基于RBF神经网络的电子节气门滑模控制器  

Sliding Mode Controller for Electronic Throttle Based on RBF Neural Network

作　　者：徐子丰 童亮[1] XU Zi-feng;TONG Liang(School of Mechanical and Electrical Engineering,Beijing University of Information and Technology,Beijing 100192,China)

机构地区：[1]北京信息科技大学机电工程学院,北京100192

出　　处：《计算机仿真》2024年第5期147-151,517,共6页Computer Simulation

基　　金：北京市自然科学基金项目(3192014)。

摘　　要：为解决车辆电子节气门在复杂环境下发生的模型参数不确定和外部扰动问题,根据电子节气门的非线性特性,选择了基于RBF神经网络的滑模控制器。通过RBF神经网络对节气门的非线性部分进行逼近,并采用Lyapunov方法设计系统的自适应律。同时系统通过对扩张状态观测器的设计,达到对阀板角速度变化的准确观测。仿真结果表明,在复杂环境下,该控制器可以对不精确的节气门模型,保持较快的响应速度和对期望开度的准确跟踪,神经网络的自学习能力提高了节气门系统的鲁棒性。In order to solve the model parameter uncertainty and external disturbance of vehicle electronic throttle in complex environment,a sliding mode controller based on RBF neural network was selected according to the nonlinear characteristics of electronic throttle.RBF neural network was used to approximate the nonlinear part of the throttle,and Lyapunov method was used to design the adaptive law of the system.At the same time,through the design of the expansion state observer,the system can accurately observe the angular velocity change of the valve plate.The simulation results show that the controller can maintain fast response speed and accurate tracking of expected opening for inaccurate throttle model under complex environment.The self-learning ability of neural network improves the robustness of throttle system.

关 键 词：径向基函数神经网络 电子节气门 滑模控制 扩张状态观测器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]