Publication detail

Self-Organizing Migrating Algorithm for Dynamic Problems: An Experimental study

KOBLIHA, M. SCHWARZ, J.

Original Title

Self-Organizing Migrating Algorithm for Dynamic Problems: An Experimental study

English Title

Self-Organizing Migrating Algorithm for Dynamic Problems: An Experimental study

Type

conference paper

Language

en

Original Abstract

This paper is an experimental study investigating the capability of well known Self-Organizing Migrating Algorithm (SOMA) to solve dynamic problems. We have proposed a new extension of SOMA algorithm with two new approaches based on multi-population and elite concept. We tested the performance of the algorithm on the artificial benchmark with moving peaks. The experimental results confirmed the capability of the proposed dynamic xSODOMA algorithm to effectively adapt the search process towards the nonstationary global optimum.

English abstract

This paper is an experimental study investigating the capability of well known Self-Organizing Migrating Algorithm (SOMA) to solve dynamic problems. We have proposed a new extension of SOMA algorithm with two new approaches based on multi-population and elite concept. We tested the performance of the algorithm on the artificial benchmark with moving peaks. The experimental results confirmed the capability of the proposed dynamic xSODOMA algorithm to effectively adapt the search process towards the nonstationary global optimum.

Keywords

SOMA algorithm, dynamic environment, limited lifetime, multi-population, AllToElite strategy.

RIV year

2007

Released

05.09.2007

Publisher

Faculty of Mechanical Engineering BUT

Location

Brno

ISBN

978-80-214-3473-8

Book

13th International Conference on Soft Computing

Pages from

24

Pages to

29

Pages count

6

Documents

BibTex


@inproceedings{BUT28815,
  author="Miloš {Kobliha} and Josef {Schwarz}",
  title="Self-Organizing Migrating Algorithm for Dynamic Problems: An Experimental study",
  annote="This paper is an experimental study investigating the capability of well known
Self-Organizing Migrating Algorithm (SOMA) to solve dynamic problems. We have
proposed a new extension of SOMA algorithm with two new approaches based on
multi-population and elite concept. We tested the performance of the algorithm on
the artificial benchmark with moving peaks. The experimental results confirmed
the capability of the proposed dynamic xSODOMA algorithm to effectively adapt the
search process towards the nonstationary global optimum.",
  address="Faculty of Mechanical Engineering BUT",
  booktitle="13th International Conference on Soft Computing",
  chapter="28815",
  howpublished="print",
  institution="Faculty of Mechanical Engineering BUT",
  year="2007",
  month="september",
  pages="24--29",
  publisher="Faculty of Mechanical Engineering BUT",
  type="conference paper"
}