何文清　麦云飞

摘 要：基于汽车疲劳试验机平台对转向机液压伺服系统进行研究，针对试验机运行过程中出现的控制精度不高、系统不稳定问题，设计了电液位置伺服系统。首先建立系统数学模型，推导相应的传递函数。接着利用Simulink软件对系统进行稳定分析，利用典型的PID参数整定策略对系统进行优化处理。最后采用神经元网络自学习算法，在simulink中设计单神经元PID自学习系统，对位置伺服系统进行参数优化整定。研究结果表明：采用单神经PID整定的电液位置伺服系统超调量比典型PID整定策略减少了90%以上，整定时间也缩短了一半以上。系统稳定性更强、精度更高，具有更好的鲁棒性。

关键词：电液位置伺服系统;Simulink软件;单神经元PID整定

DOI：10. 11907/rjdk. 191622

中图分类号：TP319   文献标识码：A                文章编号：1672-7800（2020）003-0177-05

Electro-hydraulic Position Servo System Based on Neural Network PID Tuning

HE Wen-qing，MAI Yun-fei

（School of Mechanical Engineering， University of Shanghai for Science and Technology， Shanghai 200093，China）

Abstract： This paper mainly studies the hydraulic servo system of steering gear based on the automobile fatigue testing machine of an enterprise as a platform. During the operation of the test machine， the control accuracy is not high and the system is unstable. According to the design requirements， the electro-hydraulic position servo system is designed at first. Then the stability of the system is analyzed by Simulink software， and the system is optimized by typical PID parameter tuning strategy. Finally， a single neuron PID self-learning system is designed in Simulink by using the self-learning algorithm of neuron network to optimize the parameters of the position servo system. The results show that the overshoot of electro-hydraulic position servo system tuned by single nerve PID is reduced by more than 90% and the setting time is shortened by more than half compared with the typical PID tuning strategy. The system has stronger stability， higher accuracy and stronger robustness.

Key Words： electro-hydraulic position servo system; Simulink software; single neuron PID tuning

0 引言

電动助力转向机电一体化试验台[1-2]是用于模拟汽车转向时所受地面摩擦等负载的模拟加载平台，可对汽车电动助力转向设备进行性能分析。电液位置伺服系统[3-4]是试验台控制器的重要组成部分，具有输出功率大、响应速度快、精度高等特点。由于液压伺服系统在运行过程中会出现物理参数（温度、尺寸、压力等）改变，或者受到外界干扰，导致无法达到理想的控制要求，因此学者们使用不同方法优化电液位置伺服系统。李辉[5]指出在系统中使用PID控制器可以达到较好的效果;孔令刚[6]把复合PID控制器应用到电液伺服系统中，同样取得了不错的效果;裴雨霞[7]利用模糊控制理……