Abstract：Particle Swarm Optimization(PSO)has been demonstrated that it can yield good performance for solving a large of optimization problem.However,PSO is easy to fall into “premature”,and it tends to suffer from slow converge velocity and low precision when solving continuous function optimization problems.This paper proposed a new approach,called PSO based on elite opposition-based learning.In the processing of evolution,the modified algorithm adds opposition-based learning,chooses the dimension of the global optimal particle randomly,and enlarges the searching area,so that it increases the ability of global exploration and improves the probability of optimal solution searched by our algorithm.Then we introduce 4 kinds of model using opposition-based learning into our new approach to form 4 kinds of PSO algorithms using opposition-based learning.Experiments are conducted on 12 well-known benchmark functions by 4 different models;the results show that although 4 kinds of different PSO using opposition-based learning has better performance than PSO,there are large differences on efficiency and precision among them.
赵嘉,付平,李崇侠,吕莉. 基于不同学习模型的精英反向粒子群优化算法[J]. 小型微型计算机系统, 2015, 36(6): 1368-1372.
ZHAO Jia,FU Ping,LI Chong-xia,LV Li. Particle Swarm Optimization Based on Elite Opposition Learning Using Different Learning Models. Journal of Chinese Computer Systems, 2015, 36(6): 1368-1372.
2015, Vol. 36 
