BP神经网络PID自适应控制悬架平顺性研究  

BP Neural Network PID Adaptive Control of Suspension for Ride Comfort

作　　者：刘省标 李美[1] 张少波 王德志 LIU Sheng-biao;LI Mei;ZHANG Shao-bo;WANG De-zhi(College of Mechanical and Electrical Engineering,Hainan University,Hainan Haikou 570228,China)

机构地区：[1]海南大学机电学院,海南海口570228

出　　处：《机械设计与制造》2022年第12期11-14,共4页Machinery Design & Manufacture

基　　金：海南省高等学校科学研究项目(Hnky2019ZD-2);海南省自然科学基金资助项目(2019RC144,518QN209)。

摘　　要：建立了1/4半主动悬架二自由度动力学模型并进行动力学分析。引用增量式PID控制器,采用三层前馈网络结构,设计BP神经网络PID自适应控制器。以白噪声模拟B级路面扰动激励,结合半主动悬架动力学模型与控制器在Matlab/Simulink中搭建悬架仿真模型,对悬架系统平顺性进行振动控制的仿真研究。结果表明:相比于PID半主动悬架,BP神经网络PID自适应控制半主动悬架具有更强的鲁棒性和自适应能力,能更有效的抑制路面扰动,车辆平顺性显著提高。A2-DOF dynamic model of 1/4 semi-active suspension is established and analyzed. The BP neural network PID adaptive controller is designed by using an incremental PID controller and a three-layer feedforward network. The white noise is used to simulate the B-level road disturbance excitation,and the semi-active suspension dynamic model and controler are used to build a suspension simulation model in Matlab/Simulink. The vibration control of the suspension system is simulated. The results show that compared with PID semi-active suspension,BP neural network PID self-adaptive control semi-active suspension has stronger robustness and self-adaptive ability,can more effectively suppress the road disturbance,vehicle ride comfort is significantly improved.

关 键 词：半主动悬架 PID控制策略 BP神经网络控制策略 MATLAB/SIMULINK 

分 类 号：TH16[机械工程—机械制造及自动化]