Cargando…

A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data

Intelligent fault diagnosis methods based on deep learning have achieved much progress in recent years. However, there are two major factors causing serious degradation of the performance of these algorithms in real industrial applications, i.e., limited labeled training data and complex working con...

Descripción completa

Detalles Bibliográficos
Autores principales: He, Qiuchen, Li, Shaobo, Li, Chuanjiang, Zhang, Junxing, Zhang, Ansi, Zhou, Peng
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Hindawi 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9270151/
https://www.ncbi.nlm.nih.gov/pubmed/35814590
http://dx.doi.org/10.1155/2022/3024590
_version_ 1784744396621611008
author He, Qiuchen
Li, Shaobo
Li, Chuanjiang
Zhang, Junxing
Zhang, Ansi
Zhou, Peng
author_facet He, Qiuchen
Li, Shaobo
Li, Chuanjiang
Zhang, Junxing
Zhang, Ansi
Zhou, Peng
author_sort He, Qiuchen
collection PubMed
description Intelligent fault diagnosis methods based on deep learning have achieved much progress in recent years. However, there are two major factors causing serious degradation of the performance of these algorithms in real industrial applications, i.e., limited labeled training data and complex working conditions. To solve these problems, this study proposed a domain generalization-based hybrid matching network utilizing a matching network to diagnose the faults using features encoded by an autoencoder. The main idea was to regularize the feature extractor of the network with an autoencoder in order to reduce the risk of overfitting with limited training samples. In addition, a training strategy using dropout with random changing rates on inputs was implemented to enhance the model's generalization on unseen domains. The proposed method was validated on two different datasets containing artificial and real faults. The results showed that considerable performance was achieved by the proposed method under cross-domain tasks with limited training samples.
format Online
Article
Text
id pubmed-9270151
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher Hindawi
record_format MEDLINE/PubMed
spelling pubmed-92701512022-07-09 A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data He, Qiuchen Li, Shaobo Li, Chuanjiang Zhang, Junxing Zhang, Ansi Zhou, Peng Comput Intell Neurosci Research Article Intelligent fault diagnosis methods based on deep learning have achieved much progress in recent years. However, there are two major factors causing serious degradation of the performance of these algorithms in real industrial applications, i.e., limited labeled training data and complex working conditions. To solve these problems, this study proposed a domain generalization-based hybrid matching network utilizing a matching network to diagnose the faults using features encoded by an autoencoder. The main idea was to regularize the feature extractor of the network with an autoencoder in order to reduce the risk of overfitting with limited training samples. In addition, a training strategy using dropout with random changing rates on inputs was implemented to enhance the model's generalization on unseen domains. The proposed method was validated on two different datasets containing artificial and real faults. The results showed that considerable performance was achieved by the proposed method under cross-domain tasks with limited training samples. Hindawi 2022-07-01 /pmc/articles/PMC9270151/ /pubmed/35814590 http://dx.doi.org/10.1155/2022/3024590 Text en Copyright © 2022 Qiuchen He et al. https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
spellingShingle Research Article
He, Qiuchen
Li, Shaobo
Li, Chuanjiang
Zhang, Junxing
Zhang, Ansi
Zhou, Peng
A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data
title A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data
title_full A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data
title_fullStr A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data
title_full_unstemmed A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data
title_short A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data
title_sort hybrid matching network for fault diagnosis under different working conditions with limited data
topic Research Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9270151/
https://www.ncbi.nlm.nih.gov/pubmed/35814590
http://dx.doi.org/10.1155/2022/3024590
work_keys_str_mv AT heqiuchen ahybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT lishaobo ahybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT lichuanjiang ahybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT zhangjunxing ahybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT zhangansi ahybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT zhoupeng ahybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT heqiuchen hybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT lishaobo hybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT lichuanjiang hybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT zhangjunxing hybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT zhangansi hybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata
AT zhoupeng hybridmatchingnetworkforfaultdiagnosisunderdifferentworkingconditionswithlimiteddata