An Improved Particle Swarm Optimization Algorithm and its Convergence Analysis

Abstract

To avoid falling into local optimum solution and improve global optimum efficiency and accuracy of particle swarm optimization, a novel particle swarm optimization model with centroid of population is proposed, which can enhance inter-particle cooperation and information sharing capabilities effectively, then the guidelines of parameter selection are obtained in the case of convergence of the new model. Simulation results of Benchmark functions are also analyzed in detail, and show the new algorithm is more feasible and efficient then standard particle swarm optimization method.