一种具有初始化功能的自适应惯性权重粒子群算法  

A Particle Swarm Optimization Algorithm with Initialized Adaptive Inertia Weights

作　　者：杜继永[1] 张凤鸣[1] 李建文[2] 杨骥[1] 

机构地区：[1]空军工程大学工程学院 [2]中国人民解放军95961部队

出　　处：《信息与控制》2012年第2期165-169,共5页Information and Control

摘　　要：提出了一种改进的具有初始化功能的自适应惯性权重粒子群优化(PSO)算法.该算法首先引入自适应惯性权重策略均衡全局和局部搜索能力,并针对运行过程中出现停滞现象的粒子群,围绕其加权重心位置重新初始化,引导粒子突破了局部极值的限制,提高了算法的收敛速度.最后,将此算法、PSO算法及惯性权重线性递减的PSO(LDW-PSO)算法进行了比较.实验结果表明,该算法不仅有效地增强了粒子突破局部极值的能力,而且算法的收敛速度和稳定性也有了一定的提高.An improved particle swarm optimization（PSO） algorithm with initialized adaptive inertia weights is proposed.An adaptive inertia weight strategy is introduced into the algorithm to balance the global and the local search abilities.In the evolution process,the swarm trapping at the stagnation is reinitialized around its weighted centroid position,which guides the swarm to avoid local extremum and accelerates the convergence rate of the algorithm.The comparison among the performances of the proposed approach,the standard PSO and the LDW-PSO is done.The experiment results show that the proposed algorithm can not only enhance the ability of avoiding local extremum,but also speed up the convergence rate and improve the stability to a certain degree.

关 键 词：粒子群优化算法 自适应 惯性权重 初始化 加权重心 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]