鲁棒迭代学习控制及在高精密平台中的应用  

Robust iterative learning control and its application to high precision test bench

作　　者：姜晓明[1] 王岩[1] 王程 陈兴林[1] 

机构地区：[1]哈尔滨工业大学控制科学与工程系,黑龙江哈尔滨150001 [2]上海机电工程研究所,上海200233

出　　处：《系统工程与电子技术》2013年第3期601-608,共8页Systems Engineering and Electronics

基　　金：国家科技重大专项(2009ZX02207)资助课题

摘　　要：针对不确定性系统,提出一种非因果的鲁棒迭代学习控制方法。采用反馈控制和迭代学习控制分别设计并统一进行鲁棒性分析的方法,以提高系统的反馈性能和学习性能。首先采用H∞控制方法,保证反馈部分的鲁棒稳定性;其次采用二次型优化迭代学习控制律来提高系统的学习效率;最后结合μ分析理论,对迭代学习控制系统进行鲁棒性验证,并对优化学习律的参数进行修正。该方法吸取了非因果迭代学习律学习性能好的优点,也充分利用μ分析保守性小的特点来提高系统的性能。算例结果验证了方法的有效性。A non-causal robust iterative learning control method for uncertain systems is presented. Feedback controller and iterative learning controller are designed independently, and robust analysis is then implemented together in order to ensure the feedback and learning performance of the system. Firstly, an H∞ controller is designed to guarantee the robust stability of the feedback part. Secondly, learning efficiency is improved by using a quadratic optimized iterative learning control law. Finally, the robustness of the iterative learning control system is verified by μ-analysis, and the design parameters of the optimal iterative learning control law are modified then. The proposed method not only takes the advantage of good learning performance of the non-causal iterative learning control law, but also utilizes the property of small conservation of μ-analysis to improve the performance. An example is given to testify the effectiveness of the proposed method.

关 键 词：迭代学习控制 二次型优化 鲁棒性 Μ分析 精密运动控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]