Linear Programming:

Basic Formulation and Graphical

Solution

References:
• Dantzig, G. B., & Thapa, M. N., 1977, Linear Programming 1:
Introduction, Springer-Verlag, New York.
• Hillier, F. S., & Lieberman, G. J., 2001, Introduction To Operations
Research, McGraw-Hill, New York.
• Taha, H. A., 2007, Operations Research: An Introduction, 8th ed.,
Pearson Prentice Hall, Upper Saddle River, NJ 07458.
 The History (1)

• Operation Research (OR) is a scientific method to

allocate scarce resources in an effective way to optimize
the result.
• The term operation research is emerged in World War II
(Hillier & Lieberman, 2001). As a need to allocate scarce
resources for various military operations in an effective
manner, the British and the U.S. military management
asked scientists to apply a scientific approach to dealing
with this problem. The research is called research on
(military) operations and then just called operation
research.
 The History (2)

• Industries are gradually interested in OR after it

showed a big success in military. The OR field then
grows rapidly.
• Two factors became key roles in rapid growth of OR:
the development of the method and the computer
technology (Hillier & Lieberman, 2001).
 The History (3)
• The methods of OR are quickly emerge in its early
period. An example is the development of linear
programming and the Simplex method by George
Dantzig in 1947.
• The Simplex method is an effective tool to solve linear
programming.
• It can also be used to conduct a sensitivity analysis of
the result. The power of Simplex method is surprising
even for its inventor, George Dantzig (Dantzig & Thapa,
1997).
• Another key factor is the development of computer,
both in hardware and software. By the recent capability,
computer software is able to solve a large-scale problem
in a reasonable time.
 Linear Programming (1)
• Linear programming is one technique of operation
research. Linear programming is originally from the
words “Programming in a Linear Structure” (Dantzig
& Thapa, 1997).
• Programming refer to the term in military meaning
various plans or proposed schedules of training,
logistical supply, and deployment of combat units.
At that time, the term program has not been used
to define a set of instruction in a computer; instead,
it is called codes.
• The problem is modeled in a linear structure so that
it is called linear programming.
 Linear Programming (2)

• Typically, the problem of linear programming is to

find a set of value of variables that yields the
best/optimal objective result, subject to some
constraints. The objective and the constraints are
expressed in linear form.
 Example 1:
The WYNDOR GLASS CO. produces windows (product
1) and glass doors (product 2). It has three plants.
Aluminum frames and hardware are made in Plant 1,
wood frames are made in Plant 2, and Plant 3
produces the glass and assembles the products.
 Example 1 (Cont.):
Determine what the production rates should be for
the two products in order to maximize their total
profit, subject to the restrictions imposed by the
limited production capacities available in the three
plants.
 Definition:
Decision variables:
x1 :number of batches of product 1 produced per
week
x2 :number of batches of product 2 produced per
week
Objective:
Z :total profit per week (in thousands of dollars) from
producing these two products
Constraints:
• Plants capacity (time available)
• Non-negative constraint
Formulation:
Graphical Solution:
Graphical Solution:
 Example 2:
Reddy Mikks produces both interior and exterior
paints from two raw materials, Ml and M2. The
following table provides the basic data of the
problem:
 Example 2 (Cont.):
A market survey indicates that the daily demand for
interior paint cannot exceed that for exterior paint by
more than 1 ton. Also, the maximum daily demand for
interior paint is 2 tons.
Reddy Mikks wants to determine the optimum (best)
product mix of interior and exterior paints that
maximizes the total daily profit.
 Definition:
Decision variables:
x1 :tons produced daily of exterior paint
x2 :tons produced daily of interior paint
Objective:
Z :total daily profit (in thousands of dollars)
Constraints:
• The raw material restrictions
• Demand restriction
• Non-negative constraint
Formulation:
Graphical Solution:
Graphical Solution:
 Assumptions in linear programming:
Proportionality:
Proportionality assumption means that the objectives
value, z, is proportional with the value of decision
variable xi. An objective function z = 3x12 + 5x2 is an
example of non-proportional function.
Additivity:
Additivity assumes that the objective function and the
constraints between variables/activity are in the form
of addition and do not contain any variable cross
product. Objective function z = 3x1.5x2 is an example
of additivity violation.
 Assumptions in linear programming (2):
Divisibility:
This assumption means that the value of
variables/activities can be fractional (not have to be
integer). For example, x1=0.5 and x2=1.75 is permissible
by this assumption. If the model determined that the
variable should be integer, the problem is not an ordinary
linear programming anymore, but has become integer
programming.
Certainty:
It assumes that all parameter of the model are known
constants. However, sensitivity analysis can be conducted
to see the effect of changing parameters value.
Some Cases in Linear Programming:
Unique finite optimal solution :

Unique Unique
optimal optimal

Bounded region Unbounded region

Alternative finite optimal solutions:

Alternative
optima Alternative
X* optima
X1*

X2*

Bounded region Unbounded region

Unbounded optimal solution :

Alternative
X* optima

Unbounded region
 Linear Programming:
Basic Sensitivity Analysis

References:
• Hillier, F. S., & Lieberman, G. J., 2001, Introduction To Operations
Research, McGraw-Hill, New York.
• Taha, H. A., 2007, Operations Research: An Introduction, 8th ed.,
Pearson Prentice Hall, Upper Saddle River, NJ 07458.
The Standard Formulation
The Wyndor Glass Formulation:
The Graphical Solution:
 Sensitivity Analysis

• To determine which parameters of the model are

most critical (the sensitive parameter) in
determining the solution.
• Typically, greater attention is given to performing
sensitivity analysis on the bi and cj parameters than
aij parameters.
Changing cj Parameters
The solution will not change if:

0 ≤ c1/c2 ≤ 3/2
Changing bi Parameters:

Binding
Constraints
 Shadow price (dual price)

• The shadow price for resource i (denoted by yi*)

measures the marginal value of this resource, i.e.,
the rate at which Z could be increased by (slightly)
increasing the amount of this resource (bi) being
made available.
 Tugas

Find the solution of the following problem:

Max z = 40 x1 + 50 x2

Subject to
2500 x1 + 1000 x2 ≤ 5000
250 x1 + 500 x2 ≤ 1500
x1, x2 ≥ 0
 Linear Programming:
More Formulation

References:
• Hillier, F. S., & Lieberman, G. J., 2001, Introduction To Operations
Research, McGraw-Hill, New York.
• Masruroh, N., 2011, Handout Metode Optimasi, Teknik Industri,
UGM.
• Taha, H. A., 2007, Operations Research: An Introduction, 8th ed.,
Pearson Prentice Hall, Upper Saddle River, NJ 07458.
The Standard Formulation
 Case 1 (Diet Problem)
Ozark Farms uses at least 800 lb of special feed daily.
The special feed is a mixture of corn and soybean
meal with the following compositions:

The dietary requirements of the special feed are at

least 30% protein and at most 5% fiber. Ozark Farms
wishes to determine the daily minimum-cost feed
mix.
 Definition & Formulation of Case 1:
Decision Variable:

Objective Function:
minimize Z = the total daily cost (in dollars) of the feed
mix
 Definition & Formulation of Case 1 (Cont.):
Constraint:
1. Minimum total feed

2. Protein dietary requirement

3. Fiber requirement

4. Non negative constraint

 Formulation of Case 1 (Cont.):
The problem can be formulated as:
 Case 2 (Loan Policy)
Emon Bank is in the process of devising a loan policy
that involves a maximum of $12 million. The following
table provides the pertinent data about available
types of loans.
 Case 2 (Cont.)
Bad debts are unrecoverable and produce no interest
revenue.
Competition with other financial institutions requires
that the bank allocate at least 40% of the funds to
farm and commercial loans. To assist the housing
industry in the region, home loans must equal at least
50% of the personal, car, and home loans. Emon bank
also has a stated policy of not allowing the overall
ratio of bad debts on all loans to exceed 4%.
 Definition & Formulation of Case 2:
Decision Variable:

Objective Function:
minimize Z = net return
 Definition & Formulation of Case 2 (Cont.):
Constraint:
 Definition & Formulation of Case 2 (Cont.):
Constraint:
 Case 3 (Regional Planning)
A group of 3 farming communities (kibbutz) is
planning agricultural production for the coming year.
The agricultural output is limited by both the amount
of available irrigable land and the quantity of water
allocated for irrigation by the Water. The data are:
 Case 3 (Cont.)
The crops suited for this region include sugar beets,
cotton, and sorghum. These crops differ primarily in
their expected net return per acre and their
consumption of water. In addition, the Ministry of
Agriculture has set a maximum quota for the total
acreage that can be devoted to each of these crops as
the following:
 Case 3 (Cont.)
Because of the limited water available for irrigation, the
Group will not be able to use all its irrigable land for
planting crops in the upcoming season. To ensure equity
between the three kibbutzim, it has been agreed that
every kibbutz will plant the same proportion of its
available irrigable land. For example, if kibbutz 1 plants
200 of its available 400 acres, then kibbutz 2 must plant
300 of its 600 acres, while kibbutz 3 plants 150 acres of its
300 acres. However, any combination of the crops may be
grown at any of the kibbutzim.
The problem is to plan how many acres to devote to each
crop at the respective kibbutzim while satisfying the given
restrictions. The objective is to maximize the total net
return to the group as a whole.
 Definition & Formulation of Case 3:
Decision Variable:

Objective Function:
Max Z= total net return
 Definition & Formulation of Case 3 (Cont.):
Constraints:
 Definition & Formulation of Case 3 (Cont.):
Constraints Cont.:
 Case 4 (Oil Blending Problem)

A refinery produces 3 grades of oil A, B, and C from 4

grades of crude oil 1, 2, 3, and 4.
Data
Crude Quantity Cost/barrel
available

1 3000 3
2 2000 6
3 4000 4
4 1000 5
 Case 4 (Cont.)
• Blending requirements:
– Oil A
• Not more than 30% of 1
• Not less than 40% of 2
• Not more than 50% of 3
– Oil B
• Not more than 50% of 1
• Not less than 10% of 2
– Oil C
• Not more than 70% of 1
 Case 4 (Cont.)
• Selling price
– A - $5.5
– B - $4.5
– C - $3.5

Problem: what blends of crude should be used to

make each grade of oil in order to maximize profit
subject to quantities available and blending
requirements?
Simplex Method
 Thanks to
GEORGE B. DANTZIG
(1914-2005)

The inventor of the

Simplex Method
 Introduction
• If a linear programming problem has more than two
variables, it will be difficult to be solved using
graphical method.
• The Simplex method is a general procedure to solve
linear programming problem.
• Since developed by George Dantzig in 1947, it
shows as a remarkable useful technique.
• The Simplex method is very efficient to solve linear
programming. Moreover, it provides tool for post-
optimality analysis.
 Introduction (Cont.)
• The Simplex method can only work in a problem if
the feasible solution is a convex set. The figure
shows example of (a) convex set and (b) non-convex
set
 Introduction (Cont.)
• The best solution in linear programming will always
lie in a corner point.
• Therefore, the Simplex method looks the optimal
solution by moving from a corner point to another
adjacent corner point that result a better objective
function.
• When no better adjacent corner point can be found,
the searching is stopped and the last corner point is
the optimal solution.
Algorithm of the Simplex Method

1. Initialization Step
2. Iterative Step
3. Optimality Test
 1. Initialization Step

a) Transform the formulation into an augmented

form (some books define as a standard form)
b) Choose an initial solution
 a). Transform the formulation
into an augmented form

Except for the non-negative constraints, the inequality

constraints (constraints with ≥ or ≤ relationship)
should be convert to equality constraints (=) by
adding some slack or surplus variable.
Example: x1 ≤4
becoming: x1 + s =4
x1 ≥ 1 can be converted to be x1 - s = 1
The Wyndor Glass Formulation:
 The Augmented form of
The Wyndor Glass Formulation:
 b). Choose an initial solution

•Actually, all corner point can be chosen as initial

solution, but for convenient, the initial solution is
usually set to which all of the original variable has
the value of null.
•In the Simplex method, the variables that set to be
null are called non-basic variables and the other
variables are called basic variables.
•Therefore, in initial solution, all slack/surplus
variables will have non-zero value. Thus, the initial
solution will be: x1=0, x2=0, x3=4, x4=12, and x5=18.
 Tabular form (for convenient)
The Simplex method will be more convenient if presented in a
tabular form. In the tabular form, the objective function form
is modified becoming: z - 3x1 - 5x2 = 0
 2. Iterative Step
Part 1: Enter (Tarik……….)
• Determine the entering variable by selecting a
variable that has the most negative coefficient from
the non-basic variables.
• In the example, x2 is chosen as entering variable
because it has the most negative coefficient -5,
comparing with x1 having coefficient -3.
• The column of entering variable is called pivot
column.
Part 2: Leave (Lepas……….)

• Determine the leaving variable. To do this, a ratio

between right side and each coefficient of pivot
column should be calculated.
• Then, choose the smallest positive value (not
including zero) as the leaving variable. The row of
leaving variable is called pivot row.
• The coefficient in the intersection between pivot
column and pivot row is called pivot number.
• In the initial table, the ratio of the equations (eq. (1)
to (3)) are 4/0, 12/2, and 18/2 respectively. Thus,
equation (2), with ratio value 12/2, is selected to be
the pivot row, and x4 becomes leaving variable.
 Pivot Pivot
column row

Pivot
number
Part 3: Change (Berubah……….)

• Calculate the new basic feasible solution and

reconstruct the Simplex table. The calculation
method is Gaussian elimination. New basic variable
coefficient should be 1 in the pivot row.
• Therefore:
New pivot row = old pivot row /pivot number
For other row:
New row = old row – (pivot column coefficient x
new pivot row)
Iteration 1
 3. Optimality Test
• The solution of the last iteration is then checked
whether it has reached optimal or not.
• If every coefficients of equation (0) in the table is
non-negative (≥0), the solution is optimal and the
iteration can be stopped.
• Otherwise, go to iterative step to find the best
adjacent corner point.
Final simplex table
Assignment 1
Assignment 2
 Special cases of LP solution
1. If there is a tie
a) For entering non-basic variables: Choose any one of
them.
b) For leaving basic variables: Choose any one of
them.

2. What if when you solve for basic variables, one of

them has a value of zero?
It does not matter, it is only theoretical important.
The name given to this kind of solution is
degenerate.
 Special cases of LP solution (Cont.)
3. How can you tell if the LP problem has no solution
(that is, the model is inconsistent)?
This can be detected if some artificial variables have
positive values when the LP solution is optimal.

4. How can you tell if the LP solution is unbounded?

This can be detected if an entering variable can be
increased without limit. That is, the pivot column
has no positive coefficients.
 Special cases of LP solution (Cont.)
5. How can you tell if there is more than one optimal
solution?

This can be detected if there exists a non-basic

variable which has a coefficient of zero in the final
objective function row. Note that if P1and P2 are
two optimal corner points then λP1+(1- λ )P2 are
equally good solution, for any 0 ≤ λ ≤ 1.
 Special cases of LP solution (Cont.)
6. Redundant constraint
If one or more equality constraints are linear
combination of the other equality constraints, then they
are redundant and can be eliminated. For example,
consider the following LP problem:

max z = 2x1+ 3x2

subject to
x1 + x2 = 4 (1)
-x1 + 3x2 = 7 (2)
4x2 = 11 (3)
x1 , x2 ≥ 0
Constraint (3) = (1)+(2), and hence any one of them can
be eliminated.