摘要
为了实现1种方便易用、寻优能力良好的PSO算法,寻找1种以惯性权重矩阵为基础的自适应粒子群算法(RDR-PSO)极为重要。从离散状态空间表达式入手,探讨粒子群算法的稳定性,获得改进后的自适应粒子群算法(RDR-PSO),并对RDR-PSO进行仿真验证。结果表明,与其他改进后PSO算法与标准PSO算法性能进行比较分析,得到RDR-PSO算法有着可以提升收敛准确度与跳出局部极数的特征,说明RDR-PSO收敛准确性高、全局寻优速度快、方便易用。
In order to realize a convenient PSO algorithm with good optimization ability,it is very important to find an adaptive particle swarm optimization algorithm(RDR-PSO)based on inertial weight matrix.Starting from the discrete state space expression,the stability of particle swarm optimization(PSO)was discussed,and the improved adaptive particle swarm optimization(RDR-PSO)was obtained.The result showed that RDR-PSO algorithm has the characteristics of improving convergence accuracy and jumping out of local pole number by comparing the performance of PSO algorithm with that of standard PSO algorithm.It indicates that RDR-PSO has high convergence accuracy,fast global optimization speed,convenient which easy to use,and has a very broad space for development.
作者
杨宝军
YANG Baojun(Dept.of Applied Mathematics,Taiyuan University,Taiyuan 030012,China)
出处
《黑龙江工程学院学报》
CAS
2020年第5期38-41,共4页
Journal of Heilongjiang Institute of Technology
关键词
惯性权重矩阵
自适应粒子群算法
测试
inertial weight matrix
adaptive particle swarm optimization
test
