PSOPB: A Two-population Particle Swarm Optimizer Mimicking Facultative Bio-parasitic Behavior

Abstract

Inspired by the phenomenon of bio-parasitic behavior in natural ecosystem, this paper presents a novel particle swarm optimizer named PSOPB, in which particles are composed of the host and the parasite population. In the presented algorithm, the two populations mimic facultative bio-parasitic behaviour and exchange particles according to particles' fitness values sorted of each population in a certain number of iterations. The parasite mutation and the host immunity are also considered to tie it closer to bio-parasitic behaviour as well as improve the algorithm performance. In order to embody the law of "survival of the fittest" in biological evolution, the particles with poor fitness value in the host population are removed and replaced by the same numbers of the re-initialization particles to maintain constant population size. The experimental results of a set of 10 benchmark functions demonstrate the presented algorithm's efficacy.