摘要
针对简单遗传算法(SGA)收敛速度慢、易于早熟等缺点,在前人研究成果的基础上,提出动态调整搜索空间策略,对遗传算法进行多步渐进搜索。并采用改进的自适应交叉算子和自适应变异算子,结合兼顾性能指标和响应过程平衡的适配函数,以多种改进方式相结合的遗传算法对PID参数进行迭代寻优整定。仿真结果表明:当被控对象存在较大纯滞后、时间常数特性时,采用本方法优化PID控制器参数可获得比较满意的调节效果。
A strategy of adjusting searching space dynamically for iterative optimized tuning of PID parameters was developed in this paper by use of improved genetic algorithms of multiple ways. This method can be used to overcome the shortcomings of slow-convergence and easy-premature of simple genetic algorithms (SGA). It makes use of the improved adaptive crossover operator and the adaptive mutation operator, and combines with a kind of fitness function that considers the balance of the performance target and the step-respond process. The simulated result shows that the step-respond of PID controller parameters tuned by this method is satisfied when the controlled object contains a bigger time delay and large time constant characteristics.
出处
《北京化工大学学报（自然科学版）》
CAS
CSCD
北大核心
2005年第2期101-103,共3页
Journal of Beijing University of Chemical Technology(Natural Science Edition)
基金
中国石化总公司资助项目(X503014)
中国石油天然气集团公司资助项目(03E7042)
关键词
遗传算法
自适应交叉
早熟
PID参数整定
genetic algorithms
adaptive crossover
premature
PID parameter tuning
