基于分布种群遗传算法的二自由度PID控制器参数优化整定  

The distribution population-based genetic algorithm for parameters optimization of two-degree-of-freedom PID controllers

作　　者：吴子云 黄少锋[2] 陈庆庚 

机构地区：[1]浙江机电职业学院,浙江杭州310023 [2]中船重工第715研究所 [3]浙江大学先进控制研究所工业控制技术国家重点实验室,浙江杭州310027

出　　处：《工业仪表与自动化装置》2008年第1期7-9,38,共4页Industrial Instrumentation & Automation

摘　　要：利用一种基于分布种群的遗传算法,对整定困难的二自由度PID控制器的参数进行优化整定,并给出了该遗传算法的数学描述。该算法将寻优空间划分为多个不相交子空间,分别对在各子空间内的种群进行遗传操作,评价种群适应度,细化解空间,达到在较小的范围做精细搜索的目的,兼顾了搜索的广度和深度,提高了寻优的搜索性能,克服了常规遗传算法易陷入局部极值的缺陷。仿真试验证明了其有效性。The distribution population- based genetic algorithm（DPGA） and its mathematic description are presented for two-degree-of-freedom PID controllers optimization in this paper. The searching capability of the algorithm is improved by competition between distribution populations to reduce the search zone. A good balance of search width and search length is concerned. Applying it to design optimal parameters of two-degree-of-freedom PID controllers with examples,the simulation results show that satisfactory performances are obtained.

关 键 词：分布种群 遗传算法 二自由度PID控制器 收敛性 参数优化 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]