摘要
为解决电液比例伺服系统无法精确定位的问题,推导了系统的流量动态平衡方程,并设计了一种基于死区直接补偿的BP(DZDC-BP)神经网络控制算法,即在线性化比例阀输入信号与输出流量关系的基础上,利用BP神经网络控制算法逼近非线性系统的特性,弥补系统中的非线性、未知参数、外部干扰和建模误差等问题,使得系统能够时刻跟随期望轨迹。使用MATLAB-AMESim软件对系统进行联合仿真,结果表明:无论系统有无外部干扰,DZDC-BP神经网络控制与有死区比例阀PID控制(PPID)和无死区比例阀PID控制(SPID)相比,系统的稳态误差都有显著的减少。说明DZDC-BP神经网络控制算法在一定程度上解决了系统中的死区和非线性等问题,同时具有良好的抗干扰能力,显著提高了系统的控制性能。
In order to solve the problem that the electro-hydraulic proportional servo system can't accurately realize the positioning.The system's flow dynamic balance equation is derived.The dead-zone direct compensation BP(DZDC-BP)neural network control algorithm is designed.On the basis of linearization relationship between the input signal and output flow of the proportional valve,the BP neural network control algorithm is used to approximated the nonlinear system's characteristics,and makes up for the nonlinear,unknown parameters,external interference and modeling errors in the system.This enables the system to track the desired trajectory at any time.Finally,the joint simulation of MATLAB and AMESim is used to simulate the system.The results show that,whether or not the system has external interference,the steady-state error of the DZDC-BP neural network control is significantly reduced compared with the PID control of the proportional valve with and without dead-zone(PPID and SPID).It shows that the DZDC-BP neural network control algorithm can solve the system's problem,and has good anti-jamming capability,and improves the performance of the system.
作者
刘霞勇
张潜
刘正浩
林贵华
赵建
LIU Xia-yong;ZHANG Qian;LIU Zheng-hao;LIN Gui-hua;ZHAO Jian(Marine Design & Research Institute of China, Shanghai 200011)
出处
《液压与气动》
北大核心
2022年第4期165-172,共8页
Chinese Hydraulics & Pneumatics
关键词
电液比例伺服系统
死区直接补偿
BP神经网络
联合仿真
electro-hydraulic proportional servo system
dead-zone direct compensation
BP neural network
joint simulation
