基于神经网络的IMC-PID控制方法的应用  

Application of NN-based IMC-PID in Total Pressure Control of Hydraulic Headbox

作　　者：牛绪钟[1] 汤伟[1] 胡连华[1] 王孟效[1] 

机构地区：[1]陕西科技大学工业自动化研究所,西安710021

出　　处：《化工自动化及仪表》2012年第1期6-9,共4页Control and Instruments in Chemical Industry

基　　金：科技型中小企业创新基金资助项目(10C26216103241);陕西省教育厅产业化中试项目(2010JC04)

摘　　要：水力式流浆箱对总压的控制要求较高,采用常规PID控制很难取得理想的效果,利用神经网络在线学习功能实时调整IMC-PID的控制参数,并将该算法应用到水力式流浆箱的总压控制系统中。在模型匹配和失配两种情况下的仿真和实际应用均表明,NN-IMC-PID具有较强的鲁棒性和较好的动态特性。Considering the fact that conventional PID control doesn＇ t act well in total pressure control of hydraulic headbox, making use of the NN （ neural network） online learning function to regulate IMC-PID controller parameters was proposed to have NN-based IMC-PID control applied to the hydraulic headbox pressure control. Simulation and application in both model-matching and mismatching cases prove its strong robustness and good dynamic characteristics

关 键 词：水力式流浆箱 总压控制 NN-IMC—PID S7—300PLC 

分 类 号：TH863.6[机械工程—仪器科学与技术]