基于z平面分析的粒子群优化参数选择  

Parameter Selection in Particle Swarm Optimization Based on z-Plane Analysis

作　　者：张玮[1,2] 李化[2] 贺鸿[1] 王华奎[2] 

机构地区：[1]太原理工大学化学化工学院,太原030024 [2]太原理工大学信息工程学院,太原030024

出　　处：《计算机工程》2010年第4期155-157,共3页Computer Engineering

摘　　要：针对粒子群优化算法参数选择缺乏定量理论依据的问题,在z平面上研究粒子位置期望的动态特性,讨论复特征根的模值与相角对系统动态特性的影响,提出一组收敛效率高的参数序列。采用测试函数逐个与相关参数做仿真比较实验,证明其有效性。指出在改组参数序列中,当对应的特征根在虚轴附近时,全局寻优效率最高,随着特征根远离虚轴,寻优效率逐渐降低。The dynamic characteristics of position expectations on z-plane are studied considering the insufficient quantitative estimates and theoretical basis on Particle Swarm Optimization（PSO） parameter selection. The effect of module values and phase angles of the complex characteristic roots on the dynamic characteristics is discussed, and a set of optimum parameters is proposed after numerous experiments. Benchmark functions are used to test the effectiveness of the parameters comparing with those relevant parameters. It is found that the highest efficiency is achieved when the corresponding complex characteristic roots lie near imaginary axis and the efficiency may gradually decrease when the corresponding complex characteristic roots go far from imaginary axis

关 键 词：粒子群优化 z平面 参数选择 阻尼比 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]