A Basic Study of the Adaptive Particle Swarm Optimization

抄録

It is well known that Particle Swarm Optimization (PSO), which was originally proposed by J. Kennedy et al., is a powerful algorithm for solving unconstrained and constrained global optimization problems. Appropriate adjustment of its parameters, however, requires a lot of time and, labor when PSO is applied to real optimization problems. Therefore we have so far indicated that meta-heuristics should have robustness and adaptability from the engineering viewpoint and have proposed an adaptive algorithm.

In this paper, we extend the parameters of PSO in order to realize advanced adaptability, and we treat the selection strategy of global information gbest as a new parameter. Furthermore we propose an adaptive Particle Swarm Optimization method using a new parameter to accomplish effective search in global optimization. Some numerical simulations are carried out in order to examine the adaptability of the proposed approach.