A novel method based on deep transfer unsupervised learning network for bearing fault diagnosis under variable working condition of unequal quantity

作者：

Highlights：

• A novel method called DCDBN-DMLP is proposed for bearing fault diagnosis with transfer unsupervised learning under varying working conditions.

• The dilated convolution deep belief network (DCDBN) is proposed to extract transferable characteristics from raw vibration dataset.

• Dynamic multilayer perceptron (DMLP) is proposed to classify bearing faults, and three transfer tasks with bearing fault dataset are verified.

摘要

•A novel method called DCDBN-DMLP is proposed for bearing fault diagnosis with transfer unsupervised learning under varying working conditions.•The dilated convolution deep belief network (DCDBN) is proposed to extract transferable characteristics from raw vibration dataset.•Dynamic multilayer perceptron (DMLP) is proposed to classify bearing faults, and three transfer tasks with bearing fault dataset are verified.

论文关键词：Intelligent fault diagnosis,Transfer unsupervised learning,Dilated convolution deep belief network (DCDBN),Dynamic multilayer perceptron (DMLP),Variable working conditions

论文评审过程：Received 15 June 2021, Revised 27 December 2021, Accepted 6 February 2022, Available online 12 February 2022, Version of Record 23 February 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.108381