含非严格重复扰动的抗扰迭代学习控制

Anti-disturbance iterative learning control for nonlinear systems with time-iteration-varying disturbances

中文关键词  迭代学习控制  滑模控制  参数辨识  收敛性  鲁棒性  抗扰性

英文关键词  iterative learning control  sliding mode control  parameter identification  convergence  robustness  anti- disturbance

中文摘要

      针对一类含不确定参数及未知扰动的高阶非线性系统, 采用类Lyapunov方法, 结合部分限幅学习律和滑模 控制的优点, 提出一种新的滑模鲁棒迭代学习控制算法. 根据系统中不确定量的特性, 将系统中的不确定性划分为 两类: 仅沿时间轴变化的不确定性和仅沿迭代轴变化的不确定性. 前者采用迭代辨识方法处理, 后者采用迭代滑模 技术解决. 在整个作业区间上, 随着迭代次数的增加, 控制算法确保系统的跟踪误差收敛到一个界内, 控制器信号 无抖颤, 且闭环系统中其余变量一致有界. 当系统扰动仅沿时间轴变化时, 系统跟踪误差及其各阶导数沿迭代轴渐 近收敛到0, 实现系统各个状态的精确跟踪. 相比利用连续函数近似法的传统滑模控制, 该算法对未知扰动具有更 好的鲁棒性. 理论证明和仿真结果都说明了该算法的有效性.

英文摘要

      For a class of higher-order nonlinear systems with parametric uncertainties and unknown disturbances, we propose a novel sliding-mode robust iterative learning control algorithm base on the Lyapunov-like method, which suc- cessfully combines the advantages of partially-saturated learning mechanism and sliding mode technique. Uncertainties within the system are classified into two categories, the only time-varying uncertainty and the only iteration-varying uncer- tainty. The former is treated by using the iterative identification technique, while the latter is dealt with by employing an iterative sliding mode law. In the entire time interval, it is guaranteed that, along the iteration axis, the designed chattering- free controller ensures that the tracking error converges to a given bound, while all the remaining signals are bounded. In addition, tracking errors and their derivatives converge asymptotically to zero along the iterative axis in the case of only time-dependent perturbation, which implies accurate tracking for the system states. Compared with the saturation- approximation-based conventional sliding-mode mechanism, the proposed novel control technique presents better robust- ness against unknown disturbances. Theoretical analysis and simulation results show the effectiveness of the proposed algorithm.