基于BP神经网络PID的注塑机液压控制  

Hydraulic Control of Injection Moulding Machine Based on BP Neural Network PID

作　　者：黄晓萍 候俊 HUANG Xiao-ping;HOU Jun(Nanjing Institute of Mechatronic Technology,Nanjing 211135,China)

机构地区：[1]南京机电职业技术学院

出　　处：《塑料工业》2019年第7期64-66,共3页China Plastics Industry

摘　　要：注塑机液压系统是一个非线性、大时滞性和强耦合的复杂系统,传统比例积分微分(PID)控制由于参数固定不变,导致超调量大、稳定性差、控制精度低,对注塑机液压控制效果不理想,现提出了一种改进型误差反向传播法(BP)神经网络PID控制方法。分析了液压注塑机工艺流程以及液压伺服系统控制原理,在此基础上将BP神经网络与PID控制方法相结合,通过神经网络的自学习、加权系数的调整,优化PID控制器参数 Ki 、Kp 、Kd,并将粒子群算法引入到神经网络中作为其学习算法,以有效提高BP神经网络算法的收敛速度。仿真结果表明,经过粒子群优化后的BP神经网络能够快速地对PID参数进行自适应调整,同时粒子群优化后的BP神经网络控制效果明显优于传统PID控制,该控制方法对于提升注塑机液压系统响应速度以及控制精度具有重要作用。Hydraulic system of injection moulding machine was a complex system with nonlinearity,large time delay and strong coupling.Because the parameters of traditional proportion integration differentiation(PID)control were fixed,the overshoot was large,the stability was poor and the control precision was low.The effect of hydraulic control of injection moulding machine was not ideal.An improved back propagation(BP)neural network PID control method was proposed.The process flow of hydraulic injection moulding machine and the control principle of hydraulic servo system were analyzed.On this basis,the BP neural network and the PID control method were combined.Through the self-learning of the neural network and the adjustment of the weighting coefficient,the parameters of the PID controller K i ,K p and K d were optimized.Particle swarm optimization algorithm was introduced into the neural network as its learning algorithm to effectively improve the convergence speed of the BP neural network algorithm degree.The simulation results show that the BP neural network optimized by particle swarm optimization could quickly adjust the PID parameters adaptively.At the same time,the control effect of BP neural network optimized by particle swarm optimization is obviously better than that of traditional PID control.The control method plays an important role in improving the response speed and controlling the accuracy of the hydraulic system of injection molding machine.

关 键 词：注塑机液压系统 传统PID BP神经网络 粒子群 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]