Lecture 1 Part 2:Introduction to Linear Programming

• Linear Programming Problem

• Problem Formulation
• A Simple Maximization Problem
• Graphical Solution Procedure
• Extreme Points and the Optimal Solution
• A Simple Minimization Problem
• Special Cases
• Computer solutions (next lesson)

Slide 1
 Linear Programming
Linear programming has nothing to do with computer programming.
The use of the word “programming” here means “choosing a course
of action.”
Linear programming involves choosing a course of action when the
mathematical model of the problem contains only linear functions.
Linear Programming Problem
• The maximization (of profit) or minimization (of cost) of some
quantity is the objective in all linear programming problems.
• All LP problems have constraints that limit the degree to which
the objective can be pursued.
• A feasible solution satisfies all the problem's constraints.
• An optimal solution is a feasible solution that results in the
largest possible objective function value when maximizing (or
smallest when minimizing).
• A graphical solution method can be used to solve a linear
program with two variables.
2
 Guidelines for Model Formulation
Step 1: Understand the problem thoroughly.
Step 2: Describe the objective.
Step 3: Describe each constraint.
Step 4: Define the decision variables.
Step 5: Write the objective in terms of the decision variables.
Step 6: Write the constraints in terms of the decision variables.

Example: A maximization problem (Section 2.1 Pg 32)

Step 1: Understand the problem:
Par Inc is a manufacturer for golf equipment. It wants to make two
new models of golf bag – Standard and Deluxe bag. The similar
production stages and total capacities are:
1. Cutting and dyeing of the material : 630 hrs
2. Sewing : 600 hrs
3. Finishing : 708 hrs
4. Inspection & packaging : 135 hrs
Slide 3
 A Maximization Problem
Step 1: Understand the problem:
It is found that production time (in hours) per golf bag is:
Production Department Standard Bag Deluxe Bag
1. Cutting & Dyeing 7/10 1
2. Sewing 1/2 5/6
3. Finishing 1 2/3
4. Inspection & Packaging 1/10 1/4

The accounting department analyzed that profit for each standard

bag will be $10, and for each deluxe bag will be $9.

Step 2: Describe the objective

The Objective of Par Inc is to _______________________

Slide 4
 A Maximization Problem
Step 3: Describe each constraint:
Constraint 1: Cutting and dyeing ≤ 630 hrs
Constraint 2: Sewing ≤ 600 hrs
Constraint 3: Finishing ≤ 708 hrs
Constraint 4: Inspection & packaging ≤ 135 hrs
(note: Step 2 & 3 are for lecture explanation. You only need to
write down step 4 onwards)

Step 4: Define the decision variables:

The controllable inputs for Par Inc are (1) the number of standard
bags, and (2) the number of deluxe bags produced.

Let ___ = number of standard bags

___ = number of deluxe bags to be produced

__ & ___ are referred to as decision variables

Slide 5
 A Maximization Problem
Step 5: Write the Objective Function in terms of the Decision Variables
Total Profit Contribution =
Objective Function is to:
Step 6: Write the constraints in terms of the Decision Variables:
Constraint 1: Cutting and dyeing: ≤ 630
Constraint 2: Sewing: ≤ 600
Constraint 3: Finishing: ≤ 708
Constraint 4: Inspection & packaging: ≤ 135
Nonnegativity constraint:
The complete Linear Programming Model for Par is:
Max
s.t.

Slide 6
 Graphical Solution Procedure
We need to find the feasible region for the solution.
First we draw the constraint lines and define the region of feasibility.

For cutting & dyeing constraint: (7/10) S + 1 D ≤ 630

We need to draw the line (7/10) S + 1 D = 630
Setting S = 0 => D= _______
Setting D = 0 => S= _______
We have one point on each axis of the graph D vs S
What if the points on the axis of out of the graph paper?

Where is the feasible region for (7/10)S + 1D ≤ 630? Below/above the line?
You can use test points e.g. (200,200) and (600,500) to determine the
region, or use mathematic reasoning that for ≥ constraint, the feasible area
is on above or to the right of the line (for vertical line), and vice versa.

Slide 7
 Graphical Solution Procedure
For sewing constraint: (1/2) S + (5/6) D ≤ 600
Setting S = 0 => D= _______
Setting D = 0 => S= _______
We have one point on each axis of the graph D vs S, and shade the
feasible region for constraint 2.
We continue for Constraint 3 & 4
The final feasible region is the area common to all the feasible regions. The
data in this area satisfies all the constraints.
But which pair of data meet the Objective Function (O.F.)?
Recall the O.F is Max Profit, P= 10S + 9D.
To start, we arbitrarily set 10S + 9D = 1800 and draw this line.
A set of parallel line can be draw which provide different P values.
Making D the subject, P=10S + 9D => D= (-10/9)S + (1/9)P
Hence, for any of these line, slope = -10/9, D intercept = (1/9)P
Graphically, which line give us the maximum profit? ___________
Slide 8
 Example: Maximisation
Number of Deluxe Bags

Number of Standard Bags

Slide 9
 Graphical Solution Procedure
From the graph, the optimal solution is:
S=
D=

The graphical solutions may need estimation and incur some errors.
A closer inspection shows that the optimal solution is at the intersection of the
cutting & dyeing and the finishing constraints line. A more exact solution can be
obtained from solving the two simultaneous equations:
(7/10)S + 1D = 630 -----(1)
1S + (2/3) D = 708 -------(2)
You can solve simultaneous equation by substitution or elimination,
e.g. from (2), S= 708 – (2/3) D ----(3)
Sub (3) into (1), we can solve for D.
The solution is D= 252, S = 540.
Sub into the Objective Function, max profit = 10(540) + 9(252) = $7668
Ans: To maximise profit, Par need to produce 540 standard 252 deluxe bags.
The maximum profit is $7668
Slide 10
 Slack Variables
For the optimal solution, S=540, D=252, we substitute these values into
the constraint equations to obtain the hours required:
Constraints Hrs Required Hrs Unused Hrs
available
1.Cutting & Dyeing 630 630 0
2. Sewing 480 600 120
3. Finishing 708 708 0
4. Inspection & 117 135 18
Packaging

• Notice that the optimal solution lies at the junction of C&D and
Finishing. These are called binding constraints. The capacities have
been fully used up.
• The non binding constraints are the Sewing and Inspection &
Packaging. They have unused hours.
• In Linear Programming, any unused capacity for a ≤ constraint is
referred to as a Slack associated with the constraint.
Slide 11
 Extreme Points and the Optimal Solution
• The corners or vertices of the feasible region are referred to
as the extreme points.
• An optimal solution to an LP problem can be found at an
extreme point of the feasible region.
• When looking for the optimal solution, you do not have to
evaluate all feasible solution points.
• You have to consider only the extreme points of the feasible
region. (Ref Fig 2.13 Pg 52). Substitute each point into the
O.F to find the optimal solutions (pair).
Note: Above method is feasible but not recommended for our
practice as we will conduct sensitivity analysis.

Slide 12
 A Minimization Problem (Pg 55)
Problem:
M&D Chemicals produces two products, A and B. The
management specifies that the capacity for product A and B
must total at least 350 gallons. Separately, a major customer’s
order for 125 gallons of product A must also be satisfied.
Product A requires 2 hours of processing times per gallon and
product B requires 1 hour of processing time per gallon. For the
coming month, 600 hours of processing time are available.
M&D’s objective is to satisfy these requirements with a minimum
production cost. Production cost are $2 per gallon for product A
and $3 per gallon for product B.

Solution: Define variables:

Let

Slide 13
 A Minimization Problem (Pg 55)
Model:
Objective Function:
s.t

Slide 14
 Refer to textbook for graph
Gallon of Product B

Gallon of Product A

Slide 15
 A Minimization Problem (Pg 55)
From the graph, the optimal solution lies at the junction of
constraint ___ & ____.
Solve the simultaneous equations:

Ans:
Optimal solution:

Minimum cost =

Slide 16
 Surplus Variables
For the optimal solution, A=250, B=100, we substitute these values into
the constraint equations to obtain the quantity:
Constraints LHS of RHS of Slack/Surplus
constraint Constraint
1. Demand for Prod A 1A=250 ≥ 125 125
2. Total Production 1A+1B=350 ≥ 350 0
3. Processing time 2A+1B=600 ≤ 600 0

• The non binding constraints is the demand of Product A. Product A

exceeds its minimum required level by 250-125 = 125 gallons.
• In Linear Programming, any exceed quantity for a ≥ constraint is
referred to as a Surplus associated with the constraint.
• The optimal solution lies at the junction of Total Production and
Processing Time constraint lines. The quantity has just been met (for
≤)or the capacity has been fully used up (for ≥). There is no slack nor
surplus.
Slide 17
 Summary of the Graphical Solution Procedure
• Prepare a graph of the feasible solutions for each of the
constraints.
• Determine the feasible region that satisfies all the constraints
simultaneously.
• Draw an objective function line.
• For maximisation problem:
– Move parallel objective function lines toward larger objective
function values without entirely leaving the feasible region.
– Any feasible solution on the objective function line with the
largest value is an optimal solution.
• For minimisation problem:
▪ Move parallel objective function lines toward smaller objective
function values without entirely leaving the feasible region.
▪ Any feasible solution on the objective function line with the
smallest value is an optimal solution.

Slide 18
 Special Case 1
• Alternative Optimal Solutions
In the graphical method, if the objective function line is parallel to
a boundary constraint in the direction of optimization, there are
alternate optimal solutions, with all points on this line segment
being optimal. (Ref Fig 2.18 Pg 61). Good or bad?

Slide 19
Special Case 2
Infeasibility
• No solution to the LP problem satisfies all the
constraints, including the non-negativity conditions.
• Graphically, this means a feasible region does not
exist.
• Causes include:
• A formulation error has been made (less likely),
or.
• Management’s expectations are too high.
• Too many restrictions have been placed on the
problem (i.e. the problem is over-constrained).

20
 Example: Infeasible Problem
Max 2x1 + 6x2

s.t. 4x1 + 3x2 < 12

2x1 + x2 > 8
x2
x1, x2 > 0 10

2x1 + x2 > 8
8
• There are no points that
satisfy both constraints, 6
so there is no feasible 4x1 + 3x2 < 12
region (and no feasible 4
solution).
2
• How to resolve the
problem? x1
2 4 6 8 10

Slide 21
Special Cases 3

Unbounded
• The solution to a maximization LP problem is
unbounded if the value of the solution may be
made indefinitely large without violating any of the
constraints. Isn’t that good?
• For real problems, this is the result of improper
formulation. (Quite likely, a constraint has been
inadvertently omitted.)

22
 Example: Unbounded Solution
Max 4x1 + 5x2

s.t. x1 + x2 > 5
3x1 + x2 > 8

x1, x2 > 0 x2
10
• The feasible region is 3x1 + x2 > 8
unbounded and the 8
objective function line
can be moved outward 6
from the origin without
bound, infinitely 4
x1 + x2 > 5
increasing the
objective function. 2

x1
2 4 6 8 10
Slide 23
 Homework 01
• Homeworks will be published on ELearn under the
respective week folders.
• Students have to submit handwritten solutions on A4
paper, submit before class start in the following week.
• Late submission will not be accepted.

Slide 24