Application of Demodulation Frequency Band Optimization in Compound Fault Detection of Rolling Element Bearing Based on Fast Spectrum Correlation

Abstract

The key to demodulation analysis is determining the appropriate demodulation frequency band. However, the coupling of compound faults significantly increases the difficulty. How to determine the fault-related demodulation frequency band is a challenge for the application of demodulation analysis in compound fault detection. To address this issue, a demodulation frequency band optimization method based on fast spectral correlation (Fast-SC) is proposed to detect compound faults of rolling element bearing in this paper. Firstly, the bi-variable map of vibration signal is obtained by Fast-SC. Secondly, the bi-variable map is sliced according to the theoretical fault-related frequency along the carrier frequency axis, obtaining a curve of correlation value versus carrier frequency, which is regarded as the target spectrum correlation curves (T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">scc</sub> ). Thirdly, the carrier frequency corresponding to the maximum value of the T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">scc</sub> obtained from each component is used to determine the initial demodulation frequency band. Fourthly, the crest of envelope spectrum (EC) is used to optimize the demodulation frequency band. Finally, the characteristic frequencies are exposed by utilizing the envelope demodulation on the frequency bands of each component. Simulation and experiment show that the proposed method is effective for the detection of rolling element bearing compound faults. Compared with maximum correlated kurtosis deconvolution(MCKD) and variational mode decomposition(VMD), the calculation cost of the proposed method is significantly reduced to about 6.6% of the above two methods.