Abstract:

：The removal of gear vibration noise is limited because of the rotational speed fluctuation and there are some defects in order tracking commonly used in fault diagnosis of rolling bearing under a variable rotational speed, so the method of rolling bearing fault diagnosis based on fault characteristic trend template is proposed. The instantaneous dominant meshing multiply (IDMM) trend is extracted in down sampled time domain signal and the instantaneous fault characteristic frequency (IFCF) is extracted in envelope signal obtained by Hilbert transform of down sampled signal by using the Chirplet path pursuit. The fault characteristic trend is obtained by connecting the ratio points of corresponding IDMM and IFCF. The fault diagnosis is completed by matching the fault characteristic trend and fault characteristic trend template. The innovative method does not need to obtain the frequency band containing fault information which is presently a hard process. The paper defined the concept of fault characteristic trend, and tried to find the bearing fault feature in the fault characteristic trend. The effectiveness of the proposed method has been proved by both simulated and experimental bearing vibration signals.

Key words: bearing fault diagnosis, fault characteristic trend template, IDMM, IFCF, variable rotational speed, gear noise