Observing the swarm behaviour during its evolutionary design

摘要

Evolutionary Algorithms (EAs) can be used for designing Particle Swarm Optimization (PSO) algorithms that work, in some cases, considerably better than the human-designed ones. By analyzing the evolutionary process of design PSO algorithm we can identify different swarm phenomena (such as patterns or rules) that can give us deep insights about the swarm's behaviours. The observed rules can help us to design better PSO algorithms for optimization. In this paper we investigate and analyze swarm phenomena by looking to process of evolving PSO algorithms. Several interesting facts are inferred from the strategy evolution process (the particle quality could influence the update order, some particles are updated more frequently than others are, the initial swarm size is not always optimal).