抑制初态误差影响的自适应迭代学习控制  

Adaptive Iterative Learning Control for Inhibition Effect of Initial State Random Error

作　　者：吕庆[1] 

机构地区：[1]天津师范大学计壹机与信息工程学院,天津300387

出　　处：《自动化学报》2015年第7期1365-1372,共8页Acta Automatica Sinica

摘　　要：针对一类参数化高阶不确定非线性连续系统,设计迭代学习控制算法,以解决随机初态对系统跟踪性能产生负面影响的问题.结合滑模控制思想以及部分限幅参数学习律,控制算法在预设时间段内抑制随机初态偏差对系统跟踪性能的影响.经过预设时间后,随着迭代次数的增加,系统的跟踪误差及其各阶导数一致收敛到零.且在整个运行时间段内,系统各个变量一致有界.此外,本文回避了非参数化不确定非线性系统在放宽迭代初值假设时常使用的Lipschitz假设条件,而采用类Lyapunov函数分析法设计迭代学习控制器.理论证明和仿真结果都说明了该算法的有效性.For a class of high order nonlinear continuous sys- tems with parametric uncertainties, a novel iterative learning control algorithm is presented to improve the tracking perfor- mance when arbitrary initial states are taken into account. In- hibition of poor tracking performance caused by random initial state errors depends on sliding mode technique and partially- saturated parameter learning mechanism in a pre-specified time interval. As a result, tracking errors and their derivatives will converge asymptotically to zero with the increase of iterations after the above-mentioned interval. In addition, all variables are uniformly bounded in the whole operation period. Moreover, the iterative learning controller is designed in this paper based on Lyapunov-like synthesis, instead of the Lipschitz condition used in nonlinear systems with nonparametric uncertainties. Both theoretical proof and simulation results show the effectiveness of the proposed algorithm.

关 键 词：迭代学习控制 初始状态误差 参数化系统 滑模控制 自适应控制 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]