关键词: 粒子群算法, 反向学习, 灰狼优化算法, 贝塔分布

Abstract:

To overcome the shortcomings of Particle Swarm Optimization（PSO） such as premature convergence, poor ability to escape from local optimum and low accuracy, a reverse learning PSO algorithm based on Grey Wolf Optimization（GWO） is proposed. Firstly, the inverse learning strategy is used to generate the inverse solution for the optimal particle, which enlarges the searching range of the population and enhances the global searching ability of the algorithm. Secondly, a new social learning method is used to improve the searching efficiency and mining performance of the non-optimal particle. At the same time, aiming at the problem of low convergence accuracy of PSO, the GWO algorithm is introduced and the convergence factor is disturbed to balance the global and local search performance of the algorithm and improve its accuracy. The simulation results on CEC_2017 test function show that under the same experimental conditions, the improved PSO algorithm has significantly improved the convergence accuracy and speed, and its performance is significantly better than the standard PSO algorithm.

Key words: Particle Swarm Optimization（PSO）, inverse learning, Grey Wolf Optimization（GWO） algorithm, Beta distribution