Use of Fuzzy Neural Network in Fault Diagnosis of Tilting Driving System in Tilting Train
Wu Hao · Zhongguo tiedao kexue · 2002
With the view of the fuzzy property used in fault diagnosis of the tilting driving system, the paper puts forward a fault diagnosis based on a fuzzy neural network (FNN), which combines fuzzy theory with neural network to implement fault fuzzy diagnosis. The structure of FNN is given, consisting of the fuzzy quantification section and the neural network section. The membership function is developed according to the actual situation. Meanwhile a fault discriminating method for threshold vector is employed to make fault identification even more flexible and reasonable. The table of fault identification is constructed on the basis of the related references, which are summarized in the theoretical sample of fault diagnosis of tilting driving system. It is applied to the fault diagnosis of tilting driving system. Several types of typical faults in tilting driving system are analyzed and identified with the fuzzy neural network. The simulation shows that it is useful and effective. The theoretical foundation of using the fault diagnosis in tilting driving system in real tilting train is provided.