采用前馈神经网络控制的变速泵压力控制仿真研究

Simulation of Variable Speed Pump Pressure Control Based on Feedforward Neural Network Control

针对变速泵压力控制系统存在响应速度慢、输出误差大等问题,创建了液压缸变速泵压力驱动模型简图,推导了压力变化方程式。使用3层BP神经网络算法,设计出前馈神经网络控制方法。利用3层BP神经网络的逼近特性,使控制系统快速收敛到稳定状态。通过仿真验证变速泵压力控制输出效果,结果表明,采用前馈神经网络控制方法,系统达到稳定状态时,响应时间为0.2 s,超调量为5%,稳态误差较小,控制输出效果明显优于PID。

To solve the problems of slow response speed and large output error of variable speed pump pressure control system,a feedforward neural network control method is proposed.The pressure driving model diagram of hydraulic cylinder variable speed pump is created and the pressure variation equation is deduced.The feedforward neural network control method is designed with the help of three-layer BP neural network algorithm.The control system converges rapidly to a stable state through the approximation characteristic of three-layer BP neural network.The pressure control output effect of variable speed pump is verified through simulation and the results show that by using the feedforward neural network control method,the response time is 0.2 s when the system reaches the stable state,the overshoot is 5%and the steady-state error is small,which is obviously better than the PID control output effect.