Title: Associations between population topologies and Gaussian dynamic particle swarm performance
Authors: Yang Chen; Yancheng Liu; Chuan Wang; Rui Ma
Addresses: Marine Engineering College, Dalian Maritime University, Dalian 116026, China ' Marine Engineering College, Dalian Maritime University, Dalian 116026, China ' Marine Engineering College, Dalian Maritime University, Dalian 116026, China ' Marine Engineering College, Dalian Maritime University, Dalian 116026, China
Abstract: Population topology can control the spread of the flow of information, and directly affect the final performance of Gaussian dynamic particle swarm optimisation (GDPSO) algorithm. The effects of various population topologies on the GDPSO were systematically investigated in this paper. Random graphs and special graphs were both generated to specifications, and their performances on several criteria were compared. The numerical experiment results show that the average degree of the population topology affects the accuracy and success rate; for a specified problem, there should be a trade-off between accuracy and success rate by choosing an appropriate population topology.
Keywords: Gaussian dynamic PSO; particle swarm optimisation; GDPSO; graph statistic; population topology; information flow; average degree.
DOI: 10.1504/IJMIC.2015.071887
International Journal of Modelling, Identification and Control, 2015 Vol.24 No.2, pp.138 - 148
Received: 09 Aug 2014
Accepted: 24 Feb 2015
Published online: 22 Sep 2015 *