| QU Hongchun,ZHAN Yihong,LIU Xuhui,DUAN Gansen,TAI Heze.Rolling Bearing Fault Feature Extraction Method Based on TVF-EMD and CYCBD[J].航空发动机,2026,52(1):29-37 |
| Rolling Bearing Fault Feature Extraction Method Based on TVF-EMD and CYCBD |
| DOI：10.12482/ISSN.1672-3147.20241103001 |
| Key Words:fault feature extraction rolling bearing time-varying filtering empirical mode decomposition maximum second-order
cyclostationary blind deconvolution fault diagnosis aeroengine |
| Author Name | Affiliation | | QU Hongchun | College of Aeronautical Engineering,Civil Aviation University of China,Tianjin 300300,China | | ZHAN Yihong | College of Aeronautical Engineering,Civil Aviation University of China,Tianjin 300300,China | | LIU Xuhui | College of Aeronautical Engineering,Civil Aviation University of China,Tianjin 300300,China | | DUAN Gansen | College of Aeronautical Engineering,Civil Aviation University of China,Tianjin 300300,China | | TAI Heze | College of Aeronautical Engineering,Civil Aviation University of China,Tianjin 300300,China |
|
| Hits: 1737 |
| Download times: 583 |
| Abstract:To address the difficulty in extracting fault features of rolling bearings in aeroengine rotor systems under low signal-to-
noise ratio conditions, a fault feature extraction method based on time-varying filter empirical mode decomposition(TVF-EMD)and
maximum second-order cyclostationary blind deconvolution(CYCBD)was proposed.To mitigate the nonlinearity and non-stationarity of
fault signals,the decomposition parameters of TVF-EMD were optimized using the beluga whale optimization(BWO)algorithm.The intrinsic
mode components containing fault impact characteristics were selected through the envelope spectrum peak factor,and the relevant signals
were reconstructed.The filter length and cyclic frequency search range of the CYCBD algorithm were set,with the envelope spectrum
harmonic factor serving as the fitness function.The BWO was employed to select the optimal parameter combination for deconvolution
operations on the reconstructed signals.Fault characteristic frequencies were extracted through envelope spectrum analysis.The effective?
ness of the method was validated using fault simulation signals and bearing test rig datasets.The results demonstrate that the proposed
method is capable of extracting the 120 Hz fault frequency from the simulated signal and the 115.6 Hz fault frequency from the bearing test
rig dataset.It effectively extracts weak bearing fault features under noisy conditions,enhancing the accuracy of bearing fault diagnosis.By
integrating the TVF-EMD noise reduction technique with the CYCBD method,this study provides a novel and effective solution for rolling
bearing fault diagnosis. |
| View Full Text View/Add Comment Download reader |
|
|
|
