基于扩展状态观测器的量化无模型自适应迭代学习控制

Extended state observer-based quantitative model-free adaptive iterative learning control

中文关键词  迭代学习控制  数据量化  扩展状态观测器  多非重复不确定性  非线性非仿射系统

英文关键词  iterative learning control  data quantification  extended state observer  multiple non-repeated uncertainties  nonlinear nonaffine systems

中文摘要

      针对非线性非仿射离散时间系统, 本文提出了基于扩展状态观测器的量化无模型自适应迭代学习控制策略. 通过引入迭代动态线性化方法, 处理系统非线性和非仿射结构不确定性, 提出了基于偏格式的迭代线性数据模型(iLDM). 给出误差量化描述, 设计了基于量化数据的学习控制律和参数迭代自适应律, 其中后者不仅可以估计iLDM的不确定参数, 而且能够调节控制律的学习增益, 增强了控制方案的鲁棒能力. 同时, 设计迭代域中的扩展状态观测器, 对参数估计、未建模动态和外界扰动等多非重复不确定性进行估计和补偿. 理论分析和仿真研究均证明了所提出方法的有效性.

英文摘要

      A quantized model-free adaptive iterative learning control based on extended state observer is proposed for nonlinear non-affine discrete-time systems. An iterative dynamic linearization method is introduced to deal with nonlinear and non-affine structural uncertainties, and an iterative linear data model (iLDM) based on partial form is proposed. The error quantization description is given, and the learning control law based on quantized data and the parameter iteration adaptive law are designed. The latter can not only estimate the uncertain parameters of iLDM, but also adjust the learning gain of the control law, which enhances the robustness of the control scheme. At the same time, an extended state observer in the iterative domain is designed to estimate and compensate multiple non-repeated uncertainties such as parameter estimation, unmodeled dynamics and external disturbances. Both mathematical analysis and simulation study demonstrate the effectiveness of the proposed method.