Interpersonal learning Particle Swarm Optimizer
Ji Ma, Junqi Zhang, Linwei Xu · 2015
Particle Swarm Optimizer (PSO) is a population-based optimization technique that can be applied to a wide range of problems. However, except learning from itself (pbest), the way of information exchanging between a particle and the entire swarm is only learning from the best particle (gbest) in regular PSO and the interaction is “one-way”, which means a particle in the swarm just assimilate the best information passively, and its pbest will never propagate to other particles if the pbest is not the gbest. As a result, some potential useful information is ignored. This paper presents a variant of PSO that we called Interpersonal Learning Particle Swarm Optimizer (ILPSO), which uses a novel learning strategy whereby a particle can learn from any other particles in the swarm interpersonally, and the interaction is bidirectional. This strategy enhances the diversity of the swarm and alleviate premature. Experiments are performed on 20 benchmark functions and show the algorithm has competitive performance to other six well-known PSOs.