Research of on-line monitoring and fault diagnosis system for cold-rolling based on RBF neural network

Through the analysis of electric drive system of a cold-rolling steel plant and selecting detection signal reasonably, the on-line monitoring system has been exploited. It possesses the functions of real-time data display, alarm, sample data storage, data acquisition, parameter setting and others. By using MATLAB-Simulink tools, the simulation system has been built, which is for fault diagnosis of three-phase induction motor, a key machine of cold-rolling electric drives. By applying RBF neural network to diagnosis, a diagnosis system has been designed. Through verifying the trained network, the fault diagnosis system proves to have the good ability of predicting and diagnosing the faults of three-phase induction motor, and have a good application prospect.