Class 3 LP Sensitivity

Linear Programming Formulations
Sensitivity Analysis
Class 3: Linear Programming Formulations

-Continued
Reference: HH: Chapter 5; CTR Chapter 3,4
October 19, 2016
Class 3: Linear Programming Formulations -Continued

Review from Last Session
Simplex Method
Resource Allocation problems
Cost-Benefit-Trade-off problems
Mixed problems

Outline
1 Linear Programming Formulations

Investment Planning
Blending Problem
Production Planning
Multiperiod Cash-Flow Problem
2 Sensitivity Analysis

Investment Planning
Linear Programming Formulations Blending Problem
Sensitivity Analysis Production Planning
Outline

Investment Planning
Blending Problem
Production Planning

Investment Planning
Retirement Planning Services
A client wishes to invest $750,000 in the following bonds.

Company Return Maturity Rating
Acme Chemical 8.65% 11 1-Excellent
DyneStart 9.50% 10 3-Good
Eagle Vision 10.00% 6 4-Fair
Micro Modeling 8.75% 10 1-Excellent
OptiPro 9.25% 7 3-Good
Sabre Systems 9.00% 13 2-Very Good

Investment Planning
Restrictions and Problem
Investment Restrictions
No more than 25% can be invested in any single company.
At least 50% should be invested in long-term bonds
(maturing in 10+ years).
No more than 35% can be invested in DynaStar, Eagle
Vision, and OptiPro.

Investment Planning
Restrictions and Problem
Investment Restrictions
No more than 25% can be invested in any single company.
At least 50% should be invested in long-term bonds
(maturing in 10+ years).
No more than 35% can be invested in DynaStar, Eagle
Vision, and OptiPro.
Goal:How to allocate investments to maximize income
while meeting investment restrictions

Investment Planning
Algebraic Model
Decision Variables

Investment Planning
Algebraic Model
Decision Variables
Xi = amount of money to invest in i; i = 1, . . . , 6

Investment Planning
Algebraic Model
Decision Variables
Objective Function

Investment Planning
Algebraic Model
Decision Variables
Objective Function
Maximize Total Annual Return:
0.0865X1 + 0.095X2 + 0.1X3 + 0.0875X4 + 0.0925X5 + 0.09X6

Investment Planning
Algebraic Model
Decision Variables
Objective Function
0.0865X1 + 0.095X2 + 0.1X3 + 0.0875X4 + 0.0925X5 + 0.09X6
Constraints

Investment Planning
Algebraic Model
Decision Variables
Objective Function
0.0865X1 + 0.095X2 + 0.1X3 + 0.0875X4 + 0.0925X5 + 0.09X6
Constraints
X6
Xi = 750000 Available amount
i=1
Xi ≤ 187500 ∀ i no more than 25%
X1 + X2 + X4 + X6 ≤ 375000 at least half in long term
X2 + X3 + X5 ≤ 262500 no more than 35%
Xi ≥ 0; ∀ i Nonnegativity constraints

Investment Planning
Solving the Model
Invest
ACME $112500
Dynastar $75000
Eagle Vision $187500
MicroModeling $187500
Optipro $0
Sabre Systems $187500
Total Invested $750000

Investment Planning
Solving the Model
Invest
ACME $112500
Dynastar $75000
Eagle Vision $187500
MicroModeling $187500
Optipro $0
Sabre Systems $187500
Total Invested $750000
Notice more money placed on ACME than Dynastar and
Optipro, even thought they offer higher returns. This is
because Dynastar and Optipro are both higher-risk
investments and 35% limit on higher risk is binding.

Investment Planning
Outline

Investment Planning
Blending Problem
Production Planning

Investment Planning
Blending Problem: Agri-Pro-Optimal Mix
Agri-Pro has received an order for 8,000 pounds of chicken

feed to be mixed from the following feeds.
Percent of Nutrient in
Nutrient Feed 1 Feed 2 Feed 3 Feed 4
Corn 30% 5% 20% 10%
Grain 10% 30% 15% 10%
Minerals 20% 20% 20% 30%
Cost/Pound $0.25 $0.30 $0.32 $0.15
The order must contain at least 20% corn, 15% grain, and
15% minerals.
Objective: Fill customer’s order at the lowest possible cost.

Investment Planning
Algebraic Formulation
Decision Variables

Investment Planning
Decision Variables
Xi = pounds of Feed i to use in mix; i = 1, 2, 3, 4

Investment Planning
Decision Variables
Objective Function

Investment Planning
Decision Variables
Objective Function
Minimize Total Cost:
0.25X1 + 0.30X2 + 0.32X3 + 0.15X4

Investment Planning
Decision Variables
Objective Function
0.25X1 + 0.30X2 + 0.32X3 + 0.15X4
Constraints

Investment Planning
Decision Variables
Objective Function
0.25X1 + 0.30X2 + 0.32X3 + 0.15X4
Constraints
X1 + X2 + X3 + X4 = 8000 Order

Investment Planning
Decision Variables
Objective Function
0.25X1 + 0.30X2 + 0.32X3 + 0.15X4
Constraints
X1 + X2 + X3 + X4 = 8000 Order
0.3X1 +0.05X2 +0.20X3 +0.10X4

8000
≥ 0.20 at least 20% Corn
0.1X1 +0.3X2 +0.15X3 +0.1X4

8000
≥ 0.15 at least 15% Grain
0.2X1 +0.2X2 +0.2X3 +0.3X4

8000
≥ 0.15 at least 15% Minerals

Investment Planning
Decision Variables
Objective Function
0.25X1 + 0.30X2 + 0.32X3 + 0.15X4
Constraints
X1 + X2 + X3 + X4 = 8000 Order
0.3X1 +0.05X2 +0.20X3 +0.10X4

8000
≥ 0.20 at least 20% Corn
0.1X1 +0.3X2 +0.15X3 +0.1X4

8000
≥ 0.15 at least 15% Grain
0.2X1 +0.2X2 +0.2X3 +0.3X4

8000
≥ 0.15 at least 15% Minerals
X1 , X2 , X3 , X4 ≥ 0 Nonnegativity constraints

Investment Planning
A Comment about Scaling
Notice the coefficient for X2 in the “corn” constraint is

0.05/8000 = 0.00000625
As Solver runs, intermediate calculations are made that
make coefficients larger or smaller.
Storage problems may force the computer to use
approximations of the actual numbers.
Such scaling problems sometimes prevents Solver from
being able to solve the problem accurately.
Most problems can be formulated in a way to minimize
scaling errors.

Investment Planning
Algebraic Formulation of Rescaled Model

Decision Variables
Xi = pounds of Feed i in thousands of pound to use in mix ∀i
Objective Function
Minimize Total Cost: 250X1 + 300X2 + 320X3 + 150X4
Constraints
X1 + X2 + X3 + X4 = 8 Order
0.3X1 +0.05X2 +0.20X3 +0.10X4

8 ≥ 0.20 at least 20% Corn
0.1X1 +0.3X2 +0.15X3 +0.1X4

8 ≥ 0.15 at least 15% Grain
0.2X1 +0.2X2 +0.2X3 +0.3X4

8 ≥ 0.15 at least 15% Minerals
X1 , X2 , X3 , X4 ≥ 0 Nonnegativity constraints
Investment Planning
Scaling: Before and After
Before:
Largest constraint coefficient was 8,000
Smallest constraint coefficient was
0.05/8 = 0.00000625
After:
Largest constraint coefficient is 8
Smallest constraint coefficient is
0.05/8 = 0.00625
The problem is now more evenly scaled

Investment Planning
Outline

Investment Planning
Blending Problem
Production Planning

Investment Planning
Upton Corporation- Production Planning

Upton is planning the production of their heavy-duty air
compressors for the next 6 months.
1 2 3 4 5 6
Unit Production Cost $240 $250 $265 $285 $280 $260
Units Demanded 1,000 4,500 6,000 5,500 3,500 4,000
Maximum Production 4,000 3,500 4,000 4,500 4,000 3,500
Minimum Production 2,000 1,750 2,000 2,250 2,000 1,750
Beginning inventory = 2,750 units, Maximum warehouse
capacity = 6,000 units, Safety stock = 1,500 units
Unit carrying cost = 1.5% of unit production cost

Investment Planning
Upton Corporation- Production Planning

Upton is planning the production of their heavy-duty air
compressors for the next 6 months.
1 2 3 4 5 6
Unit Production Cost $240 $250 $265 $285 $280 $260
Units Demanded 1,000 4,500 6,000 5,500 3,500 4,000
Maximum Production 4,000 3,500 4,000 4,500 4,000 3,500
Minimum Production 2,000 1,750 2,000 2,250 2,000 1,750
Beginning inventory = 2,750 units, Maximum warehouse
capacity = 6,000 units, Safety stock = 1,500 units
Unit carrying cost = 1.5% of unit production cost
Objective: To identify the production and inventory plan for the
next six months that will meet the expected demand each month
while minimizing production and inventory costs.
Investment Planning
Algebraic Formlation
Decision Variables

Investment Planning
Decision Variables
Pi = Number of units to produce in month i; i = 1, . . . , 6

Investment Planning
Decision Variables
Derived Variables
Ii = Beginning inventory for month i; i = 1, . . . , 7

Investment Planning
Decision Variables
Derived Variables
Inventory Costs: (1.5% of Production Cost in same month)

Investment Planning
Decision Variables
Derived Variables
1 2 3 4 5 6
$3.60 $3.75 $3.98 $4.28 $4.20 $3.9

Investment Planning
Decision Variables
Derived Variables
1 2 3 4 5 6
$3.60 $3.75 $3.98 $4.28 $4.20 $3.9
Objective Function
Minimize Costs:Production Costs + Inventory Costs

Investment Planning
Decision Variables
Derived Variables
1 2 3 4 5 6
$3.60 $3.75 $3.98 $4.28 $4.20 $3.9
Objective Function
Production Costs:

Investment Planning
Decision Variables
Derived Variables
1 2 3 4 5 6
$3.60 $3.75 $3.98 $4.28 $4.20 $3.9
Objective Function
Production Costs:
240P1 + 250P2 + 265P3 + 285P4 + 280P5 + 260P6

Investment Planning
Decision Variables
Derived Variables
1 2 3 4 5 6
$3.60 $3.75 $3.98 $4.28 $4.20 $3.9
Objective Function
Production Costs:
240P1 + 250P2 + 265P3 + 285P4 + 280P5 + 260P6
Inventory Costs:

Investment Planning
Decision Variables
Derived Variables
1 2 3 4 5 6
$3.60 $3.75 $3.98 $4.28 $4.20 $3.9
Objective Function
Production Costs:
240P1 + 250P2 + 265P3 + 285P4 + 280P5 + 260P6
Inventory Costs:
3.6(I1 + I2 )/2 + 3.75(I2 + I3 )/2 + 3.98(I3 + I4 )/2+
4.28(I4 + I5 )/2 + 4.20(I5 + I6 )/2 + 3.9(I6 + I7 )/2
Investment Planning
Algebraic Formulation: Constraints

Investment Planning

Production Capacity Constraints

Investment Planning

2000 ≤ P1 ≤ 4000
1750 ≤ P2 ≤ 3500
2000 ≤ P3 ≤ 4000
2250 ≤ P4 ≤ 4500
2000 ≤ P5 ≤ 4000
1750 ≤ P6 ≤ 3500

Investment Planning

2000 ≤ P1 ≤ 4000
1750 ≤ P2 ≤ 3500
2000 ≤ P3 ≤ 4000
2250 ≤ P4 ≤ 4500
2000 ≤ P5 ≤ 4000
1750 ≤ P6 ≤ 3500
Ending Inventory =Beginning Inventory + Units Produced-Units Sold

Investment Planning

2000 ≤ P1 ≤ 4000
1750 ≤ P2 ≤ 3500
2000 ≤ P3 ≤ 4000
2250 ≤ P4 ≤ 4500
2000 ≤ P5 ≤ 4000
1750 ≤ P6 ≤ 3500
Ending Inventory =Beginning Inventory + Units Produced-Units Sold
1500 ≤ I1 + P1 − 1000 ≤ 6000
1500 ≤ I2 + P2 − 4500 ≤ 6000
1500 ≤ I3 + P3 − 6000 ≤ 6000
1500 ≤ I4 + P4 − 5500 ≤ 6000
1500 ≤ I5 + P5 − 3500 ≤ 6000
1500 ≤ I6 + P6 − 4000 ≤ 6000
Investment Planning

Investment Planning
Inventory Balance Constraints

I1 = 2750
I2 = I1 + P1 − 1000
I3 = I2 + P2 − 4500
I4 = I3 + P3 − 6000
I5 = I4 + P4 − 5500
I6 = I5 + P5 − 3500
I7 = I6 + P6 − 4000

Investment Planning
Outline

Investment Planning
Blending Problem
Production Planning

Investment Planning
Multiperiod Cash-Flow Problem: Taco Viva
Taco-Viva needs a sinking fund to pay $800,000 in building

costs for a new restaurant in the next 6 months.
Payments of $250,000 are due at the end of months 2 and
4, and a final payment of $300,000 is due at the end of
month 6.
The following investments may be used.
Investment Available Months Yield at Maturity

in Month Maturity
A 1, 2, 3, 4, 5, 6 1 1.8%
B 1, 3, 5 2 3.5%
C 1, 4 3 5.8%
D 1 6 11.0%
Investment Planning
Objective
The management at Taco-Viva needs to determine the

investment plan that allows them to meet the required
schedule of payments while placing the least amount of
money in the construction fund.

Investment Planning
Decision Variables

Investment Planning
Decision Variables
Ai = amount (in $1,000s) placed in investment A
at the beginning of month i = 1, 2, 3, 4, 5, 6
Bi = amount (in $1,000s) placed in investment B
at the beginning of month i = 1, 3, 5
Ci = amount (in $1,000s) placed in investment C
at the beginning of month i = 1, 4
Di = amount (in $1,000s) placed in investment D
at the beginning of month i = 1

Investment Planning
Decision Variables
Objective Function
Minimize the total cash invested in month 1.

Investment Planning
Decision Variables
Objective Function
Minimize the total cash invested in month 1.
A1 + B1 + C1 + D1

Investment Planning
Constraints

Investment Planning
Constraints
Total Amount maturing at the beginning of the month =
Total amount reinvested at the beginning of the month +
Payment due at the beginning of the month

Investment Planning
Constraints
Cash Flow Constraints

Investment Planning
Constraints
Cash Flow Constraints
1.018A1 − A2 = 0 month 2
1.035B1 + 1.018A2 − 1A3 − 1B3 = 250 month 3
1.058C1 + 1.018A3 − 1A4 − 1C4 = 0 month 4
1.035B3 + 1.018A4 − 1A5 − 1B5 = 250 month 5
1.018A5 − 1A6 = 0 month 6
1.11D1 + 1.058C4 + 1.035B5 + 1.018A6 = 300 month 7
Nonnegativity Constraints A1 , Bi , Ci , Di ≥ 0; ∀ i

Introduction
When solving an LP problem we assume that values of all

model coefficients are known with certainty.
Such certainty rarely exists.
Sensitivity analysis helps answer questions about how
sensitive the optimal solution is to changes in various
coefficients in a model.

Approaches to Sensitivity Analysis
Change the data and re-solve the model!

Sometimes this is the only practical approach.
Solver also produces sensitivity reports that can answer
various questions

Solver’s Sensitivity Report
Answers questions about:

Amounts by which objective function coefficients can
change without changing the optimal solution.
The impact on the optimal objective function value of
changes in constrained resources.
The impact on the optimal objective function value of
forced changes in decision variables.
The impact changes in constraint coefficients will have on
the optimal solution.

Class 3 LP Sensitivity

Uploaded by

Copyright:

Available Formats

You might also like

Class 3 LP Sensitivity

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Class 3 LP Sensitivity

Uploaded by

Copyright:

Available Formats

Linear Programming Formulations

Class 3: Linear Programming Formulations

October 19, 2016

Class 3: Linear Programming Formulations -Continued

Review from Last Session

Class 3: Linear Programming Formulations -Continued

1 Linear Programming Formulations

Class 3: Linear Programming Formulations -Continued

1 Linear Programming Formulations

Class 3: Linear Programming Formulations -Continued

Retirement Planning Services

A client wishes to invest $750,000 in the following bonds.

Class 3: Linear Programming Formulations -Continued

Restrictions and Problem

Class 3: Linear Programming Formulations -Continued

Restrictions and Problem

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Solving the Model

Class 3: Linear Programming Formulations -Continued

Solving the Model

Class 3: Linear Programming Formulations -Continued

1 Linear Programming Formulations

Class 3: Linear Programming Formulations -Continued

Blending Problem: Agri-Pro-Optimal Mix

Agri-Pro has received an order for 8,000 pounds of chicken

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

Class 3: Linear Programming Formulations -Continued

0.3X1 +0.05X2 +0.20X3 +0.10X4

0.1X1 +0.3X2 +0.15X3 +0.1X4

0.2X1 +0.2X2 +0.2X3 +0.3X4

Class 3: Linear Programming Formulations -Continued

0.3X1 +0.05X2 +0.20X3 +0.10X4

0.1X1 +0.3X2 +0.15X3 +0.1X4

0.2X1 +0.2X2 +0.2X3 +0.3X4

Class 3: Linear Programming Formulations -Continued

A Comment about Scaling

Notice the coefficient for X2 in the “corn” constraint is

Class 3: Linear Programming Formulations -Continued

Algebraic Formulation of Rescaled Model

0.3X1 +0.05X2 +0.20X3 +0.10X4

0.1X1 +0.3X2 +0.15X3 +0.1X4

0.2X1 +0.2X2 +0.2X3 +0.3X4

Scaling: Before and After

The problem is now more evenly scaled

Class 3: Linear Programming Formulations -Continued

1 Linear Programming Formulations

Class 3: Linear Programming Formulations -Continued