conditions of total certainty about future outcomes and under
circumstances where we can make some degree of probability about what will happen in the
future are both topics covered by decision theory.

5. Inventory Models
The inventory model makes choices that reduce the overall cost of inventory. With this
model, the entire cost of stock-outs and carrying inventory is successfully decreased.

6. Simulations
In the simulation process, a model of the process involved in the problem is built, and the
optimal answer is then sought for through a series of planned trial-and-error attempts. When
real experiments are impractical or when a model's solution is not viable, simulation is used.

7. Dynamic Programming
Multistage decision processes can be analyzed using dynamic programming. In this particular
setting, each decision is influenced by both previous decisions and outside factors.

8. Network Scheduling (PERT/CPM)

To plan, schedule, and keep track of huge projects like building, computer system
installation, etc., this technique is frequently employed. By identifying crucial elements, this
strategy seeks to reduce difficulty spots (such as interruptions, delays, etc.). Networks and
arrows, which are used to identify crucial activities and paths, are used to visually show the
many activities and their relationships to the overall project.
There are primarily two types of techniques in network scheduling, they are:
 Program Evaluation and Review Technique (PERT) – This is employed when an
activity's duration is not precisely known or only a probabilistic estimate is given,
 Critical Path Method (CPM) - This is employed when activity timing is precise.
Etc.

1.1.3 Application of Operations Research

1. Production and Facilities Planning
2. Resource Allocation and Distribution in Projects
3. Decision Making
4. Marketing
5. Organization Behaviour
6. Finance
7. Research and Development
Etc.

In-Text Question

How has the computer aided in the acceptance of Operation Research?

1.2 Concept and Classifications of Model

The primary aspect of OR is that it builds and applies mathematical models to evaluate
issues. It is ideal to grasp what a model is before learning what a mathematical model is.
What is a model?
A model is an idealized version of a real-world system regardless of whether the system in
question deals with a problem, process, operations, or occurrences.
The classification schemes for models are described here, along with the corresponding
categories.

1.2.1 Classification by Structure

1.2.1.1 Iconic or Physical Model
Physical models are scaled-down physical representations of real situations. Example: An
image, a map, a picture, a car model, etc.

1.2.1.2 Analogue or Schematic Model

The situations here are dynamic. Example: graphs that show population growth relative to
distance.

1.2.1.3 Symbolic or Mathematical Model

Models that use mathematical symbols, such as numbers and letters, to represent the decision
variables under consideration are known as mathematical models. Equations that explain the
system's properties are then created using these symbols. The process of putting a real-world
problem into mathematical terms, typically in the form of equations, and then utilizing those
equations to support and understand the original situation is known as mathematical
modelling. Example: Linear Programming

1.2.2 Classification by Function

1.2.2.1 Descriptive Models
This explains various operations in non-mathematical language e.g. questionnaire,
observation, survey etc.

1.2.2.2 Predictive Models

This clarifies or forecasts a system's behaviour. A forecast model, for instance, forecasts
future demand or sales.

1.2.2.3 Prescriptive Models

They are employed to prescribe specific optimal solutions for issues e.g. linear programming
models.
1.2.3 Classification by Nature of Environment
1.2.3.1 Deterministic Model
With no consideration of probability, the variables are fully defined, and the results are
guaranteed. Example: Transportation problem (this will be discussed later in the courseware).

1.2.3.2 Probabilistic Model

A model is considered probabilistic if it contains at least one random variable as a parameter
or decision variable. There is no guaranteed way to forecast how the variable will change.
The probability distribution of the variables, however, allows for the prediction of a pattern in
the values of the variables e.g. When demand is unknown, probabilistic inventory models are
utilized to determine the economic ordering amount (EOQ).

1.2.4 Classification by Behaviour

1.2.4.1.1 Static Models
They reflect the system at a certain point in time and do not take system changes during
changing time periods into account. They are used only for a single time. Example: Economic
Order Quantity Model.

1.2.4.1.2 Dynamic Models

They are employed when the passage of time is crucial. They are also employed for
multistage decision problem optimization. They can be applied to related issues repeatedly.
Example: Dynamic Programming model (this will be discussed later in this courseware).

Answer to in-text question

Companies can efficiently organize and manage vast amounts of information thanks to
computers. Engineers and scientists are able to construct larger, more accurate models of
organized systems thanks to the speed and data-handling capabilities of computers, which
also enable them to apply simulation techniques to arrive at useful solutions for such models.

1.3 Developing a Mathematical Model

Mathematical models are models used to depict a system using mathematical symbols and
expressions. Consequently, the set of equations and associated mathematical expressions that
capture the essential features of a business challenge are called its mathematical model. It
might be logically structured like simulations and algorithms, or it might be functional like in
linear programming. Therefore, if there are n linked measurable judgments that need to be
made, they are depicted as Decision Variables (such as x1, x2,..., xn) which corresponding
values need to be computed. Alternatively put, decision variables are the unknown values that
decision-makers would want to estimate. Following that, the proper performance indicator
(such as profit) is represented mathematically as a function of these decision variables such as
P = 70X + 60Y. The Objective Function is the name given to this function. Mathematical
expressions (for example, 4X + 3Y ≤ 490) are also used to represent any limitations on the
values that can be allocated to these decision variables, often in the form of inequality or
equations. Mathematical expressions for limitations like this are frequently referred to as
Constraints. The model's parameters are the constants found in the constraints and objective
function. The issue, according to the mathematical model, is determining the values of the
decision variables in order to maximize the objective function while still taking into account
the given limitations (constraints). One crucial and difficult step in creating a model is
choosing the appropriate values to put for each of its parameters. It takes collecting pertinent
information to determine parameter values for actual information and this is quite challenging
to perform. As a result, the value allocated to a parameter is frequently merely an estimate.
Another limitation to put under consideration when developing a mathematical model is the
Non-Negativity Constraints which states that all problems must be greater or equal to zero. It
is a common consideration in Linear Programming which will be discussed in another
module.
For example, a linear programming model can be represented like this:
DECISION VARIABLE
X is the number of product Z1 and Y is the number of product Z2.
OBJECTIVE FUNCTION
Maximise P = 50x + 75y
CONSTRAINTS
3x + 6y ≤ 45
5x + 8y ≤ 200
1x + 6y ≤ 30
NON-NEGATIVITY CONSTRAINT
X, Y ≥ 0

In-text Question
The company Philip's Woodcarving, Inc. makes two kinds of wooden toys: soldiers and
trains. A soldier sells for $27 and consumes $10 in raw materials. Each soldier produced adds
$14 to Philip's variable labour and administrative expenditures. A train costs $21 and requires
$9 in raw materials. Each train manufactured adds $10 to Philips' variable labour and
overhead costs. Carpentry and finishing are two types of skilled labour required in the
production of wooden soldiers and trains. A soldier takes two hours of finishing work and
one hour of carpentry work. A train takes 1 hour to finish and 1 hour to carpentry. Philip can
receive all of the required raw materials each week, but only 100 finishing hours and 80
carpentry hours. The demand for trains is limitless, yet only 40 soldiers are purchased each
week. Philip wishes to optimize his weekly profit.
Create a mathematical model of Philip's situation that will allow you to maximize his weekly
profit.
Answer to In-text Question
(1) Decision variables
These are clearly the number of soldiers and trains to be manufactured each week. We
assign as follows;
x1 = number of soldiers per week
x2 = number of trains per week.
(2) Objective Function
As this determines the function to be maximized or minimized, we note that in
Philips’ case, the fixed costs (such as rent and insurance) do not depend on the values
of x1 and x2.
Thus, Philip can focus on maximizing: (weekly revenues) – (raw material purchase
cost) – (other variable costs).
- The weekly revenues and costs can be expressed in terms of x 1 and x2. Where as it is
inconsiderable to manufacture more soldiers (i.e. x1 > 40) than can be sold for each
week, so we assume that all toys produced will be sold.
- Proceeding,
Weekly revenues = weekly revenues from soldiers + weekly revenues from trains =
(dollars/soldier) (soldiers/week) + (dollar/train) (trains/week)
= 27 x1 + 21 x2
Also,
Weekly raw material costs = 10x1 + 9x2
Others weekly variable costs = 14x1 + 10x2
Then Philip wants to maximize
 (27x1 + 21x2) – (10x1 + 9x2) – (14x1 + 10x2) = (3x1 + 2x2) = 3x1 + 2x2
where the variable Z is conventionally used to denote objective function, Philips?
Objective function becomes,
maximize Z = 3x1 + 2x2 (1)
(Note the objective f x n coefficients or contribution to profit)
(3) Constraints
We note that, from (1), equation 1, Z can increase arbitrarily by increasing the
respective values of x1 and x2. However, the values of x1 and x2 are limited by the
following restrictions.
(i) There are only 100 hrs available for finishing each week.
(ii) There are only 80 hrs available for carpentry each week.
(iii) And at most 40 soldiers should be produced each week.

Whereas the amount of raw materials available is assumed to be unlimited, no restrictions

have been placed on this.
- In order to formulate a mathematical model of the Philip’s problem, we have to
express constraints (i) to (iii) in terms of decision variables x1 and x2.

To express constraints (i), note that total finishing hrs/week = (finishing hrs/soldiers)
(soldiers made/week) + (finishing/train) ( trains made/week)
= 2(x1) + 1(x2) = 2x1 + x2
Thus, constraint (i) implies that finishing hrs per week -
2x1 + x2 ≤ 100 (2)

- To express constraint (ii), we note that total carpentry hrs/week = (carpentry

her/soldier) (soldiers/week) + (carpentry hrs/train) ( train/week) = 1(x 1) + 1(x2) =
x1 + x2
Thus constraints (ii) may be written as carpentry hrs per week -
x1 + x2 ≤ 80 (3)

- Finally, we express the fact that at most 40 soldiers per week can be sold by limiting
the weekly production of soldiers to at most 40 soldiers. This yields the following
constraint;
Maximum demand for soldiers: x1 ≤ 40 (4)

(4) Sign restrictions: clearly the sign restrictions for Philip problems
x1 ≥ 0 (5)
and x2 ≥ 0 (6)

We therefore formulate the optimization model for the Philips’ problem as follows;
maximize = 3x1 + 2x2 (1)
subject to (s.t)
2x1 + x2 ≤ 100 (2)
x1 + x2 ≤ 80 (3)
x1 ≤ 40 (4)
 x1 ≥ 0 (5)
x2 ≥ 0 (6)

“subject to” (s.t) means that the values of the decision variable x1 and x2 must satisfy all
constraints and all sign restrictions. Note however, that the sign restrictions are usually
considered as being separate from the functional constraints.
This mathematical model can be seen as a Linear Programming equation (This will be
explained in details in the next unit)

1.4 Unit Summary

In this unit, you have learned that
1.Operations research is a branch of science that can also be referred to as management
science. For an efficient execution of operations, in order to address recognized issues, teams
of individuals engage in research.
2. Models facilitate understanding of complex situations and the development of problem-
solving strategies and the various classifications of model.
3. Mathematical models are models that employ mathematical symbols and expressions to
portray a system and it is used well in operation research to represent real life problems.

1.5 Self Assessment Questions (SAQs) for Study Unit1

Now that this study session is over, you can assess how well you met its learning objectives
by responding to the following questions:

SAQ 1.1. (Tests learning outcome 1.1)

From the statements made below, identify the TRUE/FALSE and fill the gap where
applicable;
1. Management research is another term for operation research.
2. ................................ is the use of scientific techniques to identify the best answers to an
issue.
3. A ............................... is a simplified representation of an actual system.
4. Operations Research study and analyses human using mathematical models.

SAQ 1.2. (Tests learning Outcome 1.2)

1. What issues does Operations Research address.

2. Give a major Objective of Operations Research.

SAQ 1.3. (Tests learning Outcome 1.3)

1. Explain any three phase of Operations Research.

2. Mention five (5) techniques of Operation Research.

SAQ 1.4. (Tests learning Outcome 1.4)

1. What do you understand by the term model in Operation Research.
2. What classification of model does the mathematical model fall into.
3. How do you differentiate a mathematical model from other existing models.

References and Further Reading

1. Operations research by P. Rama Murthy (2007), 2nd Edition. New age international
(p) Ltd.
2. Operations Research. By Prem Kumar Gupta and D.S. Hira (2007), S.Chand; Revised
Edition.
3. Chemma D.S, Operation Research, First edition, 2005
4. Tiwari N.k, Shandilya. K. Shishir, Operation Research, 2006.
Study Unit 2: Solving Linear Programming Problems and its Graphical Method

Introduction
Linear programming is a method for allocating scarce or constrained resources to
competing goods or activities under the suppositions of certainty, unchanging technology,
and constant profit per unit. It entails organizing activities to produce a best possible
outcome, i.e., an outcome that, in terms of the mathematical model, achieves the defined goal
the best among all realistic choices. The term "linear" denotes the necessity of all
mathematical functions in this model being linear functions. The term "programming" is not
used to describe computer programming in this context; rather, it is used as a euphemism for
planning. A mathematical model is used by linear programming to define the problem to be
addressed. This model can be applied to either maximize returns or cut (minimize)
expenditures. Although the most typical use of linear programming is to allocate resources to
activities, it also has a wide range of other crucial uses. In actuality, a linear programming
problem is any issue whose mathematical model conforms to the extremely broad definition
of the linear programming model. In this unit, we shall be discussing linear programming
problems / model and one of the methods of solving linear programming problems (Graphical
method).

Learning Outcomes from Study Unit 2

At the end of this study unit, you should be able to:
2.1 Define and use correctly keywords printed in bold. (SAQ 2.1).
2.2 Formulate problem (model) based on spoken statements (SAQ 2.2)
2.3 Describe the procedures for solving linear programming problems using graphical
method (SAQ 2.3).

2.1 LINEAR PROGRAMMING MODEL

FEATURES OF LINEAR PROGRAMMING

The following characteristics are necessary for any linear programming model (problem):
 It should have variables assigned (decision variables) to each quantity that is being
solved for.
 The model has to accomplish an objective purpose.
  The model needs to be constrained structurally.
 Variables and constraints must have a linear connection.
 There must be a non-negativity constraint in the model.
To understand the application of the aforementioned properties, let's take a look at a product
mix issue.

Before solving any LP problem, it is important to understand these terms:

Decision Variables: This describes the decision(s) to be made. The quantities that the
decision makers would like to know are described by the decision variable.
Objective Function: In any linear programming problem, the decision maker wants to
maximize (usually revenue or profit) or minimize (usually costs) of some function of the
decision variables. The function to be maximized or minimized is called the objective
function. It is a function written as a linear equation i.e. it is a linear function in an
optimization problem.

Constraint: The constraints of a linear programming problem are the linear inequality,
equation, limitations or restrictions on the decision variable for the simple and logical reason
that resources are often limited in real life.

Non-Negativity constraint: It can also be called sign restrictions in some text. Regardless of
whether the goal function is to maximize or minimize the net present value of an activity,
every decision variable in any LP model needs to be positive. Examples: products, ie physical
products, have to be nonnegative and represented as xi ≥ 0, while cash balance, if money
owed is more than money on hand, could be considered negative.

In-Text Question
A manufacturer produces soap and detergent, both of which need the resources listed below.
The machines with the capacities M1, M2, and M3 are the resources. In the planning phase,
the capacities are 40, 20, and 17 hours, respectively. Machine M1 is needed for soap for 2
hours, Machine M2 for 3 hours, and Machine M3 for 1 hour. 2 hours of machine M1 and 3
hour of machine M3 are needed for detergent. Soap and detergent each provide N10 and N7
to overall profits.

Now, products soap and detergent are alternative variables in the aforementioned issue. The
available resources include machine capacity. The profits from soap and detergent per unit
are provided.
The following is a summary of the ideas of the problem statement:
Machines Products Availability (in hours)
SOAP DETERGENT

M1 2 2 40
M2 3 0 20
M3 1 3 17
Profit per unit (in Naira) 10 7

Solution to In-Text Question

Let's imagine that the business produces x units of soap and y units of detergent. Since soap
and detergent both contributed 10 and 7 Naira in profits, respectively. The goal of the issue is
to maximize profit Z.

Maximize Z = 10x + 7y..................................... (OBJECTIVE FUNCTION)

It is crucial to keep in mind that the capacity should not be exceeded when it comes to the
utilization of machine hours by soap and detergent products. For illustration, consider the
following:

I Machine M1 2x + 2y ≤ 40
II Machine M2 3x + 0y ≤ 20
III Machine M3 1x + 3y ≤ 17
I, II and III above constitutes the CONSTRAINTS of the problem.

However, the manufacturer has the option to stop producing soap and detergent ( = 0) or to
produce any quantity of x and y ( > 0). It cannot produce x and y in negative quantities (it
cannot be < 0). Therefore, we must write,
x and y are ≥ 0 i.e. x,y ≥ 0. This is termed the NON -NEGATIVITY CONSTRAINT.
The problem above satisfies the characteristics of a linear programming problem since it has
an objective function, structural constraints, non-negativity constraints, and a linear
relationship between the variables and the constraints in the form of inequalities.

2.1.1 GENERAL LINEAR PROGRAMMING (LP) MODEL

As stated earlier, LP model can be applied to either maximize returns or cut (minimize)
expenditures i.e. the OBJECTIVE FUNCTION. The following is a general mathematical
representation of a linear programming problem (L.P.P.):

The model can be written explicitly as:

Minimize / Maximize (Z) = c1x1 + c 2 x2 + … + cn xn (OBJECTIVE FUNCTION)

subjects to the conditions (constraints),

a11x1 + a12 x2 + a13 x 3 + …+a1n xn (≥, =, ≤) b1
a21 x1 + a22 x2 + a23 x3 + …………. + a2n xn ( ≥, =, ≤) b2
...... ...... ...... ...... ...... ...... ......

...... ...... ...... ...... ...... ...... ......

...... ...... ...... ...... ...... ...... ......

am1 x1 + am2 x2 + am3 x3 + …………. + amn xn ( ≥, =, ≤) bm as Structural Constraints

x 1 ≥ 0, x 2 ≥ 0,....., x n ≥ 0 (Non Negative Constraint)

Where aij , bi and cj are constants and x is the decision variable.

However, the model can be written in compact form as,

Minimize / Maximize

Subject to

(≥, =, ≤) bi where i = 1,2,.....,m

xj ≥ 0 where i = 1,2,.....,n
2.1.2 FORMULATION STEPS FOR LINEAR PROGRAMMING PROBLEMS
The stages involved in creating linear programming are as follows:
1. Recognize and symbolize the unknown decision variables that need to be
determined.
2. Express the objective function to be optimized as a linear function of the decision
variables (either maximized or minimized).
3. List all the limitations (conditions) in the issue such as resource limits, market
restraints, interdependencies across variables, etc., and express them as
inequalities or linear equations of the choice variables.
4. Include non-negativity constraints among the factors to prevent the decision
variables' negative values from having any real-world physical meaning.

2.2 FORMULATING OPTIMIZATION PROBLEMS

2.2.1 MAXIMIZATION PROBLEMS / MODELS
Let’s formulate model for a maximization problem.

In-Text Question
M and D are two products that a company makes. Product M sales volume accounts for at
least 60% of M and D's combined sales. However, the business is only permitted to sell 400
units of product M per day. One raw material, with a daily availability limit of 700kg, is used
to make both goods. Product D needs 2kg per unit production, while Product M only needs
1.2kg. Profit is N70 for one unit of M and N90 for one unit of D. Identify the company's ideal
product combination.

Answer to In-text Question

To form our model, we need to understand the problem statement.

Decision Variables
From the problem, a company produces two products, M and D,
Therefore,
let x be number of Products M and
y be number of Products D.
Objective Function
The profit for one unit of M is N70 and for D is N90.
Therefore,
Maximize Z = 70x + 90y

Constraints (limitation / conditions)

The sales volume for M is at least 60% of the total sales of both M and D.
Therefore,
x ≥ 0.6 (x + y) ........1st constraint (equation 1)

The company cannot sell more than 400 units of product M per day
x ≤ 400 ...................2nd constraint (equation 2)

Both products are made of one raw material and availability of that is maximum 700kg daily.
Product M needs 1.2kg and product D needs 2kg per unit production.
1.2x + 2y ≤ 700 .......3rd constraint (equation 3)

Non-Negativity Constraint
x,y≥0

Therefore,

Decision Variables Objective Function

x be number of Products M and Maximize Z = 70x + 90y
y be number of Products D.

Constraints Non-Negativity Constraint

x ≥ 0.6 (x + y) ........i x,y≥0
x ≤ 400 ...................ii
1.2x + 2y ≤ 700 .......iii

In summary, the model for the above problem is:

Maximize Z = 70x + 90y

x ≥ 0.6 (x + y) ........i
x ≤ 400 ...................ii
1.2x + 2y ≤ 700 .......iii
x,y≥0

2.2.2 MINIMIZATION PROBLEMS / MODELS

Let’s formulate model for a minimization problem. `

In-Text Question
A patient visits a doctor to have his condition evaluated. After examining him, the doctor
informs him that he is vitamin A and vitamin D deficient. In order to regain his health, the
doctor instructs him to consistently consume vitamin A and D for a while. The doctor
recommends tonics X and Y that contain specific amounts of vitamin D and A. The patient
should also take at least 60 units of vitamin A and 30 units of vitamin D daily, according to
the doctor. The table below lists the price of tonics X and Y as well as the percentage of
vitamins A and D that each tonic contains. Create a model to minimize the cost of tonics.
Vitamins Tonic Daily requirements (unit)
X Y
A 2 3 60
D 4 2 30
Cost in Naira per unit 7 4

Answer to In-text Question

Decision Variable
Let patient purchase x units of X and y units of Y.

Objective function
Minimize Z = 7x + 4y

Constraints
For Vitamin A is 2x + 3y ≥ 60 (The word at least in this case indicates that the patient may
take more than 60 units of vitamin A daily, but not less than 60 units.).

For Vitamin D is 4x + 2y ≥ 30.

Non–negativity constraint
x and y must be ≥ 0 because the patient cannot consume negative units.
In summary,

Minimize Z = 7x + 4y
2x + 3y ≥ 60
4x + 2y ≥ 30
x,y≥0

2.2.3 Solving Linear Programming Problem

The two most significant approaches to solving linear programming problems are
 Graphical Method
 Simplex Method

2.3 GRAPHICAL METHOD

The linear programming with two variables is optimized using the graphical technique.
Finding the ideal answer is best done using a graphical approach when there are two decision
variables in the problem. The inequalities (structural restrictions) are regarded as equations in
the graphical method. This is because inequality cannot be expressed by a graph. The
inequalities are then shown using an XY plane graphic. The intersection region will be used
to determine the feasible region once all inequalities have been drawn on the XY graph. The
ideal answer and all possible values for our model are explained by the feasible region.

Example:
Solve the following Linear Programming Problem using graphical method.
Maximize Z = 12x1 + 16x2
Subject to:
10x1 + 20x2 ≤ 120 .....i
8x1 + 8x2 ≤ 80..........ii
x1, x2 ≥ 0
Solution:
First steps to take when solving a LPP using graphical method is to convert each constraint
inequality into an equation and set each variable equal to zero so as to produce two
coordinate points for each equation.
10x1 + 20x2 = 120 ......... i
From the equation in (i), when x1 = 0 (i.e. substitute x1 for zero (0)), x2 = 6 and when x2 = 0
(i.e. substitute x2 for zero (0)),, x1 = 12.
That is,
x1 0 12
x2 6 0
This results to points (0, 6) and (12 , 0). We will sketch the graph and connect each
coordinate point with the necessary straight line.

x2
10

9
8
7
6

5
4

1
x1
0 1 2 3 4 5 6 7 8 9 10 11 12

Figure 1.1 Graph for Equation I

 8x1 + 8x2 = 80..........ii
From the equation in (ii), when x1 = 0 (i.e. substitute x1 for zero (0)), x2 = 10 and when x2 = 0
(i.e. substitute x2 for zero (0)),, x1 = 10.
That is,
x1 0 10
x2 10 0

This results to points (0, 10) and (10 , 0). We will sketch the graph and connect each
coordinate point with the necessary straight line

x2

10
9
8
7
6

5
4
3

1
x1
0 1 2 3 4 5 6 7 8 9 10 11 12

Figure 1.2 Graph for Equation II

 x2

10

9
8
7 FEASIBLE REGION
D
6

5
A
4

3 10

C
Figure 1.3 Graph for Equation I and II Combined
2

1
B
x1
0 1 2 3 4 5 6 7 8 9 11 12

Figure 1.3 shows the combination of equation I and equation II graphs. The shaded region of
the closed polygon ABCD is referred to as the FEASIBLE REGION. The feasible region is
the area shared by all the constraints in the linear programming problem, including the non-
negativity constraints. i.e. the set of locations that satisfy each inequality in a collection of
inequalities.

Test of Optimality:

The final step is to pick out the point in the feasible region that maximizes the value of Ƶ =
12x1 + 16x2
To discover how to perform this step efficiently, begin by trial and error (heuristics) . The
search for an optimal solution typically involves migrating from one corner point feasible
solution (CPFS) to the next and observing the corresponding effect on Ƶ, the objective
function. This movement represents the iterative steps towards the realization of the optimum
value of Ƶ.
The iteration continues, until an adjacent CPFS, at this point, the optimal solution is reached
beyond which no further improvement on Ƶ is possible.
To get the optimal result from our example, we will locate the closed polygon ABCD's corner
coordinates in Figure 1.3 and replace the values in the objective function. In a maximization
problem, we choose the coordinates that produce the highest value. In the minimization
problem, we choose the coordinates that yield the lowest value.

Maximize Z = 12x1 + 16x2

The CPFS are A = (0,0) , B = (10,0), C = (8,2) and D = (0, 10).

Z(A) = 12 (0) + 16 (0) = 0

Z (B) = 12 (10) + 16 (0) = 120
Z (C) = 12 (8) + 16 (2) = 128
Z (B) = 12 (0) + 16 (6) = 96.

Since we are dealing with a maximization problem, the optimal (highest) solution for the
problem is CPFS C (8,2). Therefore, the Z value is maximum for corner point C where x1 = 8
and x2 = 2, Optimum = 128.

2.4 Duality Theory

This is one of the most important discoveries in the early development of Linear
Programming. Duality theory has many important ramifications. The theory revealed that
every Linear Programming problem has associated with it another Linear Programming
problem called the dual.

The relationship between the dual problem and the original problem (called the primal)
prove to be extremely useful in a variety of ways. One of the key uses of duality theory lies in
the interpretation and implementation of sensitivity analysis. Sensitivity analysis plays an
important role in the determination of the choice of parameter values, i.e. values of decision
variables, when they are the main issues to be studied for the purposes of managerial
decisions.

Let us illustrate the essence of duality theory by using an example primal problem in a
general form and present its dual after.
A primal problem

Maximize Ƶ =

Subject to

≤ bi for i= 1, 2, …, m

and xj ≥ 0 for j = 1, 2, …., n

The Dual Problem

Minimize Ƶ =

Subject to

≤ bi for i= 1, 2, …, m

and yi ≥ 0 for i = 1, 2, ..., m.

Thus, by the provisions of the duality theory, the primal problem, if in maximization form,
the dual problem is in minimization form instead.
Furthermore, the dual problem uses exactly the same parameters as the primal problem, but in
different locations, as summarized below;

1) The coefficients in the objective function of the primal problem are the right hand
sides of the functional constraints in the dual problem.
2) The right-hand sides of the functional constraints in the primal are the coefficients in
the objective function of the dual problem.
3) The coefficients of a variable in the functional constraints of the primal problem are
the coefficients in a functional constraint of the dual problem.

Illustration:
1) Where the LP Primal Problem in algebraic form is

maximize Ƶ =3x1 + 5x2

 Subject to:
x1 < 4
2x2 < 12
3x1 + 2 x2 < 18
x1 , x2 > 0.

In Matrix notation, the above problem becomes:

For the Primal Problem:
Maximize Ƶ = [3, 5] x1
x2
Subject to
1 0 x1 4
0 2 x2 < 12
3 2 18

and x1 > 0
x2 0
The Dual Problem in matrix form is:

Minimize W = [Y1, Y2, Y3] 4

12
18
Subject to:

[Y1, Y2, Y3] 1 0

0 2 > 3, 5
3 2
and
[Y1, Y2, Y3] > [0, 0, 0]

We note that general matrix notation for an LP problem in standard form, where C and Y =
[Y1, Y2, — ; Ym] are now vectors but b and x are column vectors is of the form;

Primal Problem Dual Problem

 Maximize Ƶ = Cx, Minimize W = Yb,
Subject to Subject to
Ax < b YA > C.
and x > 0 and Y > 0.

2.4.1 The tabular form of the Duality theorem

The tabular form of the duality theorem is an adaptation of the tabular form of the general LP
problem.
Primal Problem
Coefficient of:
Dual Problem RHS
x1 x2 . . . xn
Coefficient Y1 a11 a12 . . . a1n < b1 Coefficient
of Y2 a21 a22 . . . a2n < b2 of
Objective fxn
Ym am1 am2 . . . amn (Minimize)
< bm
RHS v1 v1 . . . v1
C1 C2 . . . Cn
Coefficient of objective fxn
(Maximize)

From the above table, consequently, we can deduce that the following general relationships
exist between the primal and dual problem;

1) The parameters for a (functional) constraint in either problem are the coefficients of a
variable in the other problem.

2) The coefficients in the objective function of either problem are the right-hand sides
for the other problem.

Thus, there is a direct correspondence between these entities in the two problems, as
summarized in the table below;
Table: Correspondence between entities is primal as dual problem.
 One Problem Other Problem
Constraint i Variables i
Objective function Right hand sides

These correspondences are a key to some of the applications of duality theory,

including sensitivity analysis.
2.5 The Pivot
In ordinary terms, the pivot or pivot element is the element of a matrix, or an array, which is
selected first by an algorithm, for example; Gaussian elimination, simplex algorithm, to do
certain calculations. In the case of matrix algorithms, a pivot entry is usually required to be at
least distinct from zero, and often distant from it; in this case finding this element is called
pivoting. Pivoting may be followed by an interchange of rows or columns to bring the Pivot
to a fixed position and allow the algorithm to proceed successfully and possibly to reduce
round-off error. It often used for verifying row echelon form. Pivoting might be thought of as
swapping or sorting rows or columns in a matrix, and thus it can be represented as
multiplication by permutation matrices. However, algorithms rarely move the matrix
elements because this would cost too much time; instead, they just keep track of the
permutations.

Overall, pivoting adds more operations to the computations cost of an algorithm. These
additional operations are sometimes necessary for the algorithm to work at all. Other times
these additional operations are worthwhile because they add numerical stability to the final
result. The simplex algorithm is one of such algorithms/methods that makes use of pivoting.
Thus, we carry over further discussion of the concept of pivoting and its illustration to the
section on simplex algorithm.

2.6 Summary of Study Unit 2

In this unit, you have learnt;

1. A component of operations research called linear programming uses the creation of a

mathematical model to address allocation issues.
2. The characteristics of linear programming models include:
a. Objective functions
 b. Constraints
c. Non-negative constraint.
3. Only a linear programming model with two choice variables can be solved
graphically.

2.7 Self-Assessment Questions (SAQs) for Study Unit 2

Now that this study session is over, you can assess how well you met its learning objectives
by responding to the following questions:

SAQ 2.1 (Tests learning outcome 2.1)

From the statements made below, identify the TRUE/FALSE and fill the gap where
applicable;
1. ............................ is a mathematical technique used in finding a solution for optimizing a
given objective, such as profit maximization or cost reduction under certain constraints.

2. Constraints in an LP model represent the function to be minimized or maximized.

3. A linear function in an optimization problem is referred to as Objective function.

4. When there are only two choice variables, the graphic method of linear programming is
useless.

5. The ........................ is the area shared by all the constraints in the linear programming
problem, including the non-negativity constraints.

6. In LP, the limitations, restrictions or boundaries on the total amount of a specific resource
needed to complete the tasks that determine the degree of success in the decision variables is
referred to as ........................................

SAQ 2.2 (Tests learning outcome 2.2)

Jonas Woodcarving, inc, manufactures two types of wooden toys: Soldiers and trains. A
soldier sells for N27 and uses N10 worth of raw materials. Each soldier that is manufactured
increases Jonas’s variable labour and overhead cost by N14. A train sells for N21 and uses
N9 worth of raw materials. Each train built increases Jonas’s variable labour and overhead
costs by N10. The manufacture of wooden soldiers and trains requires two types of skilled
labour’s carpentry and finishing. A soldier requires 2 hours of finishing labour and 1 hour of
carpentry labour. A train requires 1 hour of finishing and 1 hour of carpentry labour.
Each week, Jonas can obtain all the needed raw materials, but only 100 finishing hours and
80 carpentry hours. Demand for trains is unlimited, but at most 40 soldiers are bought each
week. Jonas wants to maximize weekly profit (revenue-costs). Formulate a mathematical
model of Jonas’s situation that can be used to maximize Jonas’s weekly profit.

SAQ 2.3 (Tests learning outcome 2.3)

1. What do you understand by Feasible Region.

2. Show the feasible region using the linear programming problem's graphical approach.

Maximize Ƶ = 3x1 + 5x2

Subject to:
x1 < 4
2x2 < 12
3x1 + 2x2 < 18
and x1 , x2 > 0

References and Further Reading

1. Operations research by P. Rama Murthy (2007), 2nd Edition. New age international
(p) Ltd.
2. An introduction to management science, quantitative approaches to decision
making(2017), 11th Edition . Cengage
Study Unit 3: Simplex Method, Canonical Method and Two Phase Simplex Method

Introduction
One may easily handle issues with two variables, as we saw in the last study unit on linear
programming problems. When there are more than two variables, the answer is exceedingly
laborious and difficult. The graphical method of the LPP can be used to answer problems
with two variables, however solving problems with three or more variables is quite difficult.
In these situations, the simplest and most popular simplex approach is applied. The linear
programming fundamental theorem serves as the foundation for the algorithm provided by
the simplex approach. By using a series of repetitive operations, the simplex method is an
algebraic technique that helps us gradually get closer to the optimal solution. The simplex
technique is based on the principle that, if an ideal solution to a linear programming problem
exists, it can always be identified in one of the fundamental feasible solutions. It deals with
an iterative process that starts with creating a basic feasible solution or a program and moves
on to the OPTIMAL SOLUTION, checking each feasible solution for Optimality along the
way to determine whether the current solution is optimal or not. If there isn't an optimal
solution, modify the program and retest till it test results are optimal. We can therefore
conclude that the Simplex Method is dependent on the ideas of FEASIBILITY and
OPTIMALITY. Since there are multiple steps in this process, it can be used to solve issues
with any number of restrictions and variables; however, problems with more than four
variables cannot be solved manually and must instead be solved by a computer. In this study
unit, Simplex method will be explained.

Learning Outcomes from Study Unit 3

At the end of this study unit, you should be able to:
3.1 Define and use correctly keywords printed in bold (SAQ 3.1).
3.2 Identify the Procedures needed to use a simplex approach (SAQ 3.2).
3.3 Use the Simplex method to solve maximization and minimization problem of LPP
(SAQ 3.3).

3.1 The Simplex Method

The simplex method is a general procedure for solving linear programming problems. This
method was developed by George Dantzig in 1947. It has become an efficient method used
routinely to solve huge problems on today’s computers. Thus, there are sophisticated
software packages in use currently for the simple method. We note that there are extensions
and variations of the simplex method also used to perform post optimality analysis (including
the already mentioned sensitivity analysis) on the model. Essentially, the simplex is an
algebraic procedure. However, its underlying concepts are geometric. Understanding these
geometric concepts provides a strong intuitive feeling for how the simplex method operates
and what makes it so efficient.
A typical graphical solution to LP problem can be used to illustrate the general geometric
concepts as well as the algebraic concept, the simplex method. Basically, the constraint
boundaries and their points of intersection are highlighted in the graphs because they are keys
to the analysis. In the graph, or geometric shape, each constraint boundary is a line that forms
the boundary of what is permitted by the corresponding constraint. The points of intersection
are the corner-point solutions of the problem. All the points that lie on the corners of the
feasible region are the Corner-Point Feasible solutions (CPFS), while the points that lie on
outside of the feasible region are the corner-point infeasible solutions. In any LP problem
with a number of decision variables, two CPF solutions are said to be adjacent to each other if
they share n-1 constraint boundaries. The two adjacent CPF solutions are connected by a line
segment that lies on these same shared constraint boundaries such a line segment is referred
to as an edge of the feasible region.

Optimality Test:
Adjacency of CPFS solutions provides a very useful way of checking whether a CPFS
solution is optimal. Where Optimality test, in LP problem with optimal solution (s), involves
finding a CPFS that has no adjacent CPFS solutions that are better (as measured by the
objective function). This optimality test is the one used by the simplex method for
determining when an optimal solution has been reached. In short, finding an optimal solution,
if it exists, by simplex method simply involves migrating from one CPFS solution to another
until no better one can be reached.
3.2 DIFFERENCE BETWEEN GRAPHICAL AND SIMPLEX METHODS
 1. When there are two choice variables in the problem, the graphical approach is
employed. In contrast, the problem may contain any number of choice variables when
using the Simplex technique.
2. The inequalities are treated as equations in the graphical method so that straight lines
can be drawn. In contrast, the Simplex technique turns inequalities into equations by
adding a SLACK VARIABLE in the case of maximization problems and
subtracting SURPLUS VARIABLE in the case of minimization problems.
3. While in the Simplex technique, the existence of the slack variable shows the idle
capacity of the resources, in the graphical method, the areas outside the feasible area
(the area covered by all the lines of constraints in the problem) indicate idle capacity
of resources.

3.2.1 The Simplex Method: Maximization (Algebraic form)

For an LP problem in standard form, the algebraic form of the simplex method involves
considering the augmented form of the LP model and proceeding from initialization step to
optimality test through iterative procedures of moving from on Basic Feasible Solution
(BFS) to another until an optimal solution is reached. To this extent, further movement from
the optimal solution will have a negative impact on the objective function.

In summary, the algebraic form of the simplex method involves the following steps;

1) Initialization step: How the initial corner point feasible solution is selected.
2) Iteration step: The criterion for selecting the defining equation (non-basic variable) to
be replaced by a basic variable. Thus, the following questions arise,
i) How is the replacement identified?
ii) How can a new solution be identified without completely resolving a new system of
defining equations? (The system of equations with the new set of non-basic
variables).
3) Stopping Rule: how is the test of optimality performed?
In-Text Question
Solve the LP problem below using the algebraic form of the simplex method.
Maximize Ƶ = 3x1 + 2x2 _________(1)
 Subject to
2x1 + x2 < 100 _________ (2)
x1 + x2 < 80 _________ (3)
x1 < 40 _________ (4)
x1 , x2 > 0 _________ (5)

Answer to In-text Question

Maximize Ƶ = 3x1 + 2x2 _________(1)
Subject to
2x1 + x2 < 100 _________ (2)
x1 + x2 < 80 _________ (3)
x1 < 40 _________ (4)
x1 , x2 > 0 _________ (5)

The first step is to convert the model to equivalent equality model by introducing slack
variables. This conversion yields the augmented model from which an augmented feasible
solution can be generated, initially and then continue with the other relevant steps.

Introducing slack variables gives the following set of equations:

Maximize Ƶ = 3x1 + 2x2 __________(1)
Subject to
2x1 + x2 + x3 = 100 _____(2)
x1 + x2 + x4 = 80 ______(3)
x1 + x5 = 40 ______(4)
x1, x2 > 0 (5) and (6)

The above model in equation form will correspond to an initial set of feasible solution of the
form (x1, x2, x3, x4, x5, Ƶ), being the augmented solution of the simplex method in
algebraic form or canonical form.

The simplex method can start at any corner point feasible solution. In our example, the non-
basic variables can be set to zero (ie origin) to give an initial solution:
(0, 0, x3, x4, x5, Ƶ) = (0, 0, 100, 80, 40, 0)
Note that, by virtue of the above, each equation has just one basic variable with a coefficient
of +1, and the basic variable does not appear in any other equation. So the solution is not
optimal.

By iterative steps the simplex method moves from current feasible solution to a better
adjacent basic feasible solution. This involves replacing one non-basic variable (called the
entering basic variable) with a new one (called the leaving basic variable).

The above move will definitely increase the size of the objective function, Ƶ, and is
performed by choosing the variable that has the biggest impact on Ƶ. In our example, x1 has
the higher coefficient of 3 and this becomes our first entering basic variable.

The leaving basic variable is the one whose functional constraint imposes the smallest upper
bound on how much the entering basic variable can be increased. In our example, where the
possible candidates are (x3, x4, x5), the determination of the leaving basic variable follows
from calculating the minimum ratio test as follows:

Basic Variable Equation Result of Minimum Ratio Test

100
x3 (2) x3 = 100 -2x1 -x2 x1 < /2 = 50
x4 (3) x4 = 80 – x1 – x2 x1 < 80
x5 (4) x5 = 40 – x1 x1 < 40

From the above table, since x5 imposes the smallest upper bound on x1, the leaving basic
variable is x5. Therefore, x5 becomes a basic variable and its value is set to zero in the new
basic feasible solution.

How can the new basic feasible solution be identified most conveniently?

We can solve for the new values of the remaining basic variables.
We convert the system of equation into the same convenient form we had at the initialization
step: i.e. each equation will have just one basic variable does not appear in any other
equation.
 a) Multiplying an equation by a non-zero constant.
b) Adding/subtracting a multiple of one equation to another equation.

Continuing with the example; our equation of reference is equation (4)

Where x1 + x5 = 40 __________(4)

as x1 has become a basic variable and has a coefficient of +1, we must eliminate it from other
equations including equation (1)
Thus,
New equation (1) = old equation (1) + 3x new equation (4)
= Ƶ – 3x1 – 2x2 + 3 (x1 + x3) = 120
Ƶ – 2x2 + 3x5 = 120 ___________(1)

New (2) = old equation (2) — 2(new equation (4) )

= 2x1 + x2 + x3 — 2(x1 + x5) = 20 i.e. 100-80
= x2 + x3 — 2 x5 = 20 ________(2)

New (3) = old (3) — 1 (new equation (4) )

= x1 + x2 —1 x4 —(x1 + x5) = 40) i.e. 80-40
= x2 + x4—x5 = 40 _________(3)

Thus, our new set of standard equations:

Max Ƶ — 2x2 + 3x5 = 120 _________(1)

S.t
x2 + x3 — 2x5 = 20 __________ (2)
x2 + x4 — x5 = 40 __________ (3)
x1 + x5 = 40___________(4)
While x1, x2 > 0 _________(5) and (6)

The new solution is

(x1, x2, x3, x4, x5, Ƶ) = (40, 0, 20, 40, 0, 120)
Stopping Rule: the objective function equation, equation (1), is checked for optimality. If all
the non-basic variables have negative coefficient, so that increasing any of them would
decrease Ƶ, the current solution is optimal, and the algorithm/method stops.

However, from equation (1), with x5 remaining 0, increasing the value of x2 will have a
further increasing impact on Ƶ, thus the solution is not yet optimal.

We therefore continue to determine the 2nd leaving basic variable.

Basic variable: Equation Result of MRT:

x3 x3 = 20 -x2 + 2x5qq x2 < 20
x4 x4 = 40 – x2 – x5 x2 < 40
x1 x5 = 40 – x5 + 0x2 x2 < ∞

The new leaving basic variable is x3

While the entering basic variable is x2

Thus, new equation (2) becomes:

x2 + x3 – 2x5 = 20 ____________(2)
The new (2) will have no change as the coefficient of x2 is +1

Whereas:
New equation (1) = old equation (1) + 2x (new equation (2) )
= Ƶ – 2x2 + 3x5 + 2 (x2 + x3 – 2x5) = 120 + 40
= Ƶ + 2x3 – x5 = 160 __________(1)

New equation (3) = old equation (3)—1 x (new equation (2) )

= x2 + x4 –x5 –- (x2 + x3 –2x5) = 40 –20
= x4 –x3 –x5 =20 _________(3)

Equation (4) remains unchanged.

 Thus, the new set of equations becomes:
Ƶ + 2x3 –x5 = 160 ___________(1)
x2 + x3 –2x5 = 20 _________(2)
x4 –x3 –x5 = 20___________(3)
x1 + x5 = 40 _________(4)
and x1, x2 > 0 ____________(5), (6)

Now observe that the basic variables x1, x2, and x4 all have coefficients of +1 and appear only
once each in the set of equations.

Secondly, increasing the values of the variables x3 and x5 in the objective function equation
(1) will have no further impact on Ƶ. Thus, the solution is optimal and the algorithm stops
with the following optimal solution:
(x1, x2, x3, x4, x5, Ƶ) = (40, 20, 0, 20, 0, 160)
Meaning:
x1 no of soldiers per week = 40
x2 no of trains per week = 20.

3.2.2 The Simplex Method: Maximization (Tabular form)

The algebraic form of the simplex method presents the opportunity for learning the
underlying logic of the algorithm. Nevertheless, the tabular form is seen as a more convenient
form for performing the required calculations.

The tabular form of the simplex method records only the essential information, namely;
1) The coefficient of the variable.
2) The constants on the right hand sides of the equations.
3) The basic variable appearing in each equation.
By recording the essential elements, only the numbers involved in arithmetic calculations are
highlighted and the computations are recorded more compactly and distinctly.
3.2.2.1 Steps in Solving Linear Programming Problem Using Simplex Method
1. Formulate the problem of statement to a mathematical model in the form
of LP model
2. Convert the LPP into the standard form of LPP. The properties of standard
form are listed below:
 Equations should be used to represent the constraints by
including slack or surplus and/or artificial variables.
 If the right-hand side of any restriction is not already non-
negative, it should be changed by multiplying both sides by -1.
 The objective function should be of the maximization type.
3. Identify the decision factors that will be used in the solution.
4. Identify the replacement variable.
5. Compute new row values for variable entry
6. Modify the remaining rows.
Up until an optimum solution is found, repeat steps III through VI.
The best way to illustrate this process is with the aid of a relevant example.

In-Text Question
Solve the following LPP using Simplex method.
Maximize Z = 12x1 + 16x2
Subject to the following constraints:
10x1 + 20x2 ≤ 120
8x1 + 8x2 ≤ 80
x1 and x2 ≥ 0

Answer to In-text Question

Step 1: The question is already in the right LP model.

Step 2:
To convert the LPP into the standard form of LPP, we will need to note here that the left side
of a constraint of type < or > can be changed into an equation by adding a slack variable (for
maximization problems) or removing a surplus variable (for minimization problems).
Therefore, slacks variable S1 and S2 will be added in the objective function and the
constraints since the problem is a maximization problem.

Now, in the constraint 10x1 + 20x2 ≤ 120, we are going to add a slack S1 ≥ 0 to the left side
to obtain an equation : 10x1 + 20x2 + S1 = 120, S1 ≥ 0
Also, for constraint 8x1 + 8x2 ≤ 80, we are going to add a slack S2 ≥ 0 to the left side to
obtain an equation : 8x1 + 8x2 + S2 = 80, S2 ≥ 0.
i.e.
10x1 + 20x2 + S1 = 120 ...........i
8x1 + 8x2 + S2 = 80...............ii
x1, x2, S1, S2 ≥ 0

The S1, S2 are the slack variables. The slack variables are two because of the two constraints.
Let's symmetrically modify these equations so that S1 and S2 occur in every equation.
10x1 + 20x2 + 1S1 + 0S2 = 120 ......i
8x1 + 8x2 + 0S1 + 1S2 = 80 ...........ii

Let's add slack to the objective function Z as we write it as well.

Maximize Z = 12x1 + 16x2 + 0S1 + 0S2

So we have
Maximize Z = 12x1 + 16x2 + 0S1 + 0S2
10x1 + 20x2 + 1S1 + 0S2 = 120 ......i
8x1 + 8x2 + 0S1 + 1S2 = 80 ...........ii
x1, x2, S1, S2 ≥ 0
Let’s form our first Simplex table using the information above:

Table 1.1 Initial Simplex Table

CBi Cj 12 16 0 0 Solution Ratio

B.V x1 x2 S1 S2
0 S1 10 20 1 0 120
0 S2 8 8 0 1 80
Zj

Let’s explain the initial simplex table (Table 1.1) prepared above

Where Cj refers to coefficient of the objective function

Maximize Z = 12x1 + 16x2 + 0S1 + 0S2,
Cj is 12, 16, 0, 0
B.V is the Basic Variables. From the objective function, we have 2 decision variables and 2
slack variables. Their combination results to four variable (x1, x2, S1, S2).
CBi is the coefficient of basic variable i.e the row gives the coefficients of the variables in the
objective function which will remain the same (the cost of slack variable is zero).

Zj refer to the unit contribution to profits. The subscript j refers to the specific variable being
considered.

To calculate

Z= (CB1 * x11 ) + (CB2 * x12 ) = (0*10) + (0*8) = 0

= (CB1 * x21 ) + (CB2 * x22 ) = (0*20) + (0*8) = 0
Z= (CB1 * S11 ) + (CB2 * S12 ) = (0*1) + (0*0) = 0
= (CB1 * S21 ) + (CB2 * S22 ) = (0*0) + (0*1) = 0

For the ZJ in the Solution column of the table, (0*120) + (0*80) = 0

In the initial column all the Zj - values will be zero since the CBi values are zero and no real
product is being manufactured, hence there is no gross profit to be lost if they are replaced.
Table 1.2 Initaial Simplex table showing Zj
CBi Cj 12 16 0 0 Solution
B.V x1 x2 S1 S2
0 S1 10 20 1 0 120

0 S2 8 8 0 1 80
Zj 0 0 0 0 0

Cj - Z j 12 16 0 0

Key Column
Maximum value

The Zj values under the solution column (non basic variables) is the total profit for their
solution.
The Maximum value here is the variable (automatically a non-basic variable) with the
negative coefficient having the largest absolute value (i.e. the most negative coefficient) in
objective function equation.
Key Column can also be referred to as the Pivot Column.

Step 3: Identify the decision factors that will be used in the solution.

Now, for optimality condition,

For Max; all Cj - Zj ≤ 0 ( or Zj - Cj ≥ 0 )

For Min; all Cj - Zj ≥ 0 ( or Zj - Cj ≤ 0 )

Where Cj - Zj represents the net change in the value of the objective function. The row it
appears on the table is referred to as the Net evaluation row.

Since our problem is a maximization problem, all Cj - Zj ≤ 0. The values for Cj - Zj in Table
1.2 are 12, 16 , 0 and 0. 12 and 16 are not satisfying the condition because they are greater
than zero. This means we will have to repeat the steps until Cj - Zj satisfy the optimality
condition for Max. To proceed further, we will have to select the negative coefficient having
the largest absolute value from Cj - Zj and that is 16. The column where the negative
coefficient having the largest absolute value (16) appears becomes the PIVOT
COLUMN/KEY COLUMN (COLUMN X2). Column X2 will replace a value in the basic
variable but we need to determine the value to be replaced.

Step 4: Identify the replacement variable.

Now, after finding the key column, the next thing to do is to find the replacement variable
(leaving basic variable) and to do this, we need to find the minimum ratio. The steps below
show how to achieve the minimum ratio:
1) Pick out each coefficient in the Pivot column that is strictly positive (> 0)
2) Divide each of these coefficients into the right side entry for the same row.
3) Identify the row that has the smallest of these ratios.
4) The basic variable for that row is the leaving basic variable, so replace that variable
by entering basic variable in the basic variable column of the next simplex tableau.
To get the minimum ratio using our table,
Ratio = Solution column / Key column.
Here, solution column is the right side entry in our table.
Table 1.3 Initial Simplex table showing ratio.
CBi Cj 12 16 0 0 Solution Ratio
B.V x1 x2 S1 S2
0 S1 10 20 1 0 120 120/20 = 6

0 S2 8 8 0 1 80 80/8 = 10

Zj 0 0 0 0

Cj - Z j 12 16 0 0

Key Element
Minimum value
Or Pivot Number Pivot/Key
Row
P
To get the KEY ROW to be replaced, get the minimum value from ratio i.e. 6. The row
where the minimum value (6) appears becomes the KEY/PIVOT ROW (ROW S1). The
interception point (Value) of the Key column and Key row is referred to as the KEY
ELEMENT/ PIVOT NUMBER. The key element from table 1.3 is 20. This means X2 will
replace S1.
Step 5: Compute new row values for variable entry
This step will be achieved by using elementary row operations (multiply or divide a row by a
non-zero constant) add or subtract a multiple of one row to another row) to construct a new
simplex tableau in proper form from Gaussian elimination below the current one, and then
return to the optimality test.

Back to our table, if we introduce X2 into to the basic variables, the entire S1 will be replaced.
Each value now in the S1 row is divided by the value in column X2 in the same row (key
element (20)) to generate the values of X2, the replacing row.

The row values entering variable X2 can be calculated as follows:

10 / 20 = 1/2
20 / 20 = 1
1 / 20 = 1/20
0 / 20 = 0
120/20 = 6

Step 6: Modify the remaining rows.

The only row remaining here after the other row is placed by X2 is S2.

Modifying S2 = old S2 value – ((corresponding key column value *corresponding key row
value) / key element).

8- ((8*10) / 20) = 4
8- ((8*20) / 20) = 0
0- ((8*1) / 20) = -2/5
1- ((8*0) / 20) = 1
80- ((8*120) / 20) = 32
Now, let us rewrite the initial simplex table to become the second simplex table.
Table 1.4 First Iteration
CBi Cj 12 16 0 0 Solution Ratio
B.V x1 x2 S1 S2
16 X2 1/2 1 1/20 0 6

0 S2 4 0 -2/5 1 32
Zj 8 16 4/5 0

Repeating the same procedures from steps from 3 to 6. The final outcome for the iteration
will be:

Table 1.5 First Iteration showing ratio

CBi Cj 12 16 0 0 Solution Ratio

B.V x1 x2 S1 S2
16 X2 1/2 1 1/20 0 6 6/(1/2) = 12

0 S2 4 0 -2/5 1 32 32/4 = 8
Zj 8 16 4/5 0

Cj - Z j 4 0 -4/5 0

Minimum value
Key Column
Maximum value Key Row Key Element

Since a value in Cj - Zj (4) is higher than zero, it doesn’t satisfy the optimality condition. This
means we will go further to iterate by following the procedures.
Note: Key Row can also be referred to as the Pivot Row; Key Column as Pivot Column

Table 1.6 Second Iteration

CBi Cj 12 16 0 0 Solution
B.V x1 x2 S1 S2
16 X2 0 1 1/10 -1/8 2

12 X1 1 0 -1/10 1/4 8
Zj 12 16 2/5 1 128

Cj - Z j 0 0 -2/5 -1

From Table 1.6, you will realise that all the values in Cj - Zj satisfies the optimality condition,
therefore our optimum solution to the problem is:

X1 = 12 , X2 = 16 and Z = 128

3.2.2.2 Tie Breaking in the Simplex Method.

1) Tie for the Entering Basic Variable (EBV): This is where two or more non-basic
variables are tied for having the largest negative coefficient (in absolute terms). This
tie is broken by arbitrarily selecting any of non-basic variables, the optimal solution
will be reached eventually.

2) Tie for the Leaving Basic Variable (LBV) –Degeneracy: This condition is regarded as
degenerate and has profound consequences, creating problems for the identification of
the LBV when the basic variables reach zero simultaneously as the entering basic
variable is increased.

3) No Leaving Basic Variable –Unbounded Ƶ: This happens when the entering basic
variable could be increased indefinitely without giving negative values to any of the
current basic variables. In the Tabular form, this means that every coefficient in the
Pivot column (excluding row 0/1) is either negative or zero. Normally, the simplex
method would stop with a message that the objective function is unbounded, as LP has
not discovered ways of making infinite profits, as in real life. This is effect of a model
 that has been misformulated, either by omitting relevant constraints or by stating them
incorrectly. Or it might be a serious computation error.

4) Multiple Optimal Solutions: A LP problem can have multiple optimal solutions the
simplex algorithm should normally stop when are solution is reached.

3.2.3 The Simplex Method: Minimization

The simplex method, as we have seen, is used effectively for solving LP problems, which
entails either maximization or minimization of the objective function. We have discussed the
maximization problems and the various forms it can be solved. In this section, we are going
to look at minimization problems.

For the purposes of solving minimization problems using the simplex method, we note that
there are two types of minimization problems, which we denote as type A and type B
respectively.
Type A: The first type of standard minimization problem is one in which:
1) The objective function is to be minimized. Max (Z) = -Min (-Z) i.e. P = -C
2) All variables involved in the problem are nonnegative, and
3) All other linear constraints may be written so that the expression involving the
variables is less than (<) or equal to a nonnegative constant.

Type B: The second type is quite similar to the first type, but with a significant
difference:
1) The objective function is to be minimized. Max (Z) = -Min (-Z) i.e. P = -C
2) All the variables involved in the problem are nonnegative, and
3) All other linear constraints may be written so that the expression involving the variable
is greater than or equal to ( > ) a constant.
In-Text Question
1) With respect to type A:
Minimize C = - 2x1 + 3x2
Subject to
3x1 + 4x2 < 24
7x1 – 4x2 < 16
x1, x2 >0
 Answer to In-text Question

Let P = - C
Then P = 2x1- 3x2 _____________(0)
We seek to maximize P subject to the constraint:
Step 1: 3x1 + 4x2 + x3 = 24 ______(1)
7x1—4x2 + x4 = 16 ______(2)
Where x3 and x4 are slack variable/basic.

This implies that –2x 1 + 3x2 + P = 0

Next:
We consider the above standard LP problem/model in matrix form, another way of
representing LP models and solving them:

Equation Basic Variable Coefficient of

no x1 x2 x3 x4 P RHS
1 x3 3 4 1 0 0 24
2 x4 7 -4 0 1 0 16
0 P -2 3 0 0 1 0

Please observe that the above table is simply used to identify the various coefficients
and variables. The matrix form is:

3 4 1 0 0 24
7 -4 0 1 0 16
- 2 3 0 0 1 0

The optimal solution has not been reached:

Next iteration.

3 4 1 0 0 24
7 -4 0 1 0 16
- 2 3 0 0 1 0
 Next Step:

3 +4 1 0 0 24 —3R2 + R1 R1
7 -4 0 1 0 16
- 2 3 0 0 1 0 -
1
 /7R2 R2, 2R2 +R3 R3

40 -3 120
0 /7 1 /7 0 /7
-4 1 16
1 /7 0 /7 0 /7
13 2 32
0 /7 0 /7 1 /7

The above solution is optimal.

Thus, P is maximized at 32/7 when x1 = 16/7
and x2 = 0. Therefore, C is minimized at -32/7, when x1 = 16/7 and x2 = 0

In- Text Question 2

With respect to type B;
In this case, we notice that from characteristics (3) that the direction of the inequality in the
constants has changed, and also the constant involved in the constraint may not be
nonnegative.

In order to solve any minimization problem in this category, the following steps are generally
considered.

For a case of

Minimize C= C1 x1 + C2 x2 + . . . + Cn xn
Subject to.
a11 x1 + a12x2 + … + a1n xn ≥ b1
a21 x1 + a22x2 + … + a2n xn ≥ b2
 am1 x1 + am2x2 + … + amn xn ≥ bm

Where xi ≥ 0 for all i

Answer to In-text Question 2

1) The first thing to do is to form the augmented matrix for the system of inequalities,
and add a bottom row consisting of the coefficient of the objective function.

a11 a12 . . . a1n b1

a21 a22 . . . a2n b2

am1 am2 amn bm

C1 C2 …… Cn

2) The next is to determine and form the transpose of the matrix.

a11 a21 . . . am1 C1

a12 a22 . . . am C2

a1n a2n . . . amn Cn

b1 b2 . . . bm O

3) The above transpose of the original matrix provides basis to form the dual (read
duality theory) maximization problem, corresponding to the transposed matrix.
Therefore, the task becomes to find the maximum of the objective function P, thus;
Maximize P = b1 y1 + b2 y2 + . . . + bmym.
Subject to:
a11 y1+ a21 y2 + . . . + am1 ym < C1
a12 y1 + a22 y2 + . . .+amzym < C2

a1n y1 + azn y2 + . . . +amn ym < Cn

 Where yi ≥ 0 for all i

4) Then apply the simplex method for standard maximization problem. The maximum
value of P will be the minimum value of C. Moreover, the values of x 1, x2, … , xn will
occur in the bottom row of the final matrix in the columns corresponding to the slack
variables.

In-Text Question 3

Minimize C = 2x1 + 5x2

Subject to
X1 + 2x2 > 6
3x1 + 2x2 > 6
x1, x2 > 0

Answer to In-text Question 3

This is our type B minimization problem. First, we form the coefficient matrix and take its
transpose. Then, we interpret the transpose as a standard maximization problem. Thus,
We have:

T
1 2 6 1 3 2

3 2 6 = 2 2 5

2 5 C 6 6 C
Therefore, we seek to maximize.
Maximize C = 6y1 + 6y2
Subject to
y1 + 3y2 ≤ 2
2y1 + 2y2 ≤ 5
y1, y2, ≥ 0
Step 1: Removing the inequality by using slack variables:

y1 + 3y2 + y3 = 2
2y1 + 2y2 + y4 = 5

Step 2: - 6y1 - 6y2 + C = 0

Step 3: 1 3 1 0 0 2

2 2 0 1 0 5

-6 -6 0 0 1 0

Step 4: Optimal solution has not been reached.

Step 5: 1 3 1 0 0 2

2 2 0 1 0 5

-6 -6 0 0 1 0

Step 6: 1 3 1 0 0 2

2 2 0 1 0 5 6R1 + R3 R3
Next
Page
-6 -6 0 0 1 0 -2R1 + R2 R2

1 3 1 0 0 2

0 -4 -2 1 0 1

0 12 6 0 1 12

From the above matrix, the optimal solution has been reached.
Recall that the final matrix is interpreted in a special way. C is minimized at 12 and it occurs
when x1 = 6, x2 = 0, and Ƶ = 0

3.2.4 Two Phase Simplex Method

The two-phase simplex approach can be used as an alternative to the Big M method when a
Basic Feasible Solution (BFS) is not readily available. The Two-Phase Simplex Method is so
named because the LPP solution is computed in two phases. Phase I of the solution involves
the process of removing artificial variables (finding the BFS), and phase II is used to arrive at
an optimal solution (problem is solved using simplex method).

Phase I - A final simplex table containing a basic feasible solution to the original problem is
produced in this phase after the simplex approach has been applied to a specifically designed
auxiliary linear programming problem.

Steps in Phase I
1. Set all other variables in the objective function a cost of 0, while giving each
artificial variable a cost of 1.
2. Create the Auxiliary LPP with the new objective function Z* being maximized
while considering the provided set of constraints.
3. Use the simplex approach to solve the auxiliary problem until one of the following
three scenarios occurs:
I. Max Z* < 0 and atleast one artificial vector appear in the optimum basis at a
positive level (Δj ≥ 0). In this case, given problem does not possess any
feasible solution.
II. Max Z* = 0 and at least one artificial vector appears in the optimum basis at a
zero level. In this case proceed to phase-II
III. Max Z* = 0, and there are no artificial vectors in the optimum basis. Move to
phase II in this situation as well.

Phase II - Now give each artificial variable that appears in the basis at level zero a cost of
zero and the actual cost to the variables in the objective function. Now, using the simplex
method within the provided set of constraints, this new objective function is maximized.
The modified simplex table acquired at the conclusion of phase-I is subjected to the simplex
method until an optimal basic feasible solution is reached. At the conclusion of phase I, the
artificial variables that are non-basic are eliminated.

Example
Max Z = 3x1 - x2
Subject to
2x1 + x2 ≥ 2
x1 + 3x2 ≤ 2
 x2 ≤ 4
and x1 ≥ 0, x2 ≥ 0

Solution
Standard LPP
Max Z = 3x1 - x2
Subject to
2x1 + x2 – s1+ a1= 2
x1 + 3x2 + s2 = 2
x2 + s3 = 4
x1 , x2 , s1, s2, s3,a1 ≥ 0
Auxiliary LPP
Max Z* = 0x1 - 0x2 + 0s1 + 0s2 + 0s3 -1a1
Subject to
2x1 + x2 – s1+ a1= 2
x1 + 3x2 + s2 = 2
x2 + s3 = 4
x1 , x2 , s1, s2, s3,a1 ≥ 0

Cj → 0 0 0 0 0 -1
Basic Min ratio
Variables CB XB X1 X2 S1 S2 S3 A1 XB /Xk
a1 -1 2 2 1 -1 0 0 1 1→
s2 0 2 1 3 0 1 0 0 2
s3 0 4 0 1 0 0 1 0 -
↑

Z* = -2 -2 -1 1 0 0 0 ←Δj
x1 0 1 1 1/2 -1/2 0 0 X
x2 0 1 0 5/2 ½ 1 0 X
x3 0 4 0 1 0 0 1 X

Z* = 0 0 0 0 0 0 X ←Δj
Since all Δj ≥ 0, Max Z* = 0 and no artificial vector appears in the basis, we will move on to
phase II.

Phase II
Cj → 3 -1 0 0 0
Basic Min ratio
Variables CB XB X1 X2 S1 S2 S3 XB /Xk
x1 3 1 1 ½ -1/2 0 0 -
s2 0 1 0 5/2 1/2 1 0 2→
s3 0 4 0 1 0 0 1 -
↑

Z=3 0 5/2 -3/2 0 0 ←Δj

x1 3 2 1 3 0 1 0
s1 0 2 0 5 1 2 0
s3 0 4 0 1 0 0 1

Z=6 0 10 0 3 0 ←Δj

Since all Δj ≥ 0, optimal basic feasible solution is obtained

Therefore the solution is Max Z = 6, x1 = 2, x2 = 0

3.3 Summary of Unit 3

1. A linear programming model with two or more choice variables can be solved using
simplex approaches.
2. The simplex method is an efficient and reliable algorithm for solving linear
programming problems. It also provides the basis for performing the various parts of
post-optimality and sensitivity analysis very efficiently.
3. Although it has a useful geometric interpretation, the simplex method is an algebraic
procedure. At each iteration, it moves from the current BF solution to a better adjacent
BF solution by choosing both an entering basic variable and a leaving basic variable
and then using Gaussian elimination to solve a system of linear equations. When the
current solution has no adjacent BF solution that is better, the current solution is
optimal and the algorithm stops.
4. The simplex method can be presented in algebraic form, tabular, form, and in matrix
form. Computer implementations of the simplex method and its variants have become
so powerful that they now are frequently used to solve linear programming problems
with many hundreds of thousands of functional constraints and decision variables, and
occasionally vastly larger problems.

3.4 Self Assessment Questions (SAQs) for Study Unit 3

Now that this study session is over, you can assess how well you met its learning objectives
by responding to the following questions:

SAQ 3.1 (Tests learning outcome 3.1)

From the statements made below, identify the TRUE/FALSE and fill the gap where
applicable;
1. ..................................... variables are additional variables added to the linear constraints of
a linear program in other to change inequality constraints into equality constraints.

2. It is possible to reverse the dual L.P.P to primal L.P.P.

3. A solution with the fewest possible non-zero variables is known as a ..................................

4. Dual L.P.P always reduces the amount of computation.

5. The integer that lies at the point where the key column and key row cross is known as
...........................................

6. The type ≤ inequalities are transformed into equations by the introduction of slack variables.

7. Artificial variables don't need to be added in the maximization LPP.

SAQ 3.2 (Tests learning outcome 3.2)

What are the procedures needed to use a simplex approach

SAQ 3.3 (Tests learning outcome 3.3)

1. Max Ƶ = 12x1 + 3x2 + x3

S.t

10x1 + 2 x2 + x3 < 100

7 x1 + 3x2 + 2x3 < 77

2x1 +4x2 + x3 < 80

and x1 > 0, x2 > 0, x3 > 0

a) Work through the linear programming problem above using simplex method.
References and Further Reading

1. Operations research by P. Rama Murthy (2007), 2nd Edition. New age international
(p) Ltd.
2. Operations Research. By Prem Kumar Gupta and D.S. Hira
Notes/Answers to Module One Self Assessment Questions
Answers to Unit 1
SAQ 1.1.

1. TRUE

2. Operation Research

3. Model

4. FALSE

SAQ 1.2.

1. Operation Research addresses problems like scheduling, routing, project planning,

automation, supply chain management, network optimization and transportation.
2. Finding the optimal or ideal answer for decision-making while keeping in mind that there
exist restrictions is the major of operation research.

SAQ 1.3.

1.

I. Determine the problem of interest and assemble pertinent data.

This is the initial stage of the OR. It involves developing a problem statement and having a
firm understanding of the system / problem to be worked on.
II. Prepare a mathematical model that fully represents the problem.
This is the next step after understanding the problem. It involves the development of a
mathematical model that adequately captures the nature of the problem.
III Formulate a procedure on how solutions to the problem can be obtained using the
developed model.
This is a common concept in OR. It involves locating the optimal or best solution to that issue
using the already prepared model.
2.
I. Linear Programming
II. Game Theory
III. Dynamic Programming
IV. Queuing Theory
V. Simulation.
SAQ 1.4.

1. A model in OR is an idealized depiction of a scenario in real life. It can depicts one or

more parts of reality.
2. Classification by Structure.
3. A mathematical model is a system description in mathematical terms. A set of variables
and a set of equations that construct relationships between the variables are typically used to
describe a system.

Answers to Unit 2

SAQ 2.1.

1. Linear Programming

2. FALSE

3. TRUE

4. FALSE

5. Feasible Region

6.Constraints.

SAQ 2.2.

Maximize Ƶ = 3x1 + 2x2

Subject to (St)
2x1 + x2 < 100
x1 + x2 < 80
x1 < 40
x1 > 0
x2 > 0

SAQ 2.3.

1. The feasible region is the set of locations that satisfy each inequality in a collection of
inequalities.
2
3x1 + 2x2 = 18
x2
10

8
6 2x2 = 12

4 Feasible
2 region

6 x1
4

After constructing lines respectively equivalent to 2x2 < 12 ( i.e. x2 < 6), and that of 3x1 +
2x2 < 18, which requires plotting the points (x1, x2) such that 3x1 + 2x2 = 18, another line to
complete the boundary.

Note that the point such that 3x1 + 2x2 < 18 are those that lie either underneath or on the line
3x1 + 2x2 =18, so this is the limiting line above which points do not satisfy the inequality.

Answers to Unit 3

SAQ 3.1.

1. Slack

2. TRUE

3. Basic Feasible Solution

4. FALSE

5. Key Element

6. TRUE

7. TRUE

SAQ 3.2.

1. Formulate the problem of statement to a mathematical model in the form of LP model.

2. Convert the LPP into the standard form of LPP.

3. Identify the decision factors that will be used in the solution.

4. Identify the replacement variable.
5. Compute new row values for variable entry
6. Modify the remaining rows.
Up until an optimum solution is found, repeat steps III through VI.

SAQ 3.3.

Max Ƶ = 12x1 + 3x2 + x3

S.t

10x1 + 2 x2 + x3 < 100

7 x1 + 3x2 + 2x3 < 77

2x1 +4x2 + x3 < 80

and x1 > 0, x2 > 0, x3 > 0

Slack variables x4 ≥ 0, x5 ≥ 0, x6 ≥ 0 will be introduced.

Max Ƶ = 12x1 + 3x2 + x3 + x4 + x5 + x6

S.t

10x1 + 2 x2 + x3 + x4 = 100

7 x1 + 3x2 + 2x3 + x5 = 77

2x1 +4x2 + x3 + x6 = 80

x1, x2, x3, x4, x5, x6 ≥ 0

Let’s present the initial simplex table.

Initial Simplex Table
CBi Cj 12 3 1 0 0 0 Solution
B.V x1 x2 x3 x4 x5 x6
0 x4 10 2 1 1 0 0 100
0 x5 7 3 2 0 1 0 77
0 x6 2 4 1 0 0 1 80
Zj 0 0 0 0 0 0
Zj - Cj -12 -3 -1 0 0 0
Second Simplex Table

CBi Cj 12 3 1 0 0 0 Solution
B.V x1 x2 x3 x4 x5 x6
12 x1 1 1/5 1/10 1/10 0 0 10
0 x5 0 16/10 13/10 -7/10 1 0 7
0 x6 0 18/5 4/5 -1/5 0 1 60
Zj - C j 0 -3/5 1/5 6/5 0 0

Final Simplex Table.

CBi Cj 12 3 1 0 0 0 Solution
B.V x1 x2 x3 x4 x5 x6
12 x1 1 0 -1/16 3/16 -1/8 0 73/8
3 x2 0 1 13/16 -7/16 5/8 0 35/8
0 x6 0 0 -17/8 11/8 -9/4 1 177/4

Zj - C j 0 0 11/16 15/16 3/8 0

At this stage, all the Zj - Cj ≥ 0, therefore, we obtain from the table our optimum solution as
x1 = 73/8, x2 = 35/8 and x6 = 177/4. Substituting it into our objective function Max Ƶ = 12x1 +
3x2 + x3
The maximum function will be 981/8.