A Method of Extracting Submerged Periodic Impulse-Response Based on Fractional Dual Tree Complex Wavelet Transform with Matching Pursuit and Its Application

Abstract

It is a matter of great significance how to extract incipient and hidden fault feature from heavy noise. Extracting submerged periodic impact signal is one of important approach to fault diagnosis. Bearing failure is one of the most common reasons of machine breakdowns and accidents. Based on Matching Pursuit in dual fractional Fourier-wavelet domain, a novel method is proposed to extract the fault characteristic frequency for the incipient fault diagnosis of rolling element bearings in this paper. It is carried out in two steps. First, the fractional Fourier Transform and dual tree complex wavelet transform are implemented according to the criteria of the periodicity of its autocorrelation function to achieve maximum. Secondly, matching pursuit method in dual fractional Fourier-wavelet domain is applied to extracting submerged periodic signal. Taking orthogonal wavelet bases that mutually form Hilbert transform pairs as choice object, the method can match feature more effectively by a set of elementary functions of these dual orthogonal wavelets. From the theoretical analysis and the experimental studies, it is shown that the proposed method is effective especially to extract impulsive feature of the defects impacts on high intensity noise for a roller bearing system, while other matching pursuit method based on DWT and other wavelet de-noising methods based on threshold cannot do it completely. The effectiveness of the proposed method has been demonstrated by both simulation and experimental data.