Fault Diagnosis of Rolling Bearings Based on High-order Transient-extraction Transform
The traditional time-frequency analysis method obtains the instantaneous frequency characteristics of non-stationary signals by constructing the time-frequency model of signals.However,due to the discontinuity of short-time tran-sient signals in time domain and the limitation of Heisenberg uncertainty principle,it is impossible to provide accurate time information for the short-time transient signals.In this work,aiming at the extraction of transient fault signals in rotating ma-chinery,a high-order transient extracting transform (HTET) was proposed.In this method,Taylor expansion was used to esti-mate the high-order group delay operator,which was used to replace the key parameters in the low-order time-frequency method.Through obtaining a more concentrated time-frequency representation,the accurate location of the impact character-istics formed in mechanical faults was determined.In addition,the impact component in the signal could be effectively ex-tracted and the interference of background noise was removed.Simulation analysis verified the performance of the proposed method in time-frequency feature characterization and noise suppression.The experimental analysis verified that the pro-posed method can be used for bearing fault diagnosis of artificial defects and naturally formed defects,which has practical significance for the expansion of signal processing theory and mechanical fault diagnosis.
万方数据
维普
