自适应惯性权重的分组并行粒子群优化算法-计算机工程与应用.pdf

40 2014 ，50（8） Computer Engineering and Applications 计算机工程与应用 自适应惯性权重的分组并行粒子群优化算法 1 2 周飞红 ，廖子贞 1 2 ZHOU Feihong , LIAO Zizhen 1.湖南涉外经济学院 电气与信息工程学院，长沙 410205 2.长沙市人力资源与社会保障局 信息中心，长沙 410011 1.College of Electrical and Information Engineering, Hunan International Economics University, Changsha 410205, China 2.Information Center, Changsha Human Resources and Social Security Bureau, Changsha 410011, China ZHOU Feihong, LIAO Zizhen. Grouping parallel particle swarm optimization algorithm with adaptive inertia weight. Computer Engineering and Applications, 2014, 50 （8 ）：40-44. Abstract ：No fundamental change in the particle velocity update for the island model parallel particle swarm optimiza- tion, a grouping parallel particle swarm optimization with adaptive inertia weight is proposed in this paper. This algorithm can adaptively choose the number of particles joining the group in an iterative process, besides, it can adjust the inertia weight of each group adaptively in accordance with the change of the optimal position. Each grouping uses multithreading technology to parallel processing and new information sharing mode is used between particles. The simulation results show that the algorithm has higher convergence speed and convergence precision. Key words ：grouping; parallel; particle swarm; adaptive; inertia weight 摘 要：针对岛屿模型的并行粒子群算法没有根本改变粒子速度更新的问题，提出一种自适应惯性权重的分组并行 粒子群优化算法。该算法在迭代过程中能自适应地选择加入分组的数量，同时对各组粒子的惯性权重按照组内最 优位置的变化进行自适应调整。各组运用多线程技术并行处理，粒子间采用新的信息共享的方式。仿真结果证实， 该算法具有较高的收敛速度和收敛精度。 关键词：分组；并行；粒子群；自适应；惯性权重 文献标志码：A 中图分类号：TP 18 doi ：10.3778/j.issn. 1002-8331.1207-0262 1 引言 Optimization ，FIPSO ），充分利用所有邻域个体信息，从 [3] 粒子群优化（Particle Swarm Optimization ，PSO ）算 而提高种群的多样性