LUAN Xiaochi,GAO Xiang,GUO Xiaopeng,YANG Jie,SHA Yundong.Rolling Bearing Fault Diagnosis Method Based on Multi-Parameter Fusion Joint Denoising[J].航空发动机,2026,52(1):48-58
Rolling Bearing Fault Diagnosis Method Based on Multi-Parameter Fusion Joint Denoising
DOI：10.12482/ISSN.1672-3147.20240731002
Key Words:joint denoising  rolling bearings  fault diagnosis  wavelet packet decomposition (WPD)  correlation coefficient  feature factor  envelope demodulation  aeroengine
Author NameAffiliation
LUAN Xiaochi College of Aero-Engine, Shenyang Aerospace University, Shenyang 110136, China 
GAO Xiang College of Aero-Engine, Shenyang Aerospace University, Shenyang 110136, China 
GUO Xiaopeng AECC Shenyang Engine ResearchInstitute, Shenyang 110015, China 
YANG Jie AECC Shenyang Engine ResearchInstitute, Shenyang 110015, China 
SHA Yundong College of Aero-Engine, Shenyang Aerospace University, Shenyang 110136, China 
Hits: 1762
Download times: 731
Abstract:To address the difficulty in extracting fault features from vibration signals of aeroengine rolling bearings caused by background noise, a rolling bearing fault diagnosis method featuring joint denoising based on wavelet packet decomposition (WPD), pearson correlation coefficient (PCC) criterion, and feature factor (θ-value) criterion was proposed. First, the vibration signal was decomposed into several signal components via WPD by calculating the optimal wavelet basis. The PCC and θ-value criteria were then applied for secondary screening. The retained components were reconstructed to achieve primary denoising. Subsequently, the reconstructed signal was processed using a savitzky-golay (SG) filter to perform secondary denoising and obtain the denoised signal. Finally, Hilbert envelope demodulation was applied to extract fault features. The proposed method was validated using simulated signals, the case western reserve university (CWRU) bearing datasets, and experimental data from an aeroengine intermediate bearing and main bearing. The kurtosis of the denoised simulated signal was 2.7 higher than that of the noisy signal. The results show that the proposed method can effectively suppress background noise and interference, enabling accurate identification of bearing fault features, and is therefore suitable for practical engineer? ing applications.
View Full Text  View/Add Comment  Download reader