Fault diagnosis of planet gear box based onparameter optimizations MCKD
                                          TANG Daolong, LI Hongkun, WANG Chaoge, HOU Menfan, YANG Rui
                                  (Institute of Mechanical Engineering, Dalian University of Technology, Dalian 116024, China)

Abstract: Aiming at the problem that the fault signal is difficult to extract and separate accurately when the vibration signal of planet gear box is too weak and affected by the noise pollution seriously, the transmission path is complex and changeable, a fault diagnosis method based on parameter optimizations MCKD was proposed. First of all, the MCKD method was used to reduce the strong noise of the original signal and to detect the cyclical shock components. Then the kurtosis and auto peak state coefficients as the screening criteria was set to select the optimal signal. Finally, the Hilbert envelope spectrum of the noise reduction signal was calculated to get the accurate fault character frequency. The results of simulation signal and engineering application analysis indicate that, the fault feature extraction of gearbox under strong background noise based on presented method is very effective, the noise is suppressed effectively and the fault signature is extracted successfully.

Key words: planet gearbox; wake fault diagnosis; MCKD; kurtosis; peak coefficient