高相对阶连续时间系统的间接迭代学习控制  

Indirect-type iterative learning control for high relative degree continuous-time systems

作　　者：刘娇龙 李瑾[2] 王龙 薛建平[3] 刘彬[1] LIU Jiao-long;LI Jin;WANG Long;XUE Jian-ping;LIU Bin(The Unit 66133 of PLA,Beijing 100043,China;Equipment Management and Unmanned Aerial Vehicle Engineering College,Air Force Engineering University,Xi’an Shaanxi 710051,China;Aeronautics Engineering College,Air Force Engineering University,Xi’an Shaanxi 710038,China)

机构地区：[1]中国人民解放军66133部队,北京100043 [2]空军工程大学装备管理与无人机工程学院,陕西西安710051 [3]空军工程大学航空工程学院,陕西西安710038

出　　处：《控制理论与应用》2020年第5期1127-1134,共8页Control Theory & Applications

摘　　要：本文提出了一类高相对阶线性连续时间系统的间接迭代学习控制算法,该算法相对独立于系统局部控制器,因此可以应用于已有局部反馈控制器的系统.采用具有极点配置的H∞鲁棒控制器作为系统的内环控制,而在外环通过迭代学习控制调整内环系统的指令信号.通过引入拉氏变化,构建了迭代学习系统的2-D Roesser模型,推导了系统渐近收敛条件,并研究了存在有界初始条件偏移和迭代变化外部干扰时算法的鲁棒性能.最后,利用空中加油对接控制的算例进一步验证了算法的有效性.In this paper, a kind of indirect-type iterative learning control is developed for high relative degree linear continuous-time systems. The proposed indirect-type iterative learning control design is relatively independent of the system local controller, thus it can be applied to the process which already has local feedback control. The H-infinity robust controller with pole assignment is designed in the inner loop, while the iterative learning control updating law is implemented to adjust the setpoint command of the closed-loop system in the outer loop. By introducing the Laplace transform, the two dimensional Roesser system is established for the iterative learning system, and then the system monotonic convergence condition is derived. Furthermore, the robustness performance of the iterative learning control algorithm is investigated in the presence of bounded initial state shift and iteration-varying external disturbance. Finally, the effectiveness of the proposed algorithm is illustrated by the simulation of aerial refueling docking control.

关 键 词：间接迭代学习控制 相对阶系统 鲁棒控制 2–D系统 迭代变化因素 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]