Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis
Abstract Deep learning has demonstrated notable success in mechanical signal processing with a large amount labelled data. However, the systems of the Jiaolong deep‐sea submersible prone to malfunction are typically diverse, due to the high complexity of its structure and operational environment. Co...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2024-02-01
|
Series: | Electronics Letters |
Subjects: | |
Online Access: | https://doi.org/10.1049/ell2.13122 |
_version_ | 1797306707529957376 |
---|---|
author | Penghui Zhao Xindi Wang Yi Zhang Yang Li Hongjun Wang Yang Yang |
author_facet | Penghui Zhao Xindi Wang Yi Zhang Yang Li Hongjun Wang Yang Yang |
author_sort | Penghui Zhao |
collection | DOAJ |
description | Abstract Deep learning has demonstrated notable success in mechanical signal processing with a large amount labelled data. However, the systems of the Jiaolong deep‐sea submersible prone to malfunction are typically diverse, due to the high complexity of its structure and operational environment. Consequently, this diversity gives rise to variations in the types of sensor signals and their associated data distributions that require analysis. Unsupervised domain adaptation (UDA) uses transferable knowledge derived from the source domain and applies it to an unlabelled target domain in order to improve the reusability of pre‐existing models and data. Inspired by the diffusion models that have the robust capabilities to transform data distributions across a large gap, we propose a novel diffusion‐based unsupervised domain adaptation (diffusion‐UDA) model, which leverages contrastive learning to alleviate the challenges of cross‐domain analysis for fault diagnosis within different systems of the deep‐sea submersible. Experimental results show the proposed method achieves state‐of‐the‐art performance on various benchmarks. |
first_indexed | 2024-03-08T00:45:44Z |
format | Article |
id | doaj.art-06d9781e483044988bad37d7cd382262 |
institution | Directory Open Access Journal |
issn | 0013-5194 1350-911X |
language | English |
last_indexed | 2024-03-08T00:45:44Z |
publishDate | 2024-02-01 |
publisher | Wiley |
record_format | Article |
series | Electronics Letters |
spelling | doaj.art-06d9781e483044988bad37d7cd3822622024-02-15T09:50:39ZengWileyElectronics Letters0013-51941350-911X2024-02-01603n/an/a10.1049/ell2.13122Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosisPenghui Zhao0Xindi Wang1Yi Zhang2Yang Li3Hongjun Wang4Yang Yang5School of Information Science and Engineering Shandong University Qingdao ChinaDepartment of Computer ScienceUniversity of Western OntarioLondon Ontario CanadaChina National Deep Sea Center Qingdao ChinaSchool of Information Science and Engineering Shandong University Qingdao ChinaSchool of Information Science and Engineering Shandong University Qingdao ChinaSchool of Information Science and Engineering Shandong University Qingdao ChinaAbstract Deep learning has demonstrated notable success in mechanical signal processing with a large amount labelled data. However, the systems of the Jiaolong deep‐sea submersible prone to malfunction are typically diverse, due to the high complexity of its structure and operational environment. Consequently, this diversity gives rise to variations in the types of sensor signals and their associated data distributions that require analysis. Unsupervised domain adaptation (UDA) uses transferable knowledge derived from the source domain and applies it to an unlabelled target domain in order to improve the reusability of pre‐existing models and data. Inspired by the diffusion models that have the robust capabilities to transform data distributions across a large gap, we propose a novel diffusion‐based unsupervised domain adaptation (diffusion‐UDA) model, which leverages contrastive learning to alleviate the challenges of cross‐domain analysis for fault diagnosis within different systems of the deep‐sea submersible. Experimental results show the proposed method achieves state‐of‐the‐art performance on various benchmarks.https://doi.org/10.1049/ell2.13122artificial intelligencefault diagnosissignal processing |
spellingShingle | Penghui Zhao Xindi Wang Yi Zhang Yang Li Hongjun Wang Yang Yang Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis Electronics Letters artificial intelligence fault diagnosis signal processing |
title | Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis |
title_full | Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis |
title_fullStr | Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis |
title_full_unstemmed | Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis |
title_short | Diffusion‐UDA: Diffusion‐based unsupervised domain adaptation for submersible fault diagnosis |
title_sort | diffusion uda diffusion based unsupervised domain adaptation for submersible fault diagnosis |
topic | artificial intelligence fault diagnosis signal processing |
url | https://doi.org/10.1049/ell2.13122 |
work_keys_str_mv | AT penghuizhao diffusionudadiffusionbasedunsuperviseddomainadaptationforsubmersiblefaultdiagnosis AT xindiwang diffusionudadiffusionbasedunsuperviseddomainadaptationforsubmersiblefaultdiagnosis AT yizhang diffusionudadiffusionbasedunsuperviseddomainadaptationforsubmersiblefaultdiagnosis AT yangli diffusionudadiffusionbasedunsuperviseddomainadaptationforsubmersiblefaultdiagnosis AT hongjunwang diffusionudadiffusionbasedunsuperviseddomainadaptationforsubmersiblefaultdiagnosis AT yangyang diffusionudadiffusionbasedunsuperviseddomainadaptationforsubmersiblefaultdiagnosis |