Publication: Transfer learning enabled bearing fault detection methods based on image representations of single-dimensional signals
dc.contributor.author | Deveci, Bilgin Umut | |
dc.contributor.author | Çeltikoğlu, Mert | |
dc.contributor.author | Albayrak, Özlem | |
dc.contributor.author | Ünal, Perin | |
dc.contributor.author | Kırcı, Pınar | |
dc.contributor.buuauthor | Çeltikoğlu, Mert | |
dc.contributor.buuauthor | KIRCI, PINAR | |
dc.contributor.department | Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Bilgisayar Mühendisliği Bölümü. | |
dc.contributor.researcherid | HXH-8622-2023 | |
dc.contributor.researcherid | CZK-0182-2022 | |
dc.date.accessioned | 2024-09-30T08:57:44Z | |
dc.date.available | 2024-09-30T08:57:44Z | |
dc.date.issued | 2024-02-21 | |
dc.description.abstract | Bearings are vital components in rotating machinery. Undetected bearing faults may result not only in financial loss, but also in the loss of lives. Hence, there exists an abundance of studies working on the early detection of bearing faults. The rising use of deep learning in recent years increased the number of imaging types/neural network architectures used for bearing fault classification, making it challenging to choose the most suitable 2-D imaging method and neural network. This study aims to address this challenge, by sharing the results of the training of eighteen imaging methods with four different networks using the same vibration data and training metrics. To further strengthen the results, the validation dataset size was taken as five times the training dataset size. The best results obtained is 99.89% accuracy by using Scattergram Filter Bank 1 as the image input, and ResNet-50 as the network for training. Prior to our work, Scattergram images have never been used for bearing fault classification. Ten out of 72 methods used in this work resulted in accuracies higher than 99.5%. | |
dc.identifier.doi | 10.1007/s10796-023-10371-z | |
dc.identifier.endpage | 1397 | |
dc.identifier.issn | 1387-3326 | |
dc.identifier.issue | 4, Special Issue SI | |
dc.identifier.startpage | 1345 | |
dc.identifier.uri | https://doi.org/10.1007/s10796-023-10371-z | |
dc.identifier.uri | https://link.springer.com/article/10.1007/s10796-023-10371-z | |
dc.identifier.uri | https://hdl.handle.net/11452/45486 | |
dc.identifier.volume | 26 | |
dc.identifier.wos | 000935851400001 | |
dc.indexed.wos | WOS.SCI | |
dc.language.iso | en | |
dc.publisher | Springer | |
dc.relation.journal | Information Systems Frontiers | |
dc.relation.publicationcategory | Makale - Uluslararası Hakemli Dergi | |
dc.rights | info:eu-repo/semantics/openAccess | |
dc.subject | Convolutional neural-network | |
dc.subject | Diagnosis | |
dc.subject | Frequency | |
dc.subject | Cnn | |
dc.subject | Rolling bearings fault determination | |
dc.subject | Transfer learning | |
dc.subject | Defect detection | |
dc.subject | Cnn | |
dc.subject | Deep learning | |
dc.subject | Googlenet | |
dc.subject | Resnet-50 | |
dc.subject | Squeezenet | |
dc.subject | Inception-resnet-v2 | |
dc.subject | Signal processing | |
dc.subject | Time-frequency images | |
dc.subject | Scattergram | |
dc.subject | Computer science | |
dc.title | Transfer learning enabled bearing fault detection methods based on image representations of single-dimensional signals | |
dc.type | Article | |
dc.type | Early Access | |
dspace.entity.type | Publication | |
relation.isAuthorOfPublication | 0270c3e7-f379-4f0e-84dd-a83c2bbf0235 | |
relation.isAuthorOfPublication.latestForDiscovery | 0270c3e7-f379-4f0e-84dd-a83c2bbf0235 |
Files
Original bundle
1 - 1 of 1