基于遗传算子的粒子群优化算法的比较分析  

Comparison and analysis of particle swarm optimization method based on genetic operator

作　　者：雷秀娟[1] 史忠科[2] 孙瑰琪[1] 

机构地区：[1]陕西师范大学计算机科学学院,西安710062 [2]西北工业大学自动化学院,西安710072

出　　处：《计算机工程与应用》2008年第14期65-66,79,共3页Computer Engineering and Applications

基　　金：教育部科学技术研究重点项目(No.107106);教育部高等学校科技创新工程重大项目培育基金项目

摘　　要：为了深入分析探讨改进的粒子群优化算法的性能,针对典型的函数优化问题,设计了3种方案:(1)采用线性递减惯性权重的PSO;(2)基于遗传算子的PSO;(3)在方案(2)基础上,加入收缩因子χ。在MATLAB7.0中对常用的测试函数进行优化仿真,发现当融合遗传算子和收缩因子时,算法性能最优。To analyse the performance of improved particle swarm optimization method deeply, three strategies are designed to experiment several standard test functions optimization problem. One of the strategies is linear inertia weight reduction, the second is the PSO with genetic operator, and the third is rejoining the constriction factor based on the second strategy. Through the optimization and simulation of the test functions in MATLAB 7.0 software, the performance of the PSO mixed the genetic operator and constriction factor is best of all.

关 键 词：粒子群优化 线性递减惯性权重 遗传算子 收缩因子 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]