Adversarial Based Unsupervised Domain Adaptation for Bearing Fault Diagnosis

Hongshu Wang, Rui Yang*

*Corresponding author for this work

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

3 Citations (Scopus)

Abstract

In this work we build an end-to-end adversarial domain adaptation model for bearing fault diagnosis on two different datasets. Two different adversarial based unsupervised domain adaptation models are implemented, and the obtained results are compared and analyzed. This project proposed a novel feature extractor model structure for bearing vibration signal, and pseudo-label semi-supervised learning is applied with the implemented Maximum Classifier Discrepancy (MCD) model. The proposed method outperforms the original method on XJTU and CWRU bearing datasets and achieves 97.25% accuracy.

Original languageEnglish
Title of host publication2022 27th International Conference on Automation and Computing
Subtitle of host publicationSmart Systems and Manufacturing, ICAC 2022
EditorsChenguang Yang, Yuchun Xu
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781665498074
DOIs
Publication statusPublished - 2022
Event27th International Conference on Automation and Computing, ICAC 2022 - Bristol, United Kingdom
Duration: 1 Sept 20223 Sept 2022

Publication series

Name2022 27th International Conference on Automation and Computing: Smart Systems and Manufacturing, ICAC 2022

Conference

Conference27th International Conference on Automation and Computing, ICAC 2022
Country/TerritoryUnited Kingdom
CityBristol
Period1/09/223/09/22

Keywords

  • Bearing Fault Diagnosis
  • Pseudo-label Semi-supervised Learning
  • Transfer Learning
  • Transferability Estimation
  • Unsupervised Domain Adaptation

Fingerprint

Dive into the research topics of 'Adversarial Based Unsupervised Domain Adaptation for Bearing Fault Diagnosis'. Together they form a unique fingerprint.

Cite this