基于BP神经网络PID的直驱式电液伺服系统动态特性  

Dynamic Characteristics of Direct-drive Electro-hydraulic Servo System Based on BP Neural Network PID

作　　者：杨浩[1] 张敏[1] 顾文斌[1] YANG Hao;ZHANG Min;GU Wenbin(College of Mechanical and Electrical Engineering,Hohai University,Changzhou 213022,China)

机构地区：[1]河海大学机电工程学院

出　　处：《系统仿真技术》2019年第3期188-192,共5页System Simulation Technology

基　　金：国家自然科学基金面上项目(51875171);中央高校基本科研业务费专项基金(2019B21614)

摘　　要：为了克服传统阀控液压系统易损坏性、易泄漏、对液压油污染的敏感性问题,使系统的高效能性、可靠性得到改善,采用直驱式容积控制的电液伺服系统。针对系统控制问题提出BP神经网络PID控制技术,基于PID控制建立的数学模型,BP神经网络能够实现PID的相关参数自适应与改进。结果表明,经过仿真试验的对照,与非智能PID控制比较,BP神经网络PID控制具有较好的抗干扰能力、更快的收敛速度、较小的超调量、更好的鲁棒性。采用BP神经网络PID自适应控制,为上述液压系统的最优控制提出一种控制方法。In order to overcome the traditional valve-controlled hydraulic system easy to damage,leakage problems,sensitivity to hydraulic oil pollution,improve the efficiency and reliability of the system,the use of Direct drive volume control electro-hydraulic servo system.The PID control technology of BP neural network is proposed for the control of direct-drive electro-hydraulic system,and on the basis of the traditional mathematical model of PID control,BP neural network is used to adjust the PID parameters adaptively.Results,compared with the traditional PID,the control of the simulation test,BP neural network PID control has better anti-jamming ability,faster convergence speed,smaller overshoot and better robustness.BP neural network PID adaptive control is used to propose a control method for optimal control of electro-hydraulic servo system with direct drive volume control.

关 键 词：伺服液压系统 BP神经网络 PID 自适应参数 仿真 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置]