摘要
就线性定常/时变系统以及非线性系统,依据特征模型理论,给出动态系统的一阶特征模型.其特征参数随时间变化,即以一阶时变差分方程描述受控系统的动态特性;与二阶和三阶特征模型相比较,一阶模型具更少参数.为解决由一阶特征模型描述的系统的控制问题,提出基于遗忘因子迭代学习辨识的自适应迭代学习控制方法.迭代学习辨识适于时变参数的估计,它允许被估计参数随时间快速变化,抑或突变.以直线伺服系统的位置跟踪控制为例,给出一种基于特征模型与LQ最优控制策略的自适应迭代学习控制方案.仿真与实验结果表明,提出的控制方案能够有效实现受控系统的位置跟踪控制.
This paper presents the first-order time-varying difference equation, the characteristic model. Only two parameters are involved in the equation. The adaptive iterative learning control method is suggested to deal with the characteristic-model based control problem, where the learning identification Mgorithms with a forgetting factor are introduced to estimate the time-varying unknowns. For PMLSM servo-systems, in this paper, an adaptive iterative learning control scheme is proposed based on the LQ optimal control strategy. Numerarid expereimeat results are presented to demonstrate effectiveness of the control scheme.
出处
《系统科学与数学》
CSCD
北大核心
2012年第6期666-682,共17页
Journal of Systems Science and Mathematical Sciences
基金
国家自然科学基金(60874041
61174034)资助
关键词
特征模型
遗忘因子
迭代学习辨识
自适应迭代学习控制
直线伺服系统.
Characteristic models, forgetting factor, iterative learning identification,adaptive iterative learning control, linear servo systems.
