Gaussian swarm a novel particle swarm optimization algorithm.pdf

Proceedings of the 2004 lEEE Conference on Cybernetics and Intelligent Systems Singapore, 1-3 December, 2004 Gaussian Swarm: A Novel Particle Swarm Optimization Algorithm Renato A. Krohling Lehrstuhl Elektrische Steuerung und Regclung (ESR) Fakult3t ftir Elektrotechnik und lnformationstechnik UniversiBt Dortmund D-4422 1 Dortmund, Germany E-mail: rcnato.krohling@uni-dortmund.de Abstract--In this paper, a novel particle swarm optimization algorithm based on the Gaussian probability distribution is ’ proposed. The standard Particle Swarm optimization (PSO) algorithm has some parameters that need to be specified before using the algorithm, e.g., the accclerating constants cI and c2, the inertia weight w, the maximum velocity Vmnr, and the number of particles of the swarm. The purpose of this work is the development of an algorithm based on the Gaussian distribution, which improves the convergence ability of PSO without the necessity of tuning thcsc parameters. The only parameter to be specified by the user is the number of particles. The Gaussian PSO algorithm was tested on a suite of well-known benchmark functions and thc results were compared with the results of the standard PSO algorithm. The simulation results shows that the Gaussian Swarm outpcrforms the standard one. ’ Keywords: Particle Swarm Optimiziztion, Gaussian distribution, nonlinear optimizution. 1. INTRODUCTION Particle Swarm Optimization (PSO) originally developed by Kennedy and Eberhart [l], [2] is a population-based algorithm. PSO is initialized with a population of candidate solutions. Each candidate solution in PSO, called particle, has associated a randomized vclocity, moves through the search space. Each particle keeps track of its coordinates in the search space, which are associated with the best solution ’ (fitness) it has achieved so far, pbest. Another “best’ value tracked by the global version of the particle swarm optimizer is the overall best value, g