Bearing fault feature extraction method based on complete ensemble empirical mode decomposition with adaptive noise

Journal of Vibroengineering

Kai Wen¹  Cunyi Zhang¹  Guosheng Geng¹  Maohua Xiao¹  Dan Wu¹  Longfei Xiong¹ 

[1] College of Engineering, Nanjing Agricultural University, Nanjing, 210031, China;

关键词: CEEMDAN;    EEMD;    denoising;    fault diagnosis;    feature extraction;    bearings;   

DOI  :  10.21595/jve.2018.19562

来源: DOAJ

【 摘 要 】

As an important part of rotating machinery, bearings play an important role in large-scale mechanical equipment. Abnormal bearing conditions may cause the machine to malfunction, or even evolve into a serious accident. Therefore, the accurate and timely fault diagnosis of the bearing is of great significance. Based on EMD, this paper introduces the working principles and characteristics of EEMD and CEEMDAN, respectively. Then the signal was decomposed by EEMD and CEEMDAN respectively. The simulation results show that CEEMDAN has better effect on signal decomposition. Then, comparing the effect of CEEMDAN and EEMD on bearing fault feature frequency extraction, the experiment proves that CEEMDAN has a better ability to preserve original signal and eliminate noise than EEMD method, and can extract bearing fault feature more accurately and timely.

【 授权许可】

Unknown