CFFsgram

CFFsgram: A candidate fault frequencies-based optimal demodulation band selection method for bearing fault diagnosis

How demodulate the vibration signal is an essential strategy for revealing the weak fault symptomatic of axle-box bearings. We proposed a candidate fault frequencies (CFFs)-based method, abbreviated as CFFsgram, for the optimal demodulation frequency band (DFB) identification of the axle-box bearing. The 1/3-binary tree filter bank constructed by empirical wavelet transform is adopted to divide the vibration signal into different narrowband with the same length. The local features of the squared envelope spectra of the narrowband signals are fully mined to identify the CFFs－a collection of frequencies most likely to be associated with bearing fault. An indicator calculated on the SESs of the narrowband signals is designed to guide the selection of the DFB. This file is the matlab code of CFFsgram. This code should use the EWT_Meyer_FilterBank.m that provided by Jerome Gilles (Institution: UCLA - Department of Mathematics,Year: 2012,Version: 1.0), , for which the authors are grateful.