Fault diagnosis method based on integration of RSSD and wavelet transform to rolling bearing. (January 2019)

• Record Type: Journal Article
• Authors: Chen, Baojia; Shen, Baoming; Chen, Fafa; Tian, Hongliang; Xiao, Wenrong; Zhang, Fajun; Zhao, Chunhua
• Is Part Of: Measurement. Volume 131(2019)
• Page Start: 400
• Page End: 411
• Publication Date: 2019-01
• Subjects: Quality factor -- Sub-band reconstruction -- Rolling bearing -- Resonance-based sparse signal decomposition -- Wavelet transform; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2018.07.043
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: The manuscript proposes a new kind of fault diagnosis method combined RSSD and WT. The information is fused resonance attribute and local time–frequency characteristic. The fault feature recognition ability of the method is prior to WPD and EMD. Abstract: To solve the problem of early fault diagnosis of rolling bearing under strong background noise, a fault diagnosis method based on integration of Resonance-based Sparse Signal Decomposition (RSSD) and Wavelet Transform (WT) is proposed in this paper. The RSSD method is combined with quality factor optimization using genetic algorithm and sub-band reconstruction. Firstly, the early fault vibration signal of the rolling bearing is decomposed by RSSD. The kurtosis value of the low resonance component is taken as the objective function to optimize the combination of high and low quality factors with genetic algorithm. Then, the master sub-band is selected out to reconstruct the low resonance component based on the principle of energy dominant distribution. It can reduce the noise interference and enhance the impulse characteristic of the fault signal. Finally, characteristics of local optimization and multi-resolution of wavelet analysis considered, the multi-scale wavelet decomposition is applied to the reconstructed low resonance component to extract the fault features of the bearing failure deeply. The effectiveness and application value of the method are proved by two different diagnosis cases of rolling bearing faults. By comparisons, the fault feature extraction ability of the proposed method is prior to WPD method and similar or prior to EMD method for different bearing fault signals.