具有输入未建模动态和量化输入的纯反馈非线性系统自适应控制  

The adaptive control of pure-feedback nonlinear systems with input unmodeled dynamics and quantized inputs

作　　者：施孟松 毛骏 胡汇源 SHI Mengsong;MAO Jun;HU Huiyuan(College of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou 310018,China)

机构地区：[1]中国计量大学机电工程学院,浙江杭州310018

出　　处：《中国计量大学学报》2023年第4期563-572,共10页Journal of China University of Metrology

基　　金：国家自然科学基金青年基金项目(No.62103392)。

摘　　要：目的:本文研究一类具有输入未建模动态和量化输入的纯反馈非线性系统跟踪控制问题。方法:在控制器设计中,引入M-滤波器来处理输入未建模动态,使用滞后量化器来处理量化输入,利用径向基函数神经网络(radial basis function neural networks,RBFNN)逼近未知非线性动态,选取障碍李雅普诺夫函数(barrier lyapunov function,BLF)处理输出约束,并使用Backstepping方法导出自适应控制律。结果:稳定性分析证明所得到的闭环系统全部信号均满足半全局一致终结有界(semi-globally uniformly utimately bounded,SGUUB),并且跟踪误差可以收敛到原点的一个小邻域,同时这种跟踪误差也满足输出约束。在仿真中,对于一个二阶非线性系统,可以看到当t≈0.17 s时,跟踪误差到达原点的小邻域。结论:本文提出的控制方案证明了所得到的闭环系统的所有信号都满足SGUUB。Aims:This paper investigates the tracking control problem for a class of pure-feedback nonlinear systems with input unmodeled dynamics and quantized inputs.Methods:In the controller design,an M-filter was introduced to handle the input unmodeled dynamics.A hysteresis quantizer was used to deal with the quantized inputs.The existed uncertain nonlinearities was approximated by RBFNN.BLF was selected to handle the output constraints;and an adaptive control law was exported by using the backstepping method.Results:The stability analysis demonstrated that the whole signals of the resulting closed-loop system satisfied SGUUB.In addition,the tracking error could converge to a small neighborhood of the origin.Meanwhile,such tracking error also satisfied the output constraints.In simulation,for a second-order nonlinear system,the tracking error could reach the small neighborhood of the origin when t≈0.17 s.Conclusions:The control scheme proposed in this paper demonstrates that all signals of the resulting closed-loop system are SGUUB.

关 键 词：输入未建模动态 量化输入 障碍李雅普诺夫函数 BACKSTEPPING方法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]