基于神经网络的套印系统自抗扰控制  

Auto Disturbance Rejection Control Based on Neural Network in the Gravure Register Printing System

作　　者：郑军海[1] 薛必翠[2] 刘军[1] 

机构地区：[1]山东大学控制科学与工程学院,山东济南250061 [2]济南大学,山东济南250022

出　　处：《信息技术与信息化》2008年第4期80-83,共4页Information Technology and Informatization

基　　金：山东省自然科学基金资助(Y2006G02)

摘　　要：针对凹版套色印刷机系统的特点提出基于神经网络的自抗扰控制策略,并应用于凹版套色印刷机套印误差纠偏系统的控制。分析了自抗扰控制器的特点及应用神经网络优化后的优点。结合套印纠偏控制特点应用了改进后的自抗扰控制器,并仿真得出结果。表明应用基于神经网络的自抗扰控制器后,套印纠偏设备性能获得改善。According the characteristic of the gravure register printing system, the auto disturbance controllet （ADRC） based on the neural network is proposed and applied to the gravure register system to rectify the error of printing. Analyze the peculiarity of ADRC and the merit of the ADRC which is improved by neural network. Combining the characteristic of the en：or rectification of gravure printing system, the improved ADRC is applied. And obtain the result of simulation. The result shows that the performance of error rectification is improved after the ADRC based on neural network is applied.

关 键 词：套印纠偏系统 自抗扰控制器 神经网络 时滞系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]