Gaussian Mutated Particle Swarm Optimization with Modified Velocity for Function Optimization

Abstract

This Particle swarm optimization (PSO) is a clever optimization technique, which has shown a good performance in many real world and benchmark problems. In this paper, we propose an enhanced PSO alternate i.e. Gaussian mutated PSO with the modified velocity (GMVPSO), where a concept of Gaussian mutation is introduced along with modified velocity for particles. Simulation results show that this modified GMVPSO obtains better performance than its two counterparts that is PSO with Gaussian mutation as well as PSO with modified velocity and obviously than standard PSO. The result also verifies the algorithm by the variation of population size and in all the cases the new modified algorithm dominates its ancestors in all dimensions.