摘要
为了有效地控制PSO算法的全局和局部搜索能力,着重分析了PSO算法中加速因子对粒子收敛的影响,提出加速因子采用反余弦非对称策略能有效提高PSO算法的搜索性能。对5种加速因子策略使用4个著名的基准函数进行测试,试验结果表明,反余弦非对称方法可以使粒子在搜索的初期获得更好的多样性,在算法后期则可以有效增强粒子的搜索能力,从而使算法具有更强的摆脱局部极值的能力,提高算法的收敛精度。
In order to control the global and local search of particle swarm optimization(PSO)efficiently,this paper mainly analyzed the effect of acceleration coefficients in PSO and pointed out that dissymmetrical and arc cosine method could effectively enhance the performance of PSO.We used four benchmark functions to test 5 different setting strategies of acceleration coefficient.The experimental results show that the proposed nonlinear method can reserve the diversity of swarm at the early stages of the optimization,and it also can avoid local optimum searching at the later stages to improve the convergence of the PSO algorithm.
出处
《长江大学学报(自科版)(上旬)》
CAS
2007年第4期1-4,16,共5页
JOURNAL OF YANGTZE UNIVERSITY (NATURAL SCIENCE EDITION) SCI & ENG
基金
国家自然科学基金项目(10471083)
教育部科学技术研究重点基金项目(206073)
福建省自然科学基金项目(A0610012)
福建省科技厅重点项目(2004H007)