| 遗传算法在PID控制器参数寻优中的应用 |
| |
| 引用本文： | 江敏,陈一民.遗传算法在PID控制器参数寻优中的应用[J].兰州理工大学学报,2006,32(6):85-88. |
| |
| 作者姓名： | 江敏 陈一民 |
| |
| 作者单位： | 上海大学,计算机科学与技术学院,上海,200436;上海应用技术学院,上海,200235;上海大学,计算机科学与技术学院,上海,200436 |
| |
| 摘 要： | 针对遗传算法易早熟收敛,对参数有较强的依赖性,而模拟退火法的搜索过程冗长的缺点,采用混合策略,在优化机制、结构和行为上结合两者的特点,提高算法的优化能力、效率和可靠性.在PID参数寻优中采用混合策略,将SA状态产生函数作为变异概率可控的变异操作,在高温下有较强的变异功能,在低温下有较强的趋化功能.仿真结果说明,混合策略在PID参数寻优问题中明显优于单一遗传算法和传统整定方法.进一步分析了PID参数寻优是一个多目标的优化问题,探讨了可行方向.
|
| 关 键 词： | 遗传算法 模拟退火法 混合策略 多目标寻优 PID |
| 文章编号： | 1673-5196(2006)06-0085-04 |
| 收稿时间： | 2006-08-26 |
| 修稿时间： | 2006年8月26日 |
Application of genetic algorithm to optimal tuning of PID controllers |
| |
| JIANG Min,CHEN Yi-min.Application of genetic algorithm to optimal tuning of PID controllers[J].Journal of Lanzhou University of Technology,2006,32(6):85-88. |
| |
| Authors: | JIANG Min CHEN Yi-min |
| |
| Institution: | 1. College of Computer Science and Technology, Shanghai University, Shanghai 200436, China; 2. Shanghai Institute of Technology, Shanghai 200235, China |
| |
| Abstract: | Aimed at the drawback of genetic algorithm such as its tendency to prematurely converge,stronger dependence on the parameters,and tedious process with annealing-simulated searching a hybrid policy,GASA in which the GA parallel search and SA jump probability is combined,was adopted to improve the capability,efficiency,and reliability of optimization with it. In PID tuning based on hybrid strategy,SA state production function was considered as mutation with controllable probability,so that exhibited stronger mutant function at high temperature and stronger tending function at low temperature as well.Simulation result showed that GASA was remarkably better than SGA and Z-N.Further analysis showed that PID tuning was a multi-objective optimization,and its feasibility in PID tuning was discussed. |
| |
| Keywords: | genetic algorithms simulated annealing hybrid strategy multi-objective optimization PID |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
