Applications of High-Dimensional Data Analytics in Structural Health Monitoring and Non-Destructive Evaluation: Thermal Videos Processing Using Tensor-Based Analysis
{"title":"Applications of High-Dimensional Data Analytics in Structural Health Monitoring and Non-Destructive Evaluation: Thermal Videos Processing Using Tensor-Based Analysis","authors":"Hamed Momeni, A. Ebrahimkhanlou","doi":"10.1115/imece2021-71878","DOIUrl":null,"url":null,"abstract":"\n This study reviews existing and potential applications of high-dimensional data analytics in the fields of structural health monitoring and non-destructive evaluation. Contrary to the high potential of these methods, the implemented applications in structural health monitoring and non-destructive evaluation topics are limited. In addition, with the ever-increasing development of measurement equipment, the necessity of using these methods is enhancing. In this paper, videos captured by different non-destructive evaluation techniques are studied as an example of high-dimensional data. Thermal videos are used for automatic damage detection and localization. Particularly, thermal cameras are employed to find delamination zones in composite plates, commonly used in aircraft wings. Due to the high-dimensional intrinsic of videos, using conventional statistical methods raise theoretical and practical challenges. One of the solutions to overcome these challenges is implementing tensor-based data analysis to analyze videos. Two tensor factorization methods are presented and employed to localize the damage automatically. The results show that the recorded video can be represented by a few vectors, which easily extract the time variation and extent of the damage.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"95 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1115/imece2021-71878","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
This study reviews existing and potential applications of high-dimensional data analytics in the fields of structural health monitoring and non-destructive evaluation. Contrary to the high potential of these methods, the implemented applications in structural health monitoring and non-destructive evaluation topics are limited. In addition, with the ever-increasing development of measurement equipment, the necessity of using these methods is enhancing. In this paper, videos captured by different non-destructive evaluation techniques are studied as an example of high-dimensional data. Thermal videos are used for automatic damage detection and localization. Particularly, thermal cameras are employed to find delamination zones in composite plates, commonly used in aircraft wings. Due to the high-dimensional intrinsic of videos, using conventional statistical methods raise theoretical and practical challenges. One of the solutions to overcome these challenges is implementing tensor-based data analysis to analyze videos. Two tensor factorization methods are presented and employed to localize the damage automatically. The results show that the recorded video can be represented by a few vectors, which easily extract the time variation and extent of the damage.