Gear-box rolling bearings' fault features extraction based on cepstrum editing and time domain synchronous average

In a gearbox, as rolling bearing fault signals are relatively weaker, the traditional rolling bearing fault diagnosis method is hard to extract bearing fault information from the gear-box vibration signals. In order to achieve the accurate extraction of bearing fault features, a method for bearing fault diagnosis based on cepstrum editing and time domain synchronous average was proposed here. With the proposed method, firstly the time domain synchronous average technology was used to enhance the gear-box vibration signals' components. Then, the correct position of cepstrum line corresponding to each component was obtained using the cepstrum analysis. Finally, the vibration components in its original signals were weakened with the cepstrum editing to highlight rolling bearings' fault features mixed in the original vibration signals and improve the correctness of gear-box rolling bearings' fault diagnosis. Simulation and test results verified the effectiveness of this method. ©, 2015, Chinese Vibration Engineering Society. All right reserved.