A Diagnosis Method for Fault in Rolling Bearing
Lei Ye · Journal of Chongqing Institute of Technology · 2006
In view of the fault characteristics of rolling bearings,a method for fault diagnosis of rolling bearing based on integration of wavelet packet analysis,rough sets theory and neural networks is proposed.The signal is decomposed in proper grades by using wavelet packet transform,frequency band of signal is divided precisely,distributed situation of signal energy of each bands are regarded as characteristic quantity,forming the decision table of fault diagnosis; concise diagnosis rules are obtained by processing of rough sets theory and fault diagnosis system based on neural networks is established.The validity and feasibility of this method has been demonstrated by the practical examples.