摘要
惯性权重是平衡粒子群算法中平衡全局搜索能力与局部搜索能力的重要参数。为实现快速收敛与并避免陷入局部最优,分析了PSO算法中的惯性权重与种群规模、粒子适应度以及搜索空间维度这三者的关系,并把粒子惯性权重定义为这三者的函数以改进PSO算法。该算法在每次迭代后根据此函数更新每个粒子的惯性权重,实现了自适应调整全局搜索能力与局部搜索能力,并结合动态管理种群的策略提出了改进的粒子群算法。通过在多个常用测试函数上与已有惯性权重调整算法测试比较,证明新算法具有较强的全局寻优能力与较高的搜索效率。
In order to get a better balance between global search ability and local search capabilities in the particle swarm algorithm,the relationships between inertia weight and the particle fitness,the population size and dimensions of the searching space are analyzed,and a function is constructed between them.After each iteration,the inertia weight of each particle is updated as to achieved a self-adaptive adjustment of global search ability and local search capabilities.A new improved particle swarm optimization is brought forward combined with population dynamic management strategy.The searching result of some standard testing functions proves that the new algorithm have a stronger global optimization capability and a higher search efficiency.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第13期3062-3065,共4页
Computer Engineering and Design
基金
湖南省教育厅科研基金项目(08C458)
关键词
粒子群算法
自适应惯性权重
种群规模
搜索空间维度
粒子适应度
动态管理种群
particle swarm optimization
adaptive inertia weight
population size
search space dimension
particle fitness
dynamic management of populations
