A New Method on Fault Diagnosis of Low-Speed Rolling Bearing Using Stress Waves

High-frequency stress wave analysis was used as characteristic parameter to detect the early stages of the loss of mechanical integrity in low-speed machinery in the paper. The background noise was eliminated using wavelet decomposition and the feature frequency of fault stress waves was extracted. Firstly, according to the characters of the fault stress waves obtained from a steel mill, db6 was selected to analyze the signals, because this kind of wavelet function have the best similarity with the fault stress waves. And then, multi-scaled decomposition was carried out in the analysis, and the D3 and D4 was reconstructed based on the comparison of the scale energy. At last, the characteristic signals of stress waves were extracted by applying wavelet transform successfully.