Choosing leaders for multi-objective PSO algorithms using differential evolution

Wickramasinghe, U and Li, X 2008, 'Choosing leaders for multi-objective PSO algorithms using differential evolution', in X. Li et al (ed.) The 7th International Conference on Simulated Evolution and Learning, Proceedings, Melbourne, Australia, 7-10 December, 2008.


Document type: Conference Paper
Collection: Conference Papers

Title Choosing leaders for multi-objective PSO algorithms using differential evolution
Author(s) Wickramasinghe, U
Li, X
Year 2008
Conference name The 7th International Conference on Simulated Evolution and Learning
Conference location Melbourne, Australia
Conference dates 7-10 December, 2008
Proceedings title The 7th International Conference on Simulated Evolution and Learning, Proceedings
Editor(s) X. Li et al
Publisher Springer
Place of publication Berlin, Germany
Abstract The fast convergence of particle swarm algorithms can become a downside in multi-objective optimization problems when there are many local optimal fronts. In such a situation a multi-objective particle swarm algorithm may get stuck to a local Pareto optimal front. In this paper we propose a new approach in selecting leaders for the particles to follow, which in-turn will guide the algorithm towards the Pareto optimal front. The proposed algorithm uses a Differential Evolution operator to create the leaders. These leaders can successfully guide the other particles towards the Pareto optimal front for various types of test problems. This simple yet robust algorithm is effective compared with existing multi-objective particle swarm algorithms.
Subjects Simulation and Modelling
DOI - identifier 10.1007/978-3-540-89694-4_26
Copyright notice © Springer-Verlag Berlin Heidelberg 2008
ISBN 0302 9743
Versions
Version Filter Type
Altmetric details:
Access Statistics: 222 Abstract Views  -  Detailed Statistics
Created: Mon, 19 Oct 2009, 07:39:41 EST by Catalyst Administrator
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us