A self-adaptive analysis method of fault diagnosis in roller bearing based on Local mean decomposition

In view of the nonlinear and non-stationary characteristics of fault vibration signal in roller bearing, a self-adaptive fault diagnosis method known as LMD (Local mean decomposition) is proposed. Initially the original vibration signal is decomposed into several stationary PF (product function) which possessed physical meaning and a residual component by using of LMD. Subsequently, the main components in fault signal are determined by calculation of correlation factor of each PF with the original signal aiming at obtaining amplitude and frequency information. LMD is applied in analysis of simulation signals and fault diagnosis of bearing outer-race. The results indicate that LMD method of fault diagnosis in roller bearing is equipped with high fault recognition and identification rate. The characteristics of mechanical fault signals can be effectively extracted.