Fault diagnosis of planetary gear based on wavelet real modulation zooming and resonance demodulation

In the fault diagnosis method of the resonance demodulation of the planetary gear of the wind turbine, there are some shortcomings like artificial difficulties in demodulation band and low demodulation spectral resolution. In this paper, wavelet packet decomposition and reconstruction method are used effectively to extract the resonance frequency components caused by the local fault of gear, which solves the traditional artificial limitations of resonance frequency demodulation. According to the principle of frequency shift modulation, a real modulation frequency-shift method is proposed, in which positive and negative frequency spectrum is achieved symmetrically. Wavelet packet filtering and real modulated signal are re-sampled, and then through the Hilbert transform, a real modulation zoom envelope spectrum is acquired. Finally, on the wind turbine fault simulation test bench, the fault diagnosis experimental analysis of the planetary gear shows that the method is correct and practical.