An abundant gear fault modulation signal is closely related to the gear fault type, especially to the gear fault severity. However, the modulation signal simultaneously includes coupled frequency modulation and amplitude modulation, which hinders the precise modulation separation and modulation-based gear fault severity assessment. Therefore, a new modulation separation method is proposed, which incorporates a rotation speed extraction technique based on the extreme value search, frequency modulation mechanism and Fourier series fitting. The rotational speed induced by the gear fault is first calculated by the extreme value search, which is then combined with a frequency modulation mechanism to solve the frequency modulation signal with the Fourier series fitting. Based on the vibration modulation signal model and Fourier series fitting, amplitude modulation is finally obtained. Simulation verifies the superiority of the proposed method in aspects of effectiveness and anti-noise performance compared with other modulation separation methods. The maximum relative errors of frequency modulation and amplitude modulation parameters under a signal-to-noise ratio of 0 dB are 2.9125% and 4.1143%, respectively. Two modulation intensity indicators regarding fault-induced frequency modulation and amplitude modulation signals are presented to assess gear faults. Experiment results also demonstrate the effectiveness of the proposed method in the severity assessment of misalignment and tooth breakage. Therefore, the research provides a new technique for gear fault severity assessment based on the frequency modulation or amplitude modulation signal.
Loading....