Blind source separation of single-channel statistically correlated mechanical vibration signals based on subband extraction of EEMD
Meng Zon · Zhendong yu chongji · 2014
The traditional independent component analysis is too difficult to solve the problems of underdetermined blind source separation( BSS) and statistically correlated sources separation in mechanical fault diagnosis.Under the assumption of statistical indpendence between some sub-components of correlated machine vibration sources,a novel blind source separation method based on subband extraction of ensemble empirical mode decomposition( EEMD) was proposed to solve the problem of single-channel statistically correlated mechanical signals separation.In the method,the singlechannel signal was decomposed into a series of subband observed signals by ensemble empirical mode decomposition,then the number of source signals was estimated by singular value decomposition and Bayesian information criterion.New observed signals were reconstructed by using some selected subband observed signals with high independence according to the mutual information criterion and the number of sources,and the dimension of the new observed signal was increased.The source signals were estimated through the reconstructed observed signals by using whitening preprocessing and joint approximate diagonalization.The simulations and experiments testify the validity of the proposed method.