Fault diagnosis of rolling bearing based on convolutional neural network and transfer learning

Changjiang Pan; Xiang Yang

doi:10.1117/12.2689452

7 September 2023 Fault diagnosis of rolling bearing based on convolutional neural network and transfer learning

Changjiang Pan, Xiang Yang

Proceedings Volume 12790, Eighth International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023); 1279041 (2023) https://doi.org/10.1117/12.2689452
Event: 8th International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023), 2023, Hangzhou, China

Abstract

In practical engineering, the fault of rolling bearing is very common, so the research on fault diagnosis of rolling bearing is of great significance to the safe and stable operation of equipment. Because the vibration signals collected by different bearing faults are different and not easy to obtain, it is not easy to mark them. In order to solve this problem, a fault diagnosis classification and recognition method based on one-dimensional convolutional neural network and transfer learning is proposed. First of all, the vibration signal of the source domain is directly used as the input of the onedimensional convolutional neural network, and the model is trained; then transfer learning is introduced, some of the trained model parameters are frozen, and some target domain data are fine-tuned to reduce the training time; finally, the model is used to diagnose the fault of the target domain. The experimental results show that after using the source domain model transfer learning, the average accuracy of bearing fault classification is 95.82%, which has good adaptability.

Citation Download Citation

Changjiang Pan and Xiang Yang "Fault diagnosis of rolling bearing based on convolutional neural network and transfer learning", Proc. SPIE 12790, Eighth International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023), 1279041 (7 September 2023); https://doi.org/10.1117/12.2689452

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available