Particle Swarm Optimization (PSO) Variants with Triangular Mutation

Abstract

Particle swarm optimization (PSO) is a stochastic algorithm used for the optimization problems proposed by Kenned in 1995. It is a very good technique for optimization problems. However, there is a drawback - it stuck in the local minima. To improve the performance of PSO, many researchers proposed different variants of PSO. Some of the efforts include improving the initialization of the swarm, introduce new parameters - constriction coefficient and inertia weight, define the different method of inertia weight to improve the performance of PSO, and modifying the global and local best particles by introducing the mutation operators in the PSO. In this paper, we will see the different variants of PSO with respect to initialization, inertia weight and mutation operators. We also proposed a new PSO technique using triangular mutation. The new technique is tested on the benchmarked functions. The results show better performance when compared to four previous PSO variants.