Interactive Self Improvement Based Adaptive Particle Swarm Optimization

Abstract

ABSTRACT One of the most familiar stochastic heuristic search algorithm is Particle swarm optimization (PSO), which is motivated by social behavior of animals like birds, fishes, and so forth. The significant advantages of PSO algorithm are simple structure and limited parameters to be used. Among the parameters, inertia weight is considered as the most crucial one in PSO which brings trade-off between the characteristics of exploitation and exploration. A novel Interactive Self-Improvement based Adaptive PSO (ISI-APSO) method that traits better searching efficiency and accuracy than the traditional particle swarm optimization is proposed. More precisely, it can achieve faster convergence speed while on global search over the entire search space. The simulation results show that the performance of our proposed ISI-APSO is substantially improved than other heuristic algorithms in terms of the search efficiency and convergence speed.