Yanet College

Department of Management
MBA Program

Quantitative Analysis for Management Decision Making (MBA 631)

1
By: Nuru Mohammed (Ph.D.)

2021/2022
Asella, Ehiopia
Nuru M.((Ph.D.)
 2

Chapter 2
Linear Programming (LP)
Chapter Objectives:
At the end of this chapter you should be able to;
Understand fundamental concepts and general structure of a LPM.
Express objective function and resource constraints in LP model in terms of DV and parameters.
Appreciate the limitations and assumptions of LP technique.
solve an LP problem by the graphical method
Set-up Nuru
simplex
M.((Ph.D.) tables and solve LPP using the simplex algorithm.
 2.1 Introduction
3
 Linear: refers to linear relationship among variables in a model.
 Programming: refers to the mathematical modelling and solving of a
problem that involves the use of limited resources.
 LPM: mathematical modelling technique regarding optimal use of
scarce resources.
 Scarce resources: refers to resources that are not available in infinite quantity.
 In 1947 (WW-II), George B Dantzing from US Air Force, developed LPM.
 To solve military logistics problems.
 Now, it is extensively being used in all functional areas of management

Nuru M.((Ph.D.)
 Continued…
e.g; Product mix problem - Beaver Creek Pottery Company

Resource Requirements
Labor Clay Profit
product
(Hr./Unit) (Lb./Unit) ($/Unit)
Bowl 1 4 40
Mug 2 3 50

Figure 2.1 Beaver Creek Pottery Company

How many bowls and mugs should be produced to

maximize profits given labor and materials constraints?
Product resource requirements and unit profit:
LP Model Formulation 1 (3 of 3)
Complete Linear Programming Model:

Maximize
Z = $40 x1 + $50 x2
subject to: 1x + 2 x  40
1 2

4 x2 + 3 x2  120
x1, x2  0
 2.1 Characteristics of LP
6
 LPM is characterized by its structure and model assumptions
A. Structure of LPM
Consists of following three basic components.
1. Decision variables: represent the unknown quantities to be solved
for.
If the decision-maker control the values, the variables are said to be
controllable, otherwise they are said to be uncontrollable.
In LPM all decision variables are continuous, controllable and nonnegative
Usually denoted by 𝑋1 , 𝑋2 , 𝑋3 , …, 𝑋𝑛
2. The objective function: mathematical representation of the overall
goal
 Expressed in terms of decision variables.
Nuru M.((Ph.D.)

 It could be either maximization or minimization.

 Continued…
7
 The constraints: limitations (restrictions) on the use of resources that limit the
degree to which an objective can be achieved.
 Expressed as linear equalities or inequalities in terms of DV.
 It can be system, individual, and non-negativity

Nuru M.((Ph.D.)
 Continued…
8
B. Assumptions of an LPM
1. Certainty: all parameters in the model must be known and constant.
 Availability of resources, profit (or cost) contribution per unit of DV, and consumption
of resources per unit of decision.
2. Additivity:
 The value of the objective function must be equal to the sum of the respective
individual contribution (profit or cost).
 The total amount of each resource used, must be equal to the sum of respective
resources used for each product.
3. Linearity (or proportionality):
 Each DV has a linear impact on the objective function.
 Each DV has a linear impact on the available resource (RHS).
4. Divisibility (or continuity): The solution of DV are allowed to assume continuous values.
Nuru M.((Ph.D.)
 non-integer values are acceptable.
 Continued…
9
5. Non-negativity - negative values of variables are unrealistic.

Nuru M.((Ph.D.)
 2.2 Advantages and Limitations of LP technique
10

A. Advantages : LP technique helps;

 Decision-makers to use their productive resources effectively
 To make more objective decision.
 To arrive at optimal solution of a decision problem.
 Highlight bottlenecks in the production processes.
B. Limitation
 Assumes linear relationships among decision variables.
 Some times non-integer values are meaningless.
 Does not take time and uncertainty into consideration.
 In real-life, Parameters are neither known nor constant.
 Deals with only single objective.

Nuru M.((Ph.D.)
 2.3 General mathematical model of LP problem
11

 The general LPM with “n” DV and “m” constraints can be stated;

Nuru M.((Ph.D.)
 2.4 Guidelines on LPM formulation
12
Step 1. Identification of decision variables.
Step 2. Determination of objective function.
Step 3. Identification of constraints.
Step 4. Determining the parameters of constraints or objective function.
Step 5. Building and validating the models.

Nuru M.((Ph.D.)
 Continued…
13

 Example 2.1 A manufacturing company is engaged in producing three types of

products: A, B and C. The production department produces, each day,
components sufficient to make 50 units of A, 25 units of B and 30 units of C. The
management is confronted with the problem of optimizing the daily production
of the products in the assembly department, where only 100 man-hours are
available daily for assembling the products. The following additional information
is available:

The company has a daily order commitment for 20 units of products A and a
total of 15 units of products B and C. Formulate this problem as an LP model so
as to maximize the total profit.

Nuru M.((Ph.D.)
 Continued…
14
 Soln
Z max (Profit)= 12𝑋1 +20𝑋2 +45𝑋3
s.t
0.8𝑋1 +1.7𝑋2 +2.5𝑋3 < 100
𝑋1 < 50
𝑋2 < 25
𝑋3 < 30
𝑋1 > 20
𝑋2 +𝑋3 > 15
𝑋1 , 𝑋2 , 𝑋3 > 0

Nuru M.((Ph.D.)
 Continued…
15
 Example 2.2: A farmer is preparing to plant a crop in the spring and needs to fertilize a
field. There are two brands of fertilizer to choose from, Super-gro and Crop-quick. Each
brand yields a specific amount of nitrogen and phosphate per bag, as follows:

The farmer’s field requires at least 16 pounds of nitrogen and at least 24 pounds of
phosphate. Super-gro costs $6 per bag, and Crop-quick costs $3. The farmer wants to know
how many bags of each brand to purchase to minimize the total cost of fertilizing.

Nuru M.((Ph.D.)
 Continued…
16
 Soln
Z minimize= 6𝑋1 +3𝑋2
s.t
2𝑋1 +4𝑋2 >16
4𝑋1 +3𝑋2 >24
𝑋1 , 𝑋2 , > 0

Nuru M.((Ph.D.)
 Continued…
17
Exercise!
 Example 2.2 A small scale manufacturer produce two types of products A
and B which gives a profit of 2000 and 3000 birr respectively. Each product is
produced on two machine I and II. Type A requires one hour processing on
machine I and two hour on machine II. Type B takes one hour in both
machines. Machine I is available for not mortmain 60 hours. While machine II
for 100 hours during any working week. The demand for product A does not
exceed 30 unit per week. Formulate the LPM?

Nuru M.((Ph.D.)
 Continued…
18
 Example 2.3 International Wool Company operates a large farm on which sheep
are raised. The farm manager determined that for the sheep to grow in the
desired fashion, they need at least minimum amounts of four nutrients (the
nutrients are nontoxic so the sheep can consume more than the minimum without
harm). The manager is considering three different grains to feed the sheep. Table
B-2 lists the number of units of each nutrient in each pound of grain, the minimum
daily requirements of each nutrient for each sheep, and the cost of each grain.
The manager believes that as long as a sheep receives the minimum daily amount
of each nutrient, it will be healthy and produce a standard amount of wool. The
manager wants to raise the sheep at minimum cost. Formulate LPM?

Nuru M.((Ph.D.)
 2.5 Solving LP Problems
19

 LP model can be stated in terms of two decision variables or multiple

decision variables.
 When LP problem involve only two products, it can be solved by
graphical procedure, i.e., two dimensional graphs.
 If a LP problem involves a multi product case, a more advanced
technique, i.e., simplex (algebraic) technique can be applied.

Nuru M.((Ph.D.)
 3.5.1 Graphical Method of Solving LPP
. Extreme Point Solution Method
20
. Iso-profit/cost method

 Used to solve any LP problem with only two decision variables.

 It provides a conceptual basis for solving large and complex LPP.

Nuru M.((Ph.D.)
 Continued…
21
IMPORTANT DEFINITIONS

Solution: Set of values of DV that satisfy the constraints of an LPP.

Feasible solution: Set of values of DV that satisfy all the constraints and non-negativity.
Infeasible solution: Set of values of DV that don’t satisfy all the constraints and non-
negativity.
Basic solution: For a set of “m” simultaneous equations in “n” variables (n > m) in an LP
problem, a solution obtained by setting (n – m) variables equal to zero.
Basic feasible solution: A feasible solution to an LP problem which is also the basic
solution.
i. Degenerate: if the value of at least one basic variable is zero.
ii. Non-degenerate: if value of all “m” basic variables is non-zero and positive.

Nuru M.((Ph.D.)
 Continued…
22
Optimum basic feasible solution: A basic feasible solution that optimizes
(maximizes or minimizes) the objective function.

Unbounded solution: A solution that can increase or decrease infinitely the value
of the objective function.

Nuru M.((Ph.D.)
 Continued…
23
Steps
1. Develop an LPM
2. Plot constraints on graph paper and decide the feasible region.
3. Examine extreme points of the feasible solution space to find an optimal
solution

Extreme or Corner Point Theorem

It states that for problems that have optimal solutions, a solution will occur at the corner point in the
case of unique solution, while in the case of multiple solutions, at least one will occur at a corner point as
these multiple solutions will be combinations of those points between two corner points.

Nuru M.((Ph.D.)
 Continued…
24
 Example 3.1 Use the graphical method to solve the following LP problem
𝑀𝑎𝑥𝑖𝑚𝑖𝑧𝑒 𝑍 = 15𝑥1 + 10𝑥2
S.t
4𝑥1 + 6𝑥2 ≤ 360
3𝑥1 ≤ 180
5𝑥2 ≤ 200
𝑥1 , 𝑥2 ≥ 0

Nuru M.((Ph.D.)
 Continued…
25 Example 3.1 Use the graphical method to solve the following LP problem

a) 𝑀𝑎𝑥𝑖𝑚𝑖𝑧𝑒 𝑍 = 15𝑥1 + 10𝑥2 b) Minimize 𝑍 = 3𝑥1 + 2𝑥2

S.t S.t
4𝑥1 + 6𝑥2 ≤ 360 5𝑥1 + 𝑥2 ≥ 10
3𝑥1 ≤ 180 𝑥1 + 𝑥2 ≥ 6
5𝑥2 ≤ 200 𝑥1 + 4𝑥2 ≥ 12
𝑥1 , 𝑥2 ≥ 0 𝑥1 , 𝑥2 ≥ 0

Exercise
C) 𝑀𝑎𝑥𝑖𝑚𝑖𝑧𝑒 𝑍 = 7𝑥1 + 3𝑥2
S.t
𝑥1 + 2𝑥2 ≥3
𝑥1 + 𝑥2 ≤ 4
𝑥1 ≤ 5/2
Nuru M.((Ph.D.) 𝑥2 ≤ 3/2
𝑥1 , 𝑥2 ≥ 0
 Continued…
26
a)

Nuru M.((Ph.D.)
Hence the optimal solution to the given LP problem is: x1 = 60, x2 = 20 and Max Z = 1,100.
 Continued…
27
b)

 Hence, the optimal solution to the given LP problem is: x1 = 1, x2 = 5, and Min Z = 13.

Nuru M.((Ph.D.)
 Special case in LP
(Assignment)
28

 LP has the following Special cases;

1. Alternative or Multiple solution
2. Unbounded solution
3. Infeasible solution
4. Redundancy

Nuru M.((Ph.D.)
 3.5.2 Linear Programming: The Simplex Method
29
 Most real-life problems LPM have more than two variables.
 Simplex method is appropriate.
Iterative method.
Examines corner points, using matrix row operations, until an optimal
solution is found.

Nuru M.((Ph.D.)
 LPM Standard form
30
 The problem be converted into its standard form.
 Constraints should be converted into equality, using slack or surplus
and/or artificial variables.
 The RHS of each constraint should be made non-negative.
 The objective function should be of the maximization type.

Nuru M.((Ph.D.)
 3.5.2.1 Simplex algorithm (maximization case)
31
 Steps to follow:
1. Formulation of the mathematical model
2. Develop the standard form of the model.
3. Setup the initial feasible solution
4. Test for optimality: Calculate the cj – zj value for all non-basic variables, if
i. cj – zj ≤ 0, then the basic feasible solution is optimal.
ii. at least one column of the coefficients matrix for which ck – zk > 0 and all
other elements are negative (i.e. 𝑎𝑖𝑘 < 0), then there exists an unbounded
solution.
iii. at least one cj – zj > 0, and each of these columns have at least one positive
element (i.e. 𝑎𝑖𝑗 ) for some row, then this indicates that an improvement is
possible.
Nuru M.((Ph.D.)
 Continued…
32

5. Select entering variable:

The
a variable that has the largest cj – zj value (key or pivot column). intersection
is called
6. Select leaving Variable: key or pivot
a variable with smallest non-negative RR (key or pivot row) element

7. Finding the new solution

 Apply the elementary row operation to transform the pivot element into 1
and other pivot column element into 0.
8. Repeat the procedure from step 4-7

Nuru M.((Ph.D.)
33

Flow Chart of
Nuru M.((Ph.D.)
Simplex Algorithm
 Continued…
34

 e.g: Solve the following LPP.

Maximize Z = 3𝑋1 + 5𝑋2 + 4𝑋3
S.t
2𝑋1 + 3𝑋2 ≤ 8
2𝑋2 + 5𝑋3 ≤ 10
3𝑋1 + 2𝑋2 + 4𝑋3 ≤ 15
𝑋1 , 𝑋2 , 𝑋3 ≥ 0

Note:
Elementary raw operations
1. Multiply or divide the entire raw by non-zero constant.
2. Add or deduct the multiple of one raw with/from other row
3. Interchange one raw with other.
Nuru M.((Ph.D.)
 Continued…
35
Step 2: standard form Step 3. initial feasible solution

 Introducing slack variables

Maximize Z = 3x1 + 5x2 + 4x3 + 0s1 + 0s2 + 0s3
s.t

2x1 + 3x2 + s1 = 8
2x2 + 5x3 + s2= 10
3x1 + 2x2 + 4x3 + s3 = 15
x1, x2, x3, s1, s2, s3 ≥ 0

Step 4, 5, and 6
4. The table is not optimal, positive value at 𝐶𝑗 − 𝑍𝑗 𝑟𝑎𝑤
5. 𝑋2 leaves the base (pivot column )
Nuru M.((Ph.D.)
6. 𝑆1 inter into the base (pivot row )
 Continued…
36

 To find the improved solution the “elementary raw operation” will be

used.
 Transform pivot element into “1”
 Transform other pivot column element into”0”

N.B: the three elementary raw operations are;

1. Multiply or divide the entire raw by non-zero constant.
2. Add or subtract a multiple of one raw to or from another
3. Interchange raw.

Nuru M.((Ph.D.)
 Continued…
37 Step 7: Finding the new solution
 Improved solution

Step 8: repeat steps3-7, till final solution

Nuru M.((Ph.D.)
 Continued…
38
 The final iteration looks like

 all cj – zj < 0 for non-basic variables.

 Therefore, the optimal solution is reached with, x1 = 89/41, x2 = 50/41, x3 = 62/41
and the optimal value of Z = 765/41.
Nuru M.((Ph.D.)
 Exercise
39
Exercise 1; (Class Work) Exercise 2
 Solve the following LPM
6
Maximize Z=𝑋1 + 𝑋2
5
S.t
2𝑋1 + 𝑋2 < 180
𝑋1 + 3𝑋2 < 300
𝑋1 , 𝑋2 > 0

Nuru M.((Ph.D.)
 3.5.2.2 Simplex algorithm (minimization case)
. Two-Phase Method
40 . Big-M Method or Method of Penalties

 Note:
 Artificial (A) variables have no meaning
 Before the optimal solution all artificial variables must be removed.
Big-M method is method of removing artificial variables.
 Large undesirable coefficients to artificial variables are assigned on the
objective function.
+ M for minimization and – M for maximization problem.
Steps are all the same except the optimality criteria
The table is optimal if the 𝐶𝑗 − 𝑍𝑗 row values are > 0
If for a column, k, ck – zk is most negative and all entries in this column are negative,
then the problem has an unbounded solution.
 The interning variable is the one with large with –ve 𝐶𝑗 − 𝑍𝑗 value
Nuru M.((Ph.D.)
 Continued…
41

 Solve the following LPP?

Minimize Z =5𝑋1 + 3𝑋2
S,t
2𝑋1 + 4𝑋2 ≤ 12
2𝑋1 + 2𝑋2 = 10
5𝑋1 + 2𝑋2 > 10
𝑋1 ,𝑋2 > 0

Nuru M.((Ph.D.)
 Continued…
42

Sep2: Standard form Step 4, 5, and 6

Minimize Z = 5𝑋1 +3𝑋2 +0𝑆1 +0𝑆3 +𝑀𝐴2 +𝑀𝐴3 4. The table is not optimal, negative value at 𝐶𝑗 − 𝑍𝑗 𝑟𝑎𝑤

S.t 5. 𝑋2 leaves the base (pivot column )

2𝑋1 +4𝑋2 +𝑆1 = 12 6. 𝑆1 inter into the base (pivot row )
2𝑋1 +2𝑋2 +𝐴1 = 10
5𝑋1 +2𝑋2 -𝑆3 +𝐴3 = 10
𝑋1 , 3𝑋2 , 𝑆1 , 𝑆3 , 𝐴2 , 𝐴3 > 0

Step 3. initial feasible solution

Nuru M.((Ph.D.)
 Continued…
43 Step 7: Finding the new solution

Step 8: repeat steps3-7, till final solution

Nuru M.((Ph.D.)
 Continued…
44
Next iteration Interpretation

 Thus an optimal solution is arrived

at with the value of variables as:
𝑋1 = 4, 𝑋2 = 1, 𝑆1 = 0, 𝑆2 = 12 and
Min Z = 23

N.B: the same procedure is applied for mixed

maximization problem and simple
minimization problem.

Nuru M.((Ph.D.)
 Exercise
45
Exercise 1, (Class-work) Exercise 2
• Solve the following LP model.  Solve the following LP model.
Max Z = 6X1+8X2 Minimize ɵ = 16x1 + 9x2 + 21x3
S.t: s.t.
X2 ≤ 4
x1 + x2 + 3x3 ≥ 12
X1+X2 = 9
2x1 + x2 + x3 ≥ 16
6X1+2X2 ≥ 24
X1, X2 ≥ 0 x1, x2, x3 ≥ 0

Nuru M.((Ph.D.)
 3.5.2.3 Some complications and their resolution
(Assignment)
46
 The complications that may arise in applying the simplex method may
include;
1. Unrestricted Variables: when DV take either positive, negative or zero value
2. Tie for Entering Basic Variable (Key Column)
3. Tie for Leaving Basic Variable (Key Row) – Degeneracy
4. Types of linear programming solutions
 Three types of solutions may exist:
a) Alternative (Multiple) Optimal Solutions
b) Unbounded Solution
c) Infeasible Solution

Nuru M.((Ph.D.)
 3.5.2.4 Post optimality Analysis
(Assignment)
47

 Duality
 Sensitivity Analysis

Nuru M.((Ph.D.)
48

Any Question?
Nuru M.((Ph.D.)