首页|汽车EPS伺服变幅加载系统PID建立DNN控制优化

汽车EPS伺服变幅加载系统PID建立DNN控制优化

扫码查看
为了进一步提高汽车电动助力转向电液伺服加载系统运动精度,设计了一种PID建立DNN控制系统.通过柔性联接的模式构建电液位置伺服系统,确保在瞬态平衡响应过程中快速完成控制.搭建了柔性电液位置伺服调控测试,并开展调控效果分析.研究结果表明:通过PID建立的DNN控制器在校正柔性联接电液位置伺服参数方面发挥了关键作用,整体响应时间大幅缩短,确保系统频段动态控制效果获得明显改善.采用DNN控制器调节柔性连接电液伺服系统时,能够有效增强系统低频负载位移精度,避免发生幅值衰减的问题,能够以更高精度跟踪柔性连接电液位置伺服参数,促进了系统跟踪性能获得明显改善.该研究对提高汽车运行稳定性以及节能具有很好的理论支撑作用.
DNN Control Optimization of EPS Servo Variable Amplitude Loading System PID Establishment
In order to further improve the motion accuracy of electro-hydraulic servo loading system of electric power steering,a PID control system was designed to establish DNN control system.The electro-hydraulic position servo system is constructed by the mode of flexible connection to ensure the rapid dynamic control in the transient equilibrium response.The flexible electro-hy-draulic position servo control test was built and the control effect was analyzed.The research results show that the DNN controller established by PID plays a key role in the correction of electro-hydraulic position servo parameters of flexible coupling,and the overall response time is greatly shortened,ensuring that the dynamic control effect of the system frequency band is significantly improved.When DNN controller is used to adjust the electro-hydraulic servo system of flexible connection,the low-frequency load displacement accuracy of the system can be effectively enhanced,the problem of amplitude attenuation can be avoided,and the electro-hydraulic position servo parameters of flexible connection can be tracked with higher accuracy,which promotes the obvi-ous improvement of the tracking performance of the system.This research has a good theoretical support for improving the stability of vehicle operation and energy saving.

Electro-Hydraulic Servo SystemDisplacement ControlCerebellar Model Neural NetworkTracking Performance

王培、张杰飞、张昊

展开 >

河南交通职业技术学院汽车学院,河南 郑州 450015

东南大学仪器科学与工程学院,江苏 南京 210096

电液伺服系统 位移控制 深度神经网络 跟踪性能

河南省科技发展计划项目

192102210262

2024

机械设计与制造
辽宁省机械研究院

机械设计与制造

CSTPCD北大核心
影响因子:0.511
ISSN:1001-3997
年,卷(期):2024.(7)
  • 13