Solving Linear

Programming
Models
Chapter 3
(p89 – 127)
Introduction
• Linear programming problems can be solved by
using several PC software packages
3.1. Computer Solution
• Originally (1940s) the simplex model was used
• Simplex model: a procedure involving a set of
mathematical steps to solve linear programming
problems
• The mathematical steps of the simplex method
were programmed in prewritten software
packages designed for the solution of linear
programming problems
• Many software packages with linear programming
capabilities are available
3.2. Sensitivity Analysis
• In practice, the reality model parameters are estimates
• Subject to change
• Important to study the effects of changes in parameters on
the solution to the model
• Changes may be:
• In reaction to anticipated uncertainties in parameters
• Reactions to information
• Sensitivity analysis: the analysis of the effect of parameter
changes on the optimal solution
• To study the effect of change, solve the original problem,
make changes, solve for new problems statement, and
compare
• It is also possible to study the effect of change without solving
the problem again
3.2.1. Changes in Objective Function
Coefficients
••  Let’s consider the Beaver Creek example again:

• subject to
•
•
• Optimal solution is at point B

Optimal solution point

... Changes in Objective Function
Coefficients
••  How will a change in profit per bowl from R40 to
R100 influence the solution?
• Function line is steeper
• Optimal solution point changes from B to C

Changing the objective function coefficient

... Changes in Objective Function
Coefficients
••  Alternatively, if the profit per mug is increased
from R50 to R100, the objective function line
would be flatter:
• Point A will become the optimal solution point

Objective function coefficient

... Changes in Objective Function
Coefficients
••  The objective of sensitivity analysis is to determine
the range of values (called the sensitivity range
for ) for a specific objective function coefficient
over which the optimal solution point ( and ) will
remain optimal
• E.g. if the coefficient of is increased from R40 to more
than R40, point C will become optimal
• For less than R40, point A will become optimal
• We will designate the sensitivity range as
• By considering the sensitivity range, it can be
determined how much of the selling price and
costs can be altered without resulting in a change
in production
... Changes in Objective Function
Coefficients
••  Determine the sensitivity range for in the figure
• Slope of the objective function:
•
•  the slope of the function is
• The objective function is now in the form of a
straight line [, where:
• The intercept:
• The slope:
... Changes in Objective Function
Coefficients
••  If the slope increases to , the objective function line
becomes parallel to the constraint line
• Point C becomes the optimal point along with B
• What objective function coefficient for will make
the objective function slope equal to
•
• R66,67
•  objective function slope = or
• [ - objective function coefficient for ]
Determining the sensitivity range for
... Changes in Objective Function
Coefficients
••  we have determined that the upper limit of the
sensitivity range , the coefficient, is 66,67
• If profit for a bowl increases to exactly R66,67, the
solution points will be both B and C
• If the profit for a bowl is more than R66,67, Point
C will be the optimal solution point
... Changes in Objective Function
Coefficients
••  To determine the lower limit for the sensitivity range:
• The objective function line slope decreases (becomes
flatter) from to the same slope as the constraint line [],
point A becomes optimal along with B
• The slope of this constraint line is , that is

• To get an objective function slope of , the

profit for a bowl would have to decrease to R25:

• 
... Changes in Objective Function
Coefficients
••  The complete sensitivity range for the can be
expressed as
•  The profit for a bowl can vary anywhere
between R25 and R66,67, and the optimal
solution point [, ] will not change
• Of course, the total profit will change, depending
on whatever value is

• Obviously, the sensitivity range for can be

determined in the same way
... Changes in Objective Function
Coefficients
• Simultaneous changes can be made in the
objective function coefficients as long as the
changes taken together do not change the
optimal solution point
• However, determining the effect of these
simultaneous changes is overly complex and time-
consuming using graphical analysis
• Determining the effect of simultaneous changes in
model parameters and performing sensitivity
analysis in general are much easier and more
practical using the computer
... Changes in Objective Function
Coefficients
•  
... Changes in Objective Function
Coefficients
••  As the objective function coefficient for
decreases
from R6, the objective function slope of -2
decreases [flatter line]
• When the coefficient, , equals R4, then the slope of
the objective function is , which is the same as the
constraint lime []
• This makes point B optimal along with A
• Thus, the limit for the sensitivity range for is R4
3.2.3. Changes in Constraint
Quantity Values
••  The 2nd type of sensitivity analysis
• The values to the right of the inequality signs in
the constraints
• For Beaver Creek:
• Maximize
Subject to (labour)
(clay)

• The constraint quantity values are 40 and 120

… Changes in Constraint
Quantity Values
••  Consider a scenario in which the

labour can be increased to 60 hours

• The feasible solution space changes from OABC to
OA’B’C’
• B’ is the new optimal solution
• Note: the solution mix did not change, only and
• The sensitivity range for a right-hand-side value is
the range of values over which the quantity values
can change without changing the solution variable
mix, including slack variables
Increasing the labour constraint quantity
… Changes in Constraint
Quantity Values
••  If the quantity value for the labour constraint is
increased from 40 to 80 hours, the new solution
space is OA’C and a new solution variable mix
occurs at A’
• Whereas at the original optimal point, B, both
and are in the solution, at the new optimal point
A’, only is produced
• i.e. , , ,
… Changes in Constraint
Quantity Values
••  Thus, the upper limit of the sensitivity range for
the quantity value for the first constraint, which
will be referred to as , is 80 hours
• At this value the solution mix changes such that
bowls are no longer produced
• Furthermore, as increases past 80 hours,
increases
• i.e. slack hours are created
… Changes in Constraint
Quantity Values
••  Similarly, if the value for
is decreased to 30 hr,
the new feasible solution space is OA’C
• The new optimal point is C, where no mugs () are
produced
• The new solution is , , , ,
• Again, the variable mix is changed
• Note: sensitivity range for the
constraint quantity value for labour
is
… Changes in Constraint
Quantity Values
••  The sensitivity range for clay can be determined
graphically in a similar manner
• If the quantity value for the clay constraint [] is
increased from 120 to 160, then a new solution
space, OAC’, results, with a new optimal point, C’
… Changes in Constraint
Quantity Values
••  Alternatively, if the quantity value is decreased
from 120 to 60, the new solution space is OAC’
and the new optimal point is A
• ,,,,
• Note: the sensitivity ranges for and are
… Changes in Constraint
Quantity Values
••  Just like with the sensitivity ranges for the
objective function coefficient, the sensitivity
ranges are valid for only one value
• All other values are assumed to be held constant
• However, simultaneous changes can occur, as long
as they do not change the variable mix
… Changes in Constraint
Quantity Values
• The ranges for constraint quantity values provide
useful information for the manager, especially
regarding production scheduling and planning
• If resources are reduced at the pottery company,
•  At some point one of the products will no longer be
produced
•  The support facilities and apparatus for that
production will not be needed, or extra hours of
resources will be created that are not needed
• A similar (better) result will occur if resources are
increased
•  Profit will be more than with a reduction in
resources
3.2.4. Other Forms of Sensitivity
Analysis
••  Other forms of sensitivity analysis include
• Changing individual constraint parameters
• Adding new constraints
• Adding new variables
• E.g. for Beaver Creek, if a new, less-experienced
artisan was hired to make pottery,
it might take him 1,33 hr to
produce a bowl instead of 1hr
• Thus, the labour constraint would
change from to
... Other Forms of Sensitivity
Analysis
••  Note that a change in the coefficient for
in the
labour constraint rotates the constraint line,
changing the solution space from OABC to OAC
• It also results in a new optimal solution point C, where
and and
•  1,33 hrs would be the logical upper limit for this
constraint coefficient
• However, this type of sensitivity analysis for constraint
variable coefficients is not typically provided in the
standard linear programming computer output
• Thus, the most logical way to determine the effects of
this type of change is to run the program with different
values
... Other Forms of Sensitivity
Analysis
••  Other types of sensitivity analyses:
• Add a new constraint to the model
• Add a new variable
• E.g. suppose Beaver Creek added a 3rd constraint
for packaging its products:

• This would require the model to be solved again

with the new constraint, because the new
constraint changes the optimal solution
Old New
Bowls 24 20
Mugs 8 10
Profit R1 360 R1 300
... Other Forms of Sensitivity
Analysis
••  If a new variable is added to the model, it would
require that the model be solved again to
determine the effect of the change
• E.g. suppose the pottery company was
contemplating producing a 3rd product (cups)
requiring 1,2 hr of labour and 2kg of clay
• It can secure no additional resources, and the
profit for a cup is estimated to be R30
• This change is reflected in the following model
reformulation:
• Maximize
• Subject to
... Other Forms of Sensitivity
Analysis
••  New model reformulation:
• Maximize
subject to

• Solving this new formulation with the computer

will show that this prospective change will have no
effect on the original solution, that is, the model is
not sensitive to this change
• The estimated profit from cups was not enough
to offset the production of bowls and mugs, and
the solution remained the same
Homework
• Work through the example problems (p110 –
111)
• Problems: 6, 10, 16, 25
Class representative
Problem 6
• Given
  the following graph of a linear programming model with
a single constraint and the objective function maximize ,
determine the optimal solution point:

• Determine the values by which and must decrease or

increase in order to change the current solution point to the
other extreme point.
Problem 10
• A company produces two products, A and B, which have
profits of R9 and R7, respectively. Each unit of product must
be processed on two assembly lines, where the required
production times are as follows:

a) Formulate a linear model to determine the optimal product

mix that will maximize profit.
b) Transform this model into standard form.
Problem 16
• United Aluminium Company of Cincinnati produces three
grades (high, medium, and low) of
aluminium at two mills. Each mill has a
different production capacity (in tons per
day) for each grade, as follows:

• The company has contracted with a manufacturing firm to

supply at lease 12 tons of high-grade aluminium, 8 tons of
medium-grade aluminium, and 5 tons of low-grade aluminium.
It costs United R6000 per day to operate mill 1 and R7000 per
day to operate mill 2. The company wants to know the
number of days to operate each mill in order to meet the
contract at the minimum cost.
•
Problem 25