Intelligent fault diagnosis of rolling bearings based on normalized CNN considering data imbalance and variable working conditions

作者：

Highlights：

• A stable data-driven model with low structure complexity for data imbalance scenarios is designed.

• A novel application of BN for eliminating distribution differences is illustrated and applied.

• The proposed method focuses on the rolling bearing fault diagnosis without any signal preprocessing.

• The proposed normalized CNN model can be directly employed in the different working conditions.

摘要

•A stable data-driven model with low structure complexity for data imbalance scenarios is designed.•A novel application of BN for eliminating distribution differences is illustrated and applied.•The proposed method focuses on the rolling bearing fault diagnosis without any signal preprocessing.•The proposed normalized CNN model can be directly employed in the different working conditions.

论文关键词：Rolling bearing,Fault diagnosis,Convolutional neural network,Deep learning,Data imbalance

论文评审过程：Received 26 November 2019, Revised 22 April 2020, Accepted 23 April 2020, Available online 25 April 2020, Version of Record 6 May 2020.

论文官网地址：https://doi.org/10.1016/j.knosys.2020.105971