1 / 17

# Index - PowerPoint PPT Presentation

Index. Introduction Analytic Framework Illustrative Example Conclusions. End. Introduction. GP is the most widely used MCDM approach Realistic Satisficing Philosophy Variant used: (Tamiz et al., 1995) 64%, lexicographic 21%, weighted Rest, minmax

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.

## PowerPoint Slideshow about ' Index' - elaine

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

• Introduction

• Analytic Framework

• Illustrative Example

• Conclusions

End

• GP is the most widely used MCDM approach

• Realistic Satisficing Philosophy

• Variant used: (Tamiz et al., 1995)

• 64%, lexicographic

• 21%, weighted

• Rest, minmax

• The variant chosen critically affects the final solution.

(t1, t2)

Feasible set

Introduction

Weighted

Minmax

Lexicographic

• where:

• all functions gj are convex,

• all functions fi are concave,

• all goals derive from attributes “more is better”

• Minmax:

• Lexicographic. Levels 1,..., l

Analytic Framework

Classical GP variants:

• Type 1. The percentage sum of unwanted deviation variables cannot surpass a certain bound Q1.

• Type 2. The maximum percentage deviation cannot surpass a certain bound Q2.

Q1

Meta-Satisfying

Set

Q2

Feasible set

Analytic Framework

Meta-Goal

type 2

Meta-Goal

type 1

META-GOALS

• Type 3. The percentage of unachieved goals cannot surpass a certain bound Q3

Analytic Framework

META-GOALS

• Type 3 goal on a set

• Type 2 goal on a set

Analytic Framework

• General Formulation of the Meta-Goal

• Programming model, with:

• m1 type 1 meta-goals,

• m2 type 2 meta-goals,

• m3 type 3 meta-goals

Gross margin-break-even point

g5: x1 + n5 - p5 = 300

Unwanted deviation variables

Production capacities

g6: x2 + n6 - p6 = 200

Employment

Gross margin

Illustrative Example

Hypothetical Production Planning Problem

g1: x1 + 2x2 + n1 - p1 = 300

g2: 100x1 + 300x2 + n2 - p2 = 50000

g3: 100x1 + 300x2 + n3 - p3 = 90000

g4: x1 + x2 + n4 - p4 = 500

min f ( p1, n2, n3, n4, p5, p6)

• Lexicographic Variant

• Level 1: Goals 2, 5 and 6

• Level 2: Goals 1, 3 and 4

SOLUTION:

• Decision Variables:

• x1 = 300; x2 = 66,66

• Level 1:

• n2 = 0; p5 = 0; p6 = 0

• Level 2:

• p1 = 133,33; n3 = 40000; n4 = 133,33

• D.M. says:

• With respect to the number of satisfied goals:

• - Satisfy goals 2, 5 and 6;

• - Maximize the number of satisfied goals among 1, 3 and 4;

• Aggregate percentage deviation of not more than a 100% in the second level;

• Maximum percentage disagreement of not more than a 75% in the second level

INFEASIBLE

(GP)2 Model

SOLUTION:

• Level 1: ( n2 = p5 = p6 = 0 )

• 1 = 0;

• Level 2: ( p1 = 400; n3 = n4 = 0 )

• - 2 = 1/3; (1 unsatisfied goal)

• - 3 = 0.33; (133% aggregate disagreement)

• - 4 = 0.58; (133% maximum disagreement)

• Target values for several achievement functions

META-GP MODEL

Conclusions

• Choosing a single GP variant can be a too mechanistic way of incorporating the DM’s preferences into the model