A Sparse Weighting Hoyer Indicator Fault Diagnostic Algorithm for Bearing

A bearing fault diagnosis method based on the sparse weighting Hoyer indicator (SWHI) and maximum overlap discrete wavelet packet transform (MODWPT) is proposed in response to the difficulty of identifying the optimal demodulation band in roller bearing fault diagnosis. First, given the problem that traditional sparse indicators (SIs) are vulnerable to random impacts, a more robust sparse weighting indicator is constructed by combining the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${L} _{2}$ </tex-math></inline-formula>/<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${L} _{1}$ </tex-math></inline-formula> norm and the Hoyer index (HI). Then, MODWPT is used to decompose the fault signal into all frequency bands and calculate the SWHI, and a SWHI value diagram (SWHIgram) of the fault signal is constructed. Finally, the optimal demodulation band is adaptively selected based on the maximum SWHI principle, and the original signal is filtered. The fault characteristics are recognized based on the envelope spectrum of the filtered signal. Simulation and experimental analysis of the signal show that compared with Autogram and other similar algorithms, the SWHIgram not only accurately identifies resonance bands and extracts single fault features under strong harmonic and random impact interference but also effectively separates composite fault features, demonstrating significant adaptability and robustness in complex fault scenarios.