Journal
IEEE ACCESS
Volume 8, Issue -, Pages 145194-145206Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2020.3012559
Keywords
Time-frequency analysis; Fault diagnosis; Vibrations; Rolling bearings; Feature extraction; Time-domain analysis; Rolling bearing; fault diagnosis; vibration image; empirical mode decomposition (EMD); pesudo-Wigner-Ville distribution (PWVD); fuzzy C-means (FCM) clustering
Categories
Funding
- National Natural Science Foundation of China [51605380]
- Natural Science Basic Research Project of Shaanxi Province in China [2019JLZ-08]
- Key Research and Development Project of Shaanxi Province in China [2019GY-093]
Ask authors/readers for more resources
Rolling bearing is key component of rotating machinery and its fault diagnosis is of great significance for reliable operation of machine. In this paper, an intelligent fault diagnosis method of rolling bearing based on FCM clustering of vibration images obtained by EMD-PWVD is presented. Firstly, vibration signals with different fault degrees are transformed into contour time-frequency images by EMD-PWVD. Secondly, vibration images are divided into sections and their energy distribution values are used as image feature. Then, feature vectors are constructed for known signals, which are standardized as inputs of FCM clustering to obtain classification matrix and clustering center. Finally, proximity between tested samples and clustering centers of known samples are calculated to realize identification of bearing faults. Experimental results show that identification accuracy of this proposed method is high. When adding noise, the proposed method is more stable than other vibration images such as grayscale and symmetrical polar coordinate image, and when the added noise with SNR of 5, the reduction rate of identification accuracy is obviously smaller than those of other two methods.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available