BP神经网络PID控制在电液比例加载系统中的应用  

Application of BP Neural Network PID Control in Electro-hydraulic Proportional Loading System

作　　者：李万莉[1] 李瑞[1] 

机构地区：[1]同济大学机械能源与工程学院,上海201804

出　　处：《机电一体化》2017年第7期35-39,46,共6页Mechatronics

摘　　要：实际静力加载测试中,测试试件的尺寸、刚度、载荷等存在较大的差异,加载油缸位移的扰动、流体的非线性、摩擦等都会对测试结果产生影响,这就要求控制系统应具有一定的自适应性。针对经典PID控制参数不能在线调整的缺陷,研究了一种基于BP神经网络的PID控制算法,利用BP神经网络具有的任意非线性表达能力,通过对系统性能的学习,实现具有最佳组合的PID控制。通过仿真,验证了所设计的BP神经网络PID控制系统比PID控制系统建模时间短,系统更稳定,超调量更小。In the actual static load test,there is a big difference in the size,stiffness,load of specimen.Loading cylinder displacement disturbance,fluid nonlinearity and friction will affect the results of the test,which requires the control system should have a certain degree of adaptability.According to the defects that the parameters of classical PID control can not be adjusted on line,a PID control based on BP neural network is presented in this paper.The best combination of PID control is realized by the arbitrarily nonlinear expressing ability of BP neural network and the study of system property.The simulation results show that the scheme is morestability,of smaller overshoots and shorter modeling time than PID control system.

关 键 词：BP神经网络 PID控制 电液比例溢流阀 静力加载 

分 类 号：TH137[机械工程—机械制造及自动化] TP183[自动化与计算机技术—控制理论与控制工程]