基于多层感知器的BPNN车辆稳定性最优鲁棒控制  

BPNN Optimal Robust Control for Vehicle Stability Based on Multiple Layer Perceptron

作　　者：陈凯镔 王从明 陶沙沙 李香红[2] CHEN Kai-bin;WANG Cong-ming;TAO Sha-sha;LI Xiang-hong(Chengdu Vocational and Technical College of Industry,Sichuan Chengdu 610218,China;School of Mechanical and Power Engineering,He’nan University of Technology,He’nan Jiaozuo 454003,China)

机构地区：[1]成都工业职业技术学院,四川成都610218 [2]河南理工大学能源科学与工程学院,河南焦作454003

出　　处：《机械设计与制造》2024年第5期271-277,共7页Machinery Design & Manufacture

基　　金：2018年度河南省重点研发与推广专项(182102310719)。

摘　　要：为了增强非线性车辆模型的稳定性以及鲁棒性,提出了一种基于多层感知器的反传神经网络车辆稳定性最优鲁棒控制。采用四轮主动转向模型,建立了多层感知器前馈反向传播神经网络模型作为逼近器。采用最优鲁棒控制来调节车辆的横摆角速度和侧滑角,以满足期望的车辆响应。建立的神经网络模型通过状态变量训练来区分车辆的非线性动力学特性和相应的最优反馈增益。利用Lyapunov稳定性方法对控制器的鲁棒性与稳定性进行了分析,并采用滑模控制器跟踪期望的横摆角速度和侧滑角响应。仿真结果表明,所提出的方法能显著提高车辆的鲁棒性以及控制性能。In order to enhance the stability and robustness of nonlinear vehicle model,an optimal robust control of vehicle stabili-ty based on back propagation neural network with multiple layer perceptron was proposed.Using the four-wheel active steering model,a multiple layer perceptron feed-forward back propagation neural network model was established as the approximator.The optimal robust control was used to adjust the yaw rate and sideslip angle to meet the desired vehicle response.The neural net-work model was established to distinguish the vehicle nonlinear dynamic characteristics and the corresponding optimal feedback was gained through the state variable training.Lyapunov stability method was used to analyze the robustness and stability of the controller,and sliding mode controller was used to track the desired yaw rate and sideslip angle response.Simulation results show that the proposed method can significantly improve the vehicle robustness and control performance.

关 键 词：非线性 反传神经网络 车辆控制 多层感知器 

分 类 号：TH16[机械工程—机械制造及自动化] TH133.33[自动化与计算机技术—控制理论与控制工程] TP18[自动化与计算机技术—控制科学与工程]