基于BP神经网络的冷轧薄板生产线故障诊断系统研究  

STUDY ABOUT TROUBLESHOOTING SYSTEM OF COLD-ROLLING SHEET LINE BASED ON BP NEURAL NET

作　　者：李福进[1] 安逸[2] 

机构地区：[1]河北理工大学,河北唐山063009 [2]河北理工大学计算机与自动控制学院,河北唐山063009

出　　处：《河北冶金》2008年第6期12-15,共4页Hebei Metallurgy

摘　　要：针对某钢厂冷轧薄板生产线电气传动故障类型多,传统故障诊断方法难以实现故障分离等情况,提出了一种基于BP神经网络的故障诊断方法,通过对各个传感器输出信号的分析可以获得故障特征量,将其输入神经网络后,由网络输出层的结点输出可以判断故障类型,从而实现故障诊断。Against the problems in cold - rolling sheet production line in some plant： so many kinds of electric drive breakdown, too difficult to analyze and shoot breakdown for the classical fault diagnosis method, a new method based BP neural net is proposed. By analyzing output signals from various sensors their fault characteristic quantity can be get and input into neural net, and then the fault type can be determined from node value of net output layer.

关 键 词：冷轧薄板 故障 诊断 研究 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]