On the disruption-level of polynomial mutation for evolutionary multi-objective optimisation algorithms

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

This paper looks at two variants of polynomial mutation used in various evolutionary optimisation algorithms for mutliobjective problems. The first is a non-highly disruptive and the second is a highly disruptive mutation. Both are used for problems with box constraints. A new hybrid polynomial mutation that combines the benefits of both is proposed and implemented. The experiments with three evolutionary multi-objective algorithms on well-known multi-objective optimisation problems show the difference in terms of generational distance, hypervolume, convergence speed and hit rate metrics. The hybrid polynomial mutation in general retains the advantages of both versions in the same algorithm.

Original languageEnglish
Pages (from-to)783-800
Number of pages18
JournalComputing and Informatics
Volume29
Issue number5
Publication statusPublished - 2010

Keywords

  • Evolutionary algorithms
  • Multi-objective optimisation
  • Polynomial mutation

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Networks and Communications
  • Computational Theory and Mathematics

Fingerprint Dive into the research topics of 'On the disruption-level of polynomial mutation for evolutionary multi-objective optimisation algorithms'. Together they form a unique fingerprint.

  • Cite this