摘要
分析了含维变异算子的粒子群优化算法全局搜索能力与收敛速度的矛盾,提出了动态惯性权重向量和维变异的改进粒子群优化算法。算法首先定义了维多样性的概念,根据维多样性的情况动态地调整惯性权重向量,并对维多样性最差的维进行变异。4个典型测试函数的仿真实验说明该算法具有更强的全局搜索能力和更快的收敛速度。
The contradiction of the global exploration and convergence speed of particle swarm optimization with dimension mutation operator is analyzed,and an improved algorithm（WPSO） is proposed by modifying PSO with dimension mutation based on dynamical inertial weight vector.In the proposed algorithm,the concept of dimension diversity is defined and inertial weight vector will be updated dynamically according to dimension diversity.The mutation operates on dimension whose dimension diversity is the worst.The simulation on four typical test functions indicates that the proposed algorithm has more powerful global exploration ability and faster convergence speed.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第5期29-31,49,共4页
Computer Engineering and Applications
基金
国家自然科学基金No.60874070
高校博士点基金项目(No.20070533131)~~
关键词
粒子群优化
维多样性
惯性权重向量
维变异
particle swarm optimization
dimension diversity
inertial weight vector
dimension mutation
