关键词: 粒子群优化, 改进的粒子群优化, 进化计算

Abstract: To overcome the problem of premature convergence on particle swarm optimization(PSO), this paper proposes an improved particle swarm optimization(IPSO) called keeping particles active PSO, which is guaranteed to keep the diversity of the particle swarm. When particles lose activity, this paper uses a special mutation or perturbation to activate particles and to make particles explore the search space more efficiently. Four benchmark functions are selected as the test functions. The experimental results show that the IPSO can not only significantly speed up the convergence, but also effectively solve the premature convergence problem.

Key words: particle swarm optimization, improved particle swarm optimization, evolutionary computation