基于学习自适应估计环的迭代学习控制  

Iterative Learning Control Scheme with Learning Adaptive Estimate Loop

作　　者：池荣虎[1] 侯忠生[2] 

机构地区：[1]北京交通大学交通运输学院,北京100044 [2]北京交通大学电子与信息工程学院,北京100044

出　　处：《仪器仪表学报》2005年第z1期800-802,共3页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(60474038)资助项目

摘　　要：针对一类具有强非线性和不确定性的离散时间系统,文章给出了一种基于学习自适应估计环的迭代学习控制方法。在迭代学习控制器的基础上设计了一个学习自适应估计环,用来镇定系统,给出迭代学习控制初始的控制输入值,同时根据估计出的系统参数来确定迭代学习增益的取值范围。文章基于状态空间描述,分析了迭代学习控制系统的收敛性。仿真研究表明,该控制器能够实现完全跟踪,减少系统的初始输出误差,并加快了收敛速度。An iterative learning controller is designed for a class of nonlinear and non\|affine systems with high nonlinearity and uncertainty. A learning adaptive estimate loop, designed to stabilize the system, is added on the iterative learning controller. According to the estimated parameters, the learning gains can be limited to an exactly range. State space formulation is developed to analyze the convergence. It is shown that the convergence can be guaranteed as soon as the learning gains lie in the estimated range. The simulation study demonstrates that the controller is effective to realize the perfect tracking, to reduce the initial output error and to enhance the convergence rate.

关 键 词：学习自适应估计 迭代学习控制 非线性离散系统 状态空间 

分 类 号：TH7-55[机械工程—仪器科学与技术]