Feedback-control operators for evolutionary multiobjective optimization

Ricardo H.C. Takahashi, Frederico G. Guimarães, Elizabeth F. Wanner, Eduardo G. Carrano

Research output: Chapter in Book/Published conference outputConference publication

Abstract

New operators for Multi-Objective Evolutionary Algorithms (MOEA's) are presented here, including one archive-set reduction procedure and two mutation operators, one of them to be applied on the population and the other one on the archive set. Such operators are based on the assignment of "spheres" to the points in the objective space, with the interpretation of a "representative region". The main contribution of this work is the employment of feedback control principles (PI control) within the archive-set reduction procedure and the archive-set mutation operator, in order to achieve a well-distributed Pareto-set solution sample. An example EMOA is presented, in order to illustrate the effect of the proposed operators. The dynamic effect of the feedback control scheme is shown to explain a high performance of this algorithm in the task of Pareto-set covering.

Original languageEnglish
Title of host publicationEvolutionary Multi-Criterion Optimization - 5th International Conference, EMO 2009, Proceedings
Pages66-80
Number of pages15
DOIs
Publication statusPublished - 2009
Event5th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2009 - Nantes, France
Duration: 7 Apr 200910 Apr 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5467 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference5th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2009
Country/TerritoryFrance
CityNantes
Period7/04/0910/04/09

Fingerprint

Dive into the research topics of 'Feedback-control operators for evolutionary multiobjective optimization'. Together they form a unique fingerprint.

Cite this