Adaptive mutation based particle swarm optimization algorithm

In this paper an adaptive mutation based PSO (AMBPSO) is presented for improvement of deficiencies of standard PSO, which is modified by the combination of dynamic adjustment of the inertia weights, the update of position and velocity of each particle by means of randomly adaptive mutation, and the limit of the update for the change in a reasonable range. The optimization results of two standard test functions show that these modifications can enhance particles´ activity to improve the algorithm´s search precision and convergence speed and to keep away from easily immerging in local minima efficiently compared with standard PSO and general PSO.