Repository logo
 
Publication

Sliced Wasserstein cycle consistency generative adversarial networks for fault data augmentation of an industrial robot

2023-07Journal article Restricted access
Simple item page
dc.contributor.authorPu, Ziqiang
dc.contributor.authorCabrera, Diego
dc.contributor.authorLi, Chuan
dc.contributor.authorValente de Oliveira, JOSÉ
dc.date.accessioned2023-04-20T12:26:59Z
dc.date.available2023-04-20T12:26:59Z
dc.date.issued2023-07
dc.description.abstractWe investigate the role of the loss function in cycle consistency generative adversarial networks (CycleGANs). Namely, the sliced Wasserstein distance is proposed for this type of generative model. Both the unconditional and the conditional CycleGANs with and without squeeze-and-excitation mechanisms are considered. Two data sets are used in the evaluation of the models, i.e., the well-known MNIST and a real-world in-house data set acquired for an industrial robot fault diagnosis. A comprehensive set of experiments show that, for both the unconditional and the conditional cases, sliced Wasserstein distance outperforms classic Wasserstein distance in CycleGANs. For the robot faulty data augmentation a model compatibility of 99.73% (conditional case) and 99.21% (unconditional case) were observed. In some cases, the improvement in convergence efficiency was higher than 2 (two) orders of magnitude.pt_PT
dc.description.sponsorshipNational Natural Science Foundation of China (NSFC) 52175080; Intelligent Manufacturing PHM Innovation Team Program 2018K-CXTD029pt_PT
dc.description.versioninfo:eu-repo/semantics/publishedVersionpt_PT
dc.identifier.doi10.1016/j.eswa.2023.119754pt_PT
dc.identifier.issn0957-4174
dc.identifier.urihttp://hdl.handle.net/10400.1/19480
dc.language.isoengpt_PT
dc.peerreviewedyespt_PT
dc.publisherElsevierpt_PT
dc.relationAssociate Laboratory of Energy, Transports and Aeronautics
dc.subjectSliced Wasserstein distancept_PT
dc.subjectGenerative adversarial networkspt_PT
dc.subjectCycle consistency generative adversarial networkspt_PT
dc.subjectConditional cycle consistency generative adversarial networkspt_PT
dc.subjectScarce faulty data augmentationpt_PT
dc.subjectIndustrial robotspt_PT
dc.titleSliced Wasserstein cycle consistency generative adversarial networks for fault data augmentation of an industrial robotpt_PT
dc.typejournal article
dspace.entity.typePublication
oaire.awardTitleAssociate Laboratory of Energy, Transports and Aeronautics
oaire.awardURIinfo:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F50022%2F2020/PT
oaire.citation.startPage119754pt_PT
oaire.citation.titleExpert Systems with Applicationspt_PT
oaire.citation.volume222pt_PT
oaire.fundingStream6817 - DCRRNI ID
person.familyNamePu
person.familyNameLUÍS VALENTE DE OLIVEIRA
person.givenNameZiqiang
person.givenNameJOSÉ
person.identifier.ciencia-id1F12-C1D3-7717
person.identifier.orcid0000-0003-4410-3493
person.identifier.orcid0000-0001-5337-5699
project.funder.identifierhttp://doi.org/10.13039/501100001871
project.funder.nameFundação para a Ciência e a Tecnologia
rcaap.rightsrestrictedAccesspt_PT
rcaap.typearticlept_PT
relation.isAuthorOfPublication48f95fa9-9a9c-457d-9517-6dca8f4b0264
relation.isAuthorOfPublicationbb726e73-690c-4a33-822e-c47bdac3035b
relation.isAuthorOfPublication.latestForDiscoverybb726e73-690c-4a33-822e-c47bdac3035b
relation.isProjectOfPublication9df77b70-8231-47e7-9b34-c702e9c6021c
relation.isProjectOfPublication.latestForDiscovery9df77b70-8231-47e7-9b34-c702e9c6021c

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Sliced Wasserstein cycle consistency generative adversarial networks for fault data augmentation of an industrial robot.pdf
Size:
4.32 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
3.46 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

FCT2-Artigos (em revistas ou actas indexadas)