Fault diagnosis of rolling bearing based on improved wavelet transform and envelope analysis

Fault characteristic frequency components of rolling bearing vibration signals were extracted by the method of combining an improved wavelet decomposition and reconstruction algorithm and envelope analysis. The improved wavelet decomposition and reconstruction approach avoids the defects of Mallat algorithm's confusion in frequencies. Extracting the rolling bearing fault characteristic frequencies are more accurate. Analysis of normal rolling bearing and inner ring and the outer ring failure rolling bearing vibration signals indicate that the method can effectively diagnose the fault of rolling bearings. Furthermore,using this method succeeded in diagnosing the faults of aero-engine main bearings.