Local Search Based Competitive Particle Swarm Optimizer With Opposition-Based Learning

• 1、Department of IoT Engineering,Jiangnan University, WuXi 214122

Abstract：In order to improve the optimization ability of particle swarm optimization(PSO) algorithm in complexoptimization problems,especially in high dimensional problems,a competitive particle swarm optimizationalgorithm(SW-OBLCSO) based on opposition-based learning and Solis\&Wets local search is proposed.SW-OBLCSO adopts two learning mechanisms,namely competitive swarm optimizer(CSO) and opposition-based learning,and embeds individual-based local search operators in it.In each iteration of SW-OBLCSO,competitive learning is performedbetween four randomly selected particles from the population.Particles with optimal fitness values are passed directly tothe next iteration.The particle with the worst fitness value directly obtains the position of the winner and adds an offsetto speed up the convergence of the algorithm.Two medium-adaptive particles update from the winner and by Opposition-Based Learning strategy respectively.After the performance of the solution has not been improved for many times,the SW local search strategy is added to improve the ability of the algorithm to jump out of local optimum.TheSW-OBLCSO algorithm is compared with various optimization algorithms on the 10 benchmarks functions in 100,500and 1000 dimensions.The experimental results show that the proposed algorithm exhibits outstanding performance inconvergence speed and global search ability.Testing of fuzzy cognitive map (FCM) learning problems shows thatSW-OBLCSO also has excellent performance when dealing with practical problems.

Keywords： Computer software and theory Particle Swarm Optimizer Opposition-Based Learning Competitive learning Large-scale optimization problem