The Open Cybernetics & Systemics Journal

2008, 2 : 173-179
Published online 2008 June 6. DOI: 10.2174/1874110X00802010173
Publisher ID: TOCSJ-2-173

Particle Swarms and Nonextensive Statistics for Nonlinear Optimisation

A.D. Anastasiadis and G.D. Magoulas
School of Computer Science and Information Systems, Birkbeck College, University of London, Malet Street, London WC1E 7HX, UK.

ABSTRACT

Particle swarm methods are inspired from the dynamics of social interaction and employ information sharing to seek solutions to difficult optimisation problems. In this paper we introduce an approach that combines ideas from particle swarm optimisation (PSO) and the theory of nonextensive statistical mechanics. We develop two algorithms that adopt this approach and conduct an experimental study using benchmark functions to investigate their effectiveness in nonlinear optimisation. Results appear to be promising, as the tested algorithms outperform in most cases the standard PSO and other, recently proposed, PSO variants.

Keywords:

Particle swarm optimiser, global search methods, statistical mechanics.