Fault diagnosis based on integrated neural network and its simualtion research

Fault diagnosis is a process of information fusion, for the reasons of low fault diagnosis precision and credibility of traditional diagnosis methods, a fault diagnosis method fusing BP neural Network was presented to realize fault diagnosis. Based on multi-sensor information fusion, to use fault information fully, we give a diagnosis way of multi-network diagnosis respectively from different flank, then, integrate diagnostic result of each network, receiving eventual result. At last an exemple is given to indicate it´s validity. By simulation research, indicating the technique can advance veracity and resolve the shortage of single neural network fault diagnosis.