Fault Diagnosis of Rolling Bearing Based on EMD Average Energy Method

The paper presents a fault diagnosis system based on the combination of EMD and BP Neural network.Firstly,in order to improve the signal to noise ratio,the method uses wavelet packet to pretreatment with the original vibration signal,so it will get the more suitable fault vibration signal for research.Secondly,decomposes the signal with EMD,then analysis the several interesting IMF taking the average energy of each IMF part obtained as input vector of BP neural network to train the BP network and realizes intelligent diagnosis of rolling bearing fault testified with real rolling bearing fault data.