欠阻尼对象的最优 PID 控制器参数整定方法  

Optimal Tuning of PID Controllers for Under-Damped Systems

作　　者：薛亚丽[1] 李东海[1] 吕崇德[1] 

机构地区：[1]清华大学热能动力仿真与控制研究所,北京100084

出　　处：《系统仿真学报》2004年第10期2347-2349,2366,共4页Journal of System Simulation

基　　金：国家九五科技攻关计划(2001BA401A03-07)

摘　　要：研究了带有纯延迟的欠阻尼对象最优PID控制器的参数整定及性能鲁棒性评价问题。首先依据频域判据,确定保证闭环系统稳定的PID控制器参数域,然后采用全局寻优的遗传算法,获得ITAE指标最优的PID控制器参数。采用蒙特卡罗随机实验方法,对模型参数在有界范围内摄动时的性能鲁棒性进行了定量的比较和评价。仿真结果表明,本文方法具有良好的动态性能,在对象参数摄动时与极点配置方法和幅值相位裕量方法相比具有较好的性能鲁棒性。This paper studies optimal tuning of PID controllers and its robust performance evaluation for under-damped plants with time delay. Based on frequency stability theory, the parameter regions of PID controller were acquired which could stabilize the close-loop system. Then the ITAE-optimum parameters were obtained through genetic algorithm with global searching. Monte-Carlo stochastic method was adopted to quantitatively evaluate the performance robustness of system when parameter uncertainty occurred in a bounded range. Simulation results show that the proposed method bears satisfying dynamic quality and better performance robustness compared with Pole Assignment method and Gain Phase Margin method.

关 键 词：最优PID控制器 欠阻尼对象 蒙特卡罗方法 性能鲁棒性 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]