{"title":"Combining transfer learning and numerical modelling to deal with the lack of training data in data-based SHM","authors":"","doi":"10.1016/j.jsv.2024.118710","DOIUrl":null,"url":null,"abstract":"<div><p>Structural health monitoring (SHM) involves continuously surveilling the performance of structures to identify progressive damage or deterioration that might evolve over time. Recently, machine learning (ML) algorithms have been successfully employed in various SHM applications, including damage detection. However, supervised ML algorithms often require labelled data for multiple possible damage states of the structure for successful damage identification. Although it may be feasible to gather such data for low-value structures, obtaining damage data for expensive structures such as aircraft could be highly challenging. Herein, this data insufficiency is addressed by combining Finite Element (FE) models with domain adaptation, specifically transfer component analysis (TCA) and joint domain adaptation (JDA). The proposed methodology is showcased in two case studies, a Brake–Reuß beam, where damage scenarios correspond to different torque settings on a lap joint and a wingbox laboratory structure where damage is introduced as saw-cuts. Supervised learning algorithms in the form of Artificial Neural Networks (ANNs) and K-Nearest Neighbours (KNNs) are trained based on FE data after domain adaptation is applied and are then tested with the experimental data. It is shown that even though the performance of classifiers in distinct scenarios of dual, three, four and five-class cases is sensitive to choices in the training stage, the use of TCA or JDA allows for the use of FE data for training and significantly reduces the need for expensive experimental damage data to be used for training. These results can pave the way for a broader use of ML algorithms in SHM of critical and/or expensive structures.</p></div>","PeriodicalId":17233,"journal":{"name":"Journal of Sound and Vibration","volume":null,"pages":null},"PeriodicalIF":4.3000,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S0022460X24004723/pdfft?md5=e0bc70543f8e8eb337978ada86bec52c&pid=1-s2.0-S0022460X24004723-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Sound and Vibration","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0022460X24004723","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ACOUSTICS","Score":null,"Total":0}
引用次数: 0
Abstract
Structural health monitoring (SHM) involves continuously surveilling the performance of structures to identify progressive damage or deterioration that might evolve over time. Recently, machine learning (ML) algorithms have been successfully employed in various SHM applications, including damage detection. However, supervised ML algorithms often require labelled data for multiple possible damage states of the structure for successful damage identification. Although it may be feasible to gather such data for low-value structures, obtaining damage data for expensive structures such as aircraft could be highly challenging. Herein, this data insufficiency is addressed by combining Finite Element (FE) models with domain adaptation, specifically transfer component analysis (TCA) and joint domain adaptation (JDA). The proposed methodology is showcased in two case studies, a Brake–Reuß beam, where damage scenarios correspond to different torque settings on a lap joint and a wingbox laboratory structure where damage is introduced as saw-cuts. Supervised learning algorithms in the form of Artificial Neural Networks (ANNs) and K-Nearest Neighbours (KNNs) are trained based on FE data after domain adaptation is applied and are then tested with the experimental data. It is shown that even though the performance of classifiers in distinct scenarios of dual, three, four and five-class cases is sensitive to choices in the training stage, the use of TCA or JDA allows for the use of FE data for training and significantly reduces the need for expensive experimental damage data to be used for training. These results can pave the way for a broader use of ML algorithms in SHM of critical and/or expensive structures.
期刊介绍:
The Journal of Sound and Vibration (JSV) is an independent journal devoted to the prompt publication of original papers, both theoretical and experimental, that provide new information on any aspect of sound or vibration. There is an emphasis on fundamental work that has potential for practical application.
JSV was founded and operates on the premise that the subject of sound and vibration requires a journal that publishes papers of a high technical standard across the various subdisciplines, thus facilitating awareness of techniques and discoveries in one area that may be applicable in others.