在本文中,我们考虑了为连续时间非线性系统开发控制器的问题,其中控制该系统的方程式未知。 利用这些测量结果,提出了两个新的在线方案,这些方案通过两个基于自适应动态编程(ADP)的新实现方案来合成控制器,而无需为系统构建或假设系统模型。 为了避免对系统的先验知识的需求,引入了预补偿器以构造增强系统。 通过自适应动态规划求解相应的Hamilton-Jacobi-Bellman(HJB)方程,该方程由最小二乘技术,神经网络逼近器和策略迭代(PI)算法组成。 我们方法的主要思想是通过最小二乘技术对状态,状态导数和输入信息进行采样以更新神经网络的权重。 更新过程是在PI框架中实现的。 本文提出了两种新的实现方案。 最后,给出了几个例子来说明我们的方案的有效性。 (C)2014 ISA。 由Elsevier Ltd.出版。保留所有权利。
2023-03-21 17:45:57 901KB Model-free controller; Optimal control;
1
by Hou, Zhongsheng Jin, Shangtai Introduction ...........................................................................................1 1.1 Model-Based Control ........................................................................1 1.1.1 Modeling and Identification .................................................1 1.1.2 Model-Based Controller Design ...........................................3 1.2 Data-Driven Control .........................................................................5 1.2.1 Definition and Motivation of Data-Driven Control .............6 1.2.2 Object of Data-Driven Control Methods..............................7 1.2.3 Necessity of Data-Driven Control Theory and Methods ........................................................................8 1.2.4 Brief Survey on Data-Driven Control Methods..................10 1.2.5 Summary of Data-Driven Control Methods.......................15 1.3 Preview of the Book.........................................................................16 2 Recursive Parameter Estimation for Discrete-Time Systems................19 2.1 Introduction ....................................................................................19 2.2 Parameter Estimation Algorithm for Linearly Parameterized Systems.....................................................................20 2.2.1 Projection Algorithm..........................................................21 2.2.2 Least-Squares Algorithm ....................................................22 2.3 Parameter Estimation Algorithm for Nonlinearly Parameterized Systems.....................................................................27 2.3.1 Projection Algorithm and Its Modified Form for Nonlinearly Parameterized Systems...............................27vi  ◾  Contents 2.3.2 Least-Squares Algorithm and Its Modified Form for Nonlinearly Parameterized Systems...............................32 2.4 Conclusions.......................................................................
2021-05-03 20:10:52 5.49MB Adaptive Control 控制理论
1
Model Free Adaptive Control with Disturbance Observer
2021-03-20 21:18:14 133KB 研究论文
1