镁蒸汽生成装置神经网络PID控制性能研究  

Research on Performance of Neural Network PID Control of Magnesium Vapor Generation Unit

作　　者：赵曦 苏东海[1] Zhao Xi;Su Donghai

机构地区：[1]沈阳工业大学机械工程学院,辽宁沈阳110870

出　　处：《一重技术》2022年第3期65-68,共4页CFHI Technology

摘　　要：以镁蒸汽生成装置为研究对象,利用古典控制理论推导镁蒸汽生成装置位置控制系统的传递函数并构建数学模型,为使送料系统工作过程具有高响应和高精度,采用PID控制优化系统的性能。结果表明,PID控制虽然使系统的控制性能更加理想,但有局限,后续用神经网络PID控制器对系统进行控制,使系统进一步优化。通过编写神经网络PID控制器程序,采用Matlab/Simulink软件构建仿真模型,得到阶跃响应曲线。对比结果表明,两种控制方法都可以优化系统,但是神经网络PID的优化效果能够缩短稳定时间,提高响应速度,消除超调。This paper adopts the classical control theory to derive the transfer function of the position control system of the magnesium vapor generation unit and establish the mathematical model,and adopts PID control optimization system to obtain high response and high precision during operation of the feeding system.The result shows that,although the adoption of PID control improves the control performance of the system,it has limitations,so the neural network PID controller is adopted to control the system and obtained further system optimization.Adopt Matlab/Simulink software to establish simulation model by writing neural network PID controller program,and obtain the step response curve.The comparative result shows that both control methods can optimize the system,but the neutral network PID can shorten the stabilization time,improve response speed,and eliminate overshoot.

关 键 词：镁蒸汽生成装置 神经网络PID PID 电液伺服 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]