摘要
根据带约束因子PSO算法,推导出认知因子c1、社会因子c2和惯性权重w之间应满足的关系.提出新的DCF-PSO算法,随着其中的惯性权重非线性递减,动态调整c1和c2值.通过Benchmark验证了改进后算法的高效性能.实验结果表明,算法表现优异.
According to the Constrain Factor Particle Swarm Optimization, the relation of c1, which is the self confidence factor, c2, which is the swarm confidence factor, and w, which is inertia weight factor is obtained. This paperput forward a new algorithm called DCF-PSO, in which the factor c1 and c2 will automatically adjust with the nonlinear inertia weight decreasing ofw. The Benchmark function test shows the excellent performance of DCF-PSO.
出处
《河北工业大学学报》
CAS
北大核心
2010年第3期51-55,共5页
Journal of Hebei University of Technology
基金
广东省高等教育教学改革工程项目(BKYBJG20060257)
关键词
动态约束因子
粒子群
惯性权重非线性递减
进化计算
dynamic constrain factor
particle swarm optimization
nonlinear inertia weight decreasing
evolutionary computation
