Bearing fault diagnosis method combining variable working condition calculation order tracking and spectral kurtosis

The invention discloses a bearing fault diagnosis method combining variable working condition calculation order tracking and spectral kurtosis. The bearing fault diagnosis method comprises the steps of: carrying out tree-shaped decomposition filtering on a rolling bearing vibration signal, calculating the spectral kurtosis, and drawing a spectral kurtosis graph to obtain a center frequency and anoptimal bandwidth of an optimal filtering bandwidth; carrying out band-pass filtering on the rolling bearing vibration signal, carrying out Hilbert transformation on the filtered signal, and carryingout envelope demodulation to obtain an envelope signal containing fault impact; setting a rotating speed pulse threshold value, acquiring a vibration signal pulse generation moment, solving an angulardomain time quadratic equation coefficient, acquiring an equal-angle sampling moment, carrying out interpolation fitting on the envelope signal, and acquiring an angular domain stable signal of whichthe transverse axis is an a