RBF在冷轧在线监测与故障诊断中的应用

摘要

通过对某钢厂冷轧电气传动系统分析,合理选取检测信号,构建了基于组态的实时在线监测系统,实现实时数据显示、报警、样本数据的存储、数据采集、参数设定等功能.利用MATLAB的Simulink工具构建了冷轧传动中的三相异步电动机故障诊断仿真系统.将RBF神经网络技术应用于冷轧电气传动系统的故障诊断,设计了基于RBF网络的三相异步电动机故障诊断系统.通过对训练好的网络进行验证,证明所设计的诊断方法能够对传动系统中的电动机故障进行很好的预测和判断,具有良好的实际应用前景.%The' article analyses the cold-rolling electric drives system of the steel mill. Though choosing the reasonable signal detection, on-line monitoring system based on configuration is constituted. The function of real-time data displaying, alarming, sample data collecting and storage, parameters setting is achieved. The diagnose simulation system of three wire a-synchronous motor is built though Matlab simulation tools. Three wire asynchronous motor fault diagnosis is designed though RBF neural network. Through the verification of the trajned network, the diagnosis system is proved that it can be well predicted and determine to the fault of the motor. The system has good application prospect.