具干扰抑制的线性鲁棒迭代学习控制  

Disturbance rejection issue in linear robust iterative learning control

作　　者：崔彩莲[1] 孙明轩[1] 徐建明[1] 

机构地区：[1]浙江工业大学信息工程学院,浙江杭州310032

出　　处：《浙江工业大学学报》2006年第3期237-241,共5页Journal of Zhejiang University of Technology

基　　金：国家自然科学基金资助项目(60474005)

摘　　要：针对不确定线性定常系统,考虑频域迭代学习控制器设计及干扰抑制问题.根据提出的闭环迭代学习控制律,推导出跟踪误差收敛的充分条件,并证明了它等价于系统没有采用迭代学习控制时的鲁棒性能条件,但在性能上却得到大大的改善.在完全跟踪不可能实现的情况下,讨论性能权重函数的选取,保证了跟踪误差收敛于实际工程容许的范围内,且参数选择非常简单.基于此性能权重函数,结合鲁棒控制理论,求解满足收敛条件的控制器.给出的仿真结果表明该设计方法的有效性.The issues of iterative learning controller design and disturbance rejection in ILC are addressed in this paper. A sufficient condition for convergence designed according to the proposed closed ILC algorithm is derived in frequency domain for a class of uncertain linear time-invariant （LTI） systems. Furthermore, it is proven to be equal to robust performance condition of the feedback system without ILC, but the performance is much better. When it is impossible to eliminate tracking error completely, performance weighting function is selected to guarantee small tracking error which is allowable in project. This choice method of parameter is very simple. Based on it, the controller satisfying the convergence condition is solved combining with robust control theory. Simulation results are presented to demonstrate the effectiveness of the proposed method.

关 键 词：干扰抑制 性能权重函数 迭代学习控制 鲁棒性能 不确定线性定常系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]