基于LQR指标的鲁棒PID整定方法及鲁棒稳定性裕度评估  

Robust PID parameters design method based on LQR criterion and robustness analysis

作　　者：马龙华[1] 叶正波[2] 郑泳凌[1] 钱积新[1] 

机构地区：[1]浙江大学系统工程研究所,浙江杭州310027 [2]浙江大学宁波理工学院管理工程系,浙江宁波315104

出　　处：《浙江大学学报（工学版）》2002年第5期486-489,共4页Journal of Zhejiang University：Engineering Science

摘　　要：根据 L QR指标 ,结合最小最大原理 ,使用改进遗传算法作为优化搜索算法 ,提出一种新的鲁棒 PID参数整定方法 .该方法有三个优点 :首先能通过选取衰减系数和自然频率来获取期望的闭环特性 ,并获取近似的 L QR权矩阵 ;其次为 PID控制器增加了一定程度的预测性能 ,最后用本方法设计的控制器对模型误差和外部干扰都表现出一定的鲁棒性 .仿真实例证实了该方法的有效性 .A new method for tuning Robust PID parameters, was developed based on the LQR criterion. Combining the Mini max theory and applying a meliorated genetic algorithm in optimal search, this method has three benefits. Firstly, it can help to achieve the expected closed loop performance characteristics and create a proximate LQR weight matrix, via choosing proper damping ratio and natural frequency. Secondly, it improves to certain extent the predictability of a PID controller. Finally, controllers utilizing this method have shown good robustness in its performance against introduced model errors and outside disturbances.The validity of this method is verified with simulation examples.

关 键 词：鲁棒稳定性 裕度评估 LQR指标 最小最大准则 遗传算法 鲁棒PID参数 整定 优化控制理论 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]