Adaptive particle swarm optimization with mutation

When an individual is closed to the optimal particle, its velocity will approximate to zero. This is the main reason why particle swarm optimization (PSO) algorithm is prone to trap into local minima. A new improved particle swarm optimization (IPSO) is proposed, in which is guaranteed to converge to the global optimization solution with probability one. During the running time, the mutation probability for the current particle is determined by the variance of the individual´s concentration and convergence function. The ability of IPSO to break away from the local optimum is greatly improved by the mutation. The concept of adaptive acceleration factor is introduced to the IPSO. In this manner, the global and local search capability can be coordinated to make for locating the global optimum quickly. Finally, IPSO is applied to optimize several test functions. Test results show that IPSO can find global optima effectively.