摘要
在基本PSO算法和线性权重下降PSO算法的基础上,提出一种并行PSO算法,将粒子群分成两组,分别采用不同的惯性权重,各侧重于全局搜索和局部搜索,根据进化代数动态调整两种算法中进化的粒子数。通过仿真实验,证明了并行PSO算法的寻优性能优于基本PSO算法和线性权重下降PSO算法。
A parallel particle swarm optimization （PSO） algorithm is proposed based on basic PSO algorithm and LWDPSO algorithm. The particle swarm is divided into two groups, and different inertia weights are employed for global search and local search respectively by using parallel PSO algorithm. Parallel variables are dynamically adapted according to the evolution stage. The simulations prove the parallel PSO algorithm has better optimization performance than the other two PSO algorithms.
出处
《电子技术应用》
北大核心
2010年第10期132-135,共4页
Application of Electronic Technique
基金
安徽省自然科学基金(090412065)
安庆市重点科技项目(20091003)
