Analysis of Gaussian & Cauchy Mutations in Modified Particle Swarm Optimization Algorithm

This paper provides an analysis of the application of Cauchy mutation and Gaussian mutation in one of the modified version of the popular and well known particle swarm optimization technique. The main objective is to provide better convergence as well finest results in the solutions of various real world applications. The particle swarm optimization is generally acted as a foundation in the area of swarm intelligence. The modified PSO used a improved weight factor to provide a better convergence than the conventional PSO. But many real world problems need a further refinement in the solution. Therefore, in this paper two types of well known mutations are used with the modified PSO for betterment in the refining process. The proposed mutations are Gaussian mutation and Cauchy mutation. The proposed mutated modified PSO algorithms are simulated by the use of standard benchmarking functions to establish their dominance over the already modified PSO. Between Cauchy MPSO and Gaussian MPSO, the overall performance of Cauchy MPSO is better. The use of Cauchy mutation as well as Gaussian mutation in the modified PSO helps in the improvement of the convergence of the modified PSO for approaching much better solutions.