An Analysis of Multiple Particle Swarm Optimizers with Inertia Weight for Multi-objective Optimization

• Published in: IAENG International Journal of Computer Science Vol. 39; no. 2; pp. 190 - 199
• Main Author: Zhang, Hong
• Format: Journal Article
• Language: English
• Published: 01.06.2012
• Subjects: Agglomeration; Dynamics; Inertia; Mathematical models; Optimization; Proposals; Quantitative analysis; Searching
• ISSN: 1819-656X

An improved particle swarm optimizer with inertia weight (PSOIW alpha ) was applied to multi-objective optimization (MOO). For further improving its search performance, in this paper, we propose to use a cooperative PSO method called multiple particle swarm optimizers with inertia weight (MPSOIW alpha ) to search. The crucial idea of the MPSOIW alpha , here, is to reinforce the search ability of the PSOIWq by the union's power of plural swarms, i.e. distributed processing. To demonstrate the search performance and effect of the proposal, computer experiments on a suite of 2-objective optimization problems are carried out by an aggregation-based manner. The resulting Pareto-optimal solution distributions corresponding to each given problem indicate that the linear weighted aggregation among the adopted three kinds of dynamic weighted aggregations is the most suitable for acquiring better search results. Throughout quantitative analysis to experimental data, we clarify the search characteristics and performance effect of the MPSOIW alpha contrast with that of the original PSOIW, PSOIW alpha , and MPSOIW.