Sparse Signal Decomposition Method Based on Multi-scale Chirplet and Its Application to Bearing Fault Diagnosis

Abstract

Based on the chirplet path tracing algorithm and sparse signal decomposition,a new sparse signal decomposition method based on multi-scale chirplet is proposed and applied to the decomposition of bearing failure vibration signals under non-stationary speed.The proposed method projects the signals onto the multi-scale chirplet base functions,and chooses the base functions adaptively according to the signal characteristics.Because of the multi-scale features of the base functions,this method is superior to the old sparse signal decomposition method,which adopts a single scale,and is more applicable to the decomposition of non-stationary signals whose frequency has a curve change.When the bearing has a failure,the relevant characteristic frequency in the vibration signal will fluctuate with the change of speed.The proposed method is very suitable to obtain the bearing failure characteristic frequency which fluctuates with the time,so it can be used to identify the falures of bearings.Simulation and a practical application example proves the effectiveness of the method.