半主动空气悬架BP-PID控制器设计及其随机路面验证  

Design of BP-PID Controller for Semi-active Air Suspension and Its Random Pavement Verification

作　　者：班小强 覃桂全 BAN Xiaoqiang;QIN Guiquan(School of Intelligent Manufacturing,Guangdong Southern Vocational College,Jiangmen 529000,China;Maibo Automation Machinery Co.,Ltd.,Jiangmen 529000,China)

机构地区：[1]广东南方职业学院智能制造学院,广东江门529000 [2]迈博自动化机械有限公司,广东江门529000

出　　处：《机械制造与自动化》2022年第3期217-219,235,共4页Machine Building & Automation

基　　金：2020年广东省普通高校特色创新项目(2020KTSCX388)。

摘　　要：在综合考虑算法精度与效率的条件下,设计一种新的汽车半主动悬架BP-PID控制模型,在该算法中融合神经网络控制器的鲁棒特性。通过神经网络完成PID参数的在线整定功能,并达到对非线性函数进行快速逼近与自主学习的过程,充分克服非线性悬架系统的表述偏差缺陷。利用高斯白噪声以及成形滤波器来随机调控路面轮廓的不平度,在Matlab/Simulink平台开展路面模拟信号验证分析。测试结果表明:BP-PID控制的汽车悬架可以显著减小车身加速度,获得更舒适的乘坐性能。In comprehensive consideration of algorithm accuracy and efficiency,a new BP-PID control model of semi-active suspension was designed,into which,the robustness of the neural network controller was integrated.The online tuning function of PID parameters is completed by neural network to achieve the process of fast approximation and self-learning of nonlinear functions,which fully overcomes the defect of expression deviation of nonlinear suspension system.Gaussian white noise and shaping filter were used to regulate the roughness of road profile in random and the road simulation signal was verified and analyzed on Matlab/Simulink platform.The test results show that the vehicle suspension controlled by BP-PID can significantly reduce the body acceleration and obtain more comfortable ride performance.

关 键 词：汽车悬架 神经网络 PID控制器 仿真 随机路面 

分 类 号：TH122[机械工程—机械设计及理论]