Foundations and applicability of transfer learning for structural health monitoring of bridges

Marcus Omori Yano; Eloi Figueiredo; Samuel da Silva; Alexandre Cury

doi:10.1016/j.ymssp.2023.110766

Foundations and applicability of transfer learning for structural health monitoring of bridges

Marcus Omori Yano, Eloi Figueiredo, Samuel da Silva, Alexandre Cury

Research output: Contribution to journal › Article › peer-review

24 Citations (Scopus)

Abstract

The number of bridges worldwide is extensive, making it financially and technically challenging for the authorities to install a structural health monitoring (SHM) system and collect large quantities of data for every bridge. Transfer learning has gained relevance in the last few years to extend the SHM concept for most bridges, while minimizing costs with monitoring systems and time with data measurement. It can be especially suitable for bridges structurally similar and replicated extensively, like overpasses integrated into highways. Therefore, this paper intends to lay down the foundations of transfer learning for SHM of bridges and to highlight the importance of the quality of knowledge transferred across different bridges for damage detection. Transfer Component Analysis, Joint Distribution Adaptation, and Maximum Independence Domain Adaptation methods are applied to data sets from different bridges, where classifiers have access to labeled training data from one bridge (source domain) and unlabeled monitoring test data from another bridge (target domain) that present similarities. The effectiveness of those methods is compared through the classification performance using real-world monitoring data sets collected from the Z-24 Bridge in Switzerland, and the PI-57 and PK 075+317 Bridges in France.

Original language	English
Article number	110766
Journal	Mechanical Systems and Signal Processing
Volume	204
DOIs	https://doi.org/10.1016/j.ymssp.2023.110766
Publication status	Published - 1 Dec 2023

Bibliographical note

Publisher Copyright:
© 2023 Elsevier Ltd

Funding

Marcus Omori Yano, Eloi Figueiredo, and Samuel da Silva thank the financial support provided by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES/Brazil)-Finance Code 001 and CAPES/FCT grant number 2019.00164.CBM and The Portuguese National Funding Agency for Science Research and Technology (FCT/Portugal) for promoting the collaboration between Brazil/Portugal. Marcus Omori Yano acknowledges the funding from the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES/Brazil) grant numbers 88882.433643/2019-01 and 88887.647575/2021-00. Eloi Figueiredo acknowledges the funding from the Portuguese National Funding Agency for Science Research and Technology (FCT/Portugal) through grant UIDB/04625/2020. Samuel da Silva is thankful for the Brazilian National Council of Technological and Scientific Development (CNPq) grant number 306526/2019-0 and the São Paulo Research Foundation (FAPESP) grant number 19/19684-3. Alexandre Cury thanks CNPq (Brazilian National Council of Technological and Scientific Development) grant number 303982/2022-5 and FAPEMIG (Fundação de Amparo a Pesquisa do Estado de Minas Gerais) grant number PPM-00001-18 for their financial support. The authors also thank Université Gustave Eiffel (former IFSTTAR/LCPC - Laboratoire Central des Ponts et Chaussées), SANEF (Société des Autoroutes du Nord et de l'Est de la France) for the data used in this paper for the PI-57 bridge (project 0560V407) and SNCF (Société Nationale des Chemins de fer Français) for the data used in this paper (project 01V0527 RGCU “Evaluation dynamique des ponts”). The authors also express their gratitude to the editor and anonymous reviewers for their numerous comments and positive suggestions for improving the article. Marcus Omori Yano, Eloi Figueiredo, and Samuel da Silva thank the financial support provided by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES/Brazil) -Finance Code 001 and CAPES/FCT grant number 2019.00164.CBM and The Portuguese National Funding Agency for Science Research and Technology (FCT/Portugal) for promoting the collaboration between Brazil/Portugal. Marcus Omori Yano acknowledges the funding from the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES/Brazil) grant numbers 88882.433643/2019-01 and 88887.647575/2021-00 . Eloi Figueiredo acknowledges the funding from the Portuguese National Funding Agency for Science Research and Technology (FCT/Portugal) through grant UIDB/04625/2020 . Samuel da Silva is thankful for the Brazilian National Council of Technological and Scientific Development (CNPq) grant number 306526/2019-0 and the São Paulo Research Foundation (FAPESP) grant number 19/19684-3 . Alexandre Cury thanks CNPq (Brazilian National Council of Technological and Scientific Development) grant number 303982/2022-5 and FAPEMIG (Fundação de Amparo a Pesquisa do Estado de Minas Gerais) grant number PPM-00001-18 for their financial support. The authors also thank Université Gustave Eiffel (former IFSTTAR/LCPC - Laboratoire Central des Ponts et Chaussées), SANEF (Société des Autoroutes du Nord et de l’Est de la France) for the data used in this paper for the PI-57 bridge (project 0560V407) and SNCF (Société Nationale des Chemins de fer Français) for the data used in this paper (project 01V0527 RGCU “Evaluation dynamique des ponts”). The authors also express their gratitude to the editor and anonymous reviewers for their numerous comments and positive suggestions for improving the article.

Funders	Funder number
Brazilian National Council of Technological and Scientific Development
LCPC
Laboratoire central des ponts et chaussées
Portuguese National Funding Agency for Science Research and Technology
SANEF
Société Nationale des Chemins de fer Français	01V0527
Société des Autoroutes du Nord et de l'Est de la France	0560V407
Fundação de Amparo à Pesquisa do Estado de São Paulo	303982/2022-5, 19/19684-3
Fundação para a Ciência e a Tecnologia	UIDB/04625/2020, 2019.00164, 88882.433643/2019-01, 88887.647575/2021-00
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
Conselho Nacional de Desenvolvimento Científico e Tecnológico	306526/2019-0
Fundação de Amparo à Pesquisa do Estado de Minas Gerais	PPM-00001-18
Institut français des sciences et technologies des transports, de l’aménagement et des réseaux
Université Gustave Eiffel

Keywords

Bridges
Joint distribution adaptation
Maximum independence domain adaptation
Structural health monitoring
Transfer component analysis
Transfer learning

Access to Document

10.1016/j.ymssp.2023.110766

Cite this

@article{b17b093974b843bf9f5911a0037b927d,

title = "Foundations and applicability of transfer learning for structural health monitoring of bridges",

abstract = "The number of bridges worldwide is extensive, making it financially and technically challenging for the authorities to install a structural health monitoring (SHM) system and collect large quantities of data for every bridge. Transfer learning has gained relevance in the last few years to extend the SHM concept for most bridges, while minimizing costs with monitoring systems and time with data measurement. It can be especially suitable for bridges structurally similar and replicated extensively, like overpasses integrated into highways. Therefore, this paper intends to lay down the foundations of transfer learning for SHM of bridges and to highlight the importance of the quality of knowledge transferred across different bridges for damage detection. Transfer Component Analysis, Joint Distribution Adaptation, and Maximum Independence Domain Adaptation methods are applied to data sets from different bridges, where classifiers have access to labeled training data from one bridge (source domain) and unlabeled monitoring test data from another bridge (target domain) that present similarities. The effectiveness of those methods is compared through the classification performance using real-world monitoring data sets collected from the Z-24 Bridge in Switzerland, and the PI-57 and PK 075+317 Bridges in France.",

keywords = "Bridges, Joint distribution adaptation, Maximum independence domain adaptation, Structural health monitoring, Transfer component analysis, Transfer learning",

author = "{Omori Yano}, Marcus and Eloi Figueiredo and {da Silva}, Samuel and Alexandre Cury",

note = "Publisher Copyright: {\textcopyright} 2023 Elsevier Ltd",

year = "2023",

month = dec,

day = "1",

doi = "10.1016/j.ymssp.2023.110766",

language = "English",

volume = "204",

journal = "Mechanical Systems and Signal Processing",

issn = "0888-3270",

publisher = "Academic Press",

}

TY - JOUR

T1 - Foundations and applicability of transfer learning for structural health monitoring of bridges

AU - Omori Yano, Marcus

AU - Figueiredo, Eloi

AU - da Silva, Samuel

AU - Cury, Alexandre

PY - 2023/12/1

Y1 - 2023/12/1

N2 - The number of bridges worldwide is extensive, making it financially and technically challenging for the authorities to install a structural health monitoring (SHM) system and collect large quantities of data for every bridge. Transfer learning has gained relevance in the last few years to extend the SHM concept for most bridges, while minimizing costs with monitoring systems and time with data measurement. It can be especially suitable for bridges structurally similar and replicated extensively, like overpasses integrated into highways. Therefore, this paper intends to lay down the foundations of transfer learning for SHM of bridges and to highlight the importance of the quality of knowledge transferred across different bridges for damage detection. Transfer Component Analysis, Joint Distribution Adaptation, and Maximum Independence Domain Adaptation methods are applied to data sets from different bridges, where classifiers have access to labeled training data from one bridge (source domain) and unlabeled monitoring test data from another bridge (target domain) that present similarities. The effectiveness of those methods is compared through the classification performance using real-world monitoring data sets collected from the Z-24 Bridge in Switzerland, and the PI-57 and PK 075+317 Bridges in France.

AB - The number of bridges worldwide is extensive, making it financially and technically challenging for the authorities to install a structural health monitoring (SHM) system and collect large quantities of data for every bridge. Transfer learning has gained relevance in the last few years to extend the SHM concept for most bridges, while minimizing costs with monitoring systems and time with data measurement. It can be especially suitable for bridges structurally similar and replicated extensively, like overpasses integrated into highways. Therefore, this paper intends to lay down the foundations of transfer learning for SHM of bridges and to highlight the importance of the quality of knowledge transferred across different bridges for damage detection. Transfer Component Analysis, Joint Distribution Adaptation, and Maximum Independence Domain Adaptation methods are applied to data sets from different bridges, where classifiers have access to labeled training data from one bridge (source domain) and unlabeled monitoring test data from another bridge (target domain) that present similarities. The effectiveness of those methods is compared through the classification performance using real-world monitoring data sets collected from the Z-24 Bridge in Switzerland, and the PI-57 and PK 075+317 Bridges in France.

KW - Bridges

KW - Joint distribution adaptation

KW - Maximum independence domain adaptation

KW - Structural health monitoring

KW - Transfer component analysis

KW - Transfer learning

UR - http://www.scopus.com/inward/record.url?scp=85171440512&partnerID=8YFLogxK

U2 - 10.1016/j.ymssp.2023.110766

DO - 10.1016/j.ymssp.2023.110766

M3 - Article

AN - SCOPUS:85171440512

SN - 0888-3270

VL - 204

JO - Mechanical Systems and Signal Processing

JF - Mechanical Systems and Signal Processing

M1 - 110766

ER -

Foundations and applicability of transfer learning for structural health monitoring of bridges

Abstract

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this