基于LMI和扰动观测器的电动伺服系统RBF神经网络控制  

RBF Neural Network Control of Electric Servo System Based on LMI and Disturbance Observer

作　　者：李晓飞 范元勋[1] 许鹿辉 LI Xiaofei;FAN Yuanxun;XU Luhui(School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南京理工大学机械工程学院,江苏南京210094

出　　处：《机械制造与自动化》2024年第1期113-117,共5页Machine Building & Automation

基　　金：航天一院CALT基金资助项目(CALT201512)。

摘　　要：为了提高电动伺服系统的加载力跟踪精度,基于线性矩阵不等式(LMI)设计扰动观测器和控制器。针对系统中的非线性因素,采用RBF神经网络逼近系统的数学模型;在建立系统跟踪目标模型的基础上,根据LMI设计扰动观测器对控制器进行多余力的补偿,利用李雅普诺夫函数证明扰动观测器和控制器的收敛;在MATLAB/Simulink中搭建仿真模型,分析扰动观测器和RBF神经网络在不同工况下对系统相应量的精准估计,且误差均满足所设定的性能指标,同时与PID控制相比较,证明所提控制策略的控制性能更优。In order to improve the loading force tracking accuracy of the electric servo system,a system of disturbance observer and controller is designed based on linear matrix inequality(LMI).Aimed at the nonlinear factors in the system,the RBF neural network is used to approximate the mathematical model of the system.Based on the establishment of the system tracking target model,the disturbance observer is designed according to LMI to compensate the excess force of the controller,and the Lyapunov function is applied to prove the convergence of disturbance observer and controller.A simulation model is constructed in MATLAB/Simulink,and the accurate estimation of the corresponding quantities of the system by the disturbance observer and RBF neural network is conducted under different working conditions,which shows that the errors all meet the set performance indicators,and the control performance of the proposed control strategy is better compared with PID control.

关 键 词：电动伺服系统 线性矩阵不等式 扰动观测器 RBF神经网络 

分 类 号：TM351[电气工程—电机]