An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring

Structural health monitoring (SHM) is critical for ensuring the safety and longevity of structures, yet existing methodologies often face challenges such as high data dimensionality, lack of interpretability, and reliance on extensive label datasets. Current research in SHM has primarily focused on...

Full description

Saved in:
Bibliographic Details
Main Authors: Tales Boratto, Heder Soares Bernardino, Alex Borges Vieira, Tiago Silveira Gontijo, Matteo Bodini, Dmitriy A. Martyushev, Camila Martins Saporetti, Alexandre Cury, Flávio Barbosa, Leonardo Goliatt
Format: Article
Language:English
Published: MDPI AG 2025-01-01
Series:Infrastructures
Subjects:
Online Access:https://www.mdpi.com/2412-3811/10/2/32
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849718959584051200
author Tales Boratto
Heder Soares Bernardino
Alex Borges Vieira
Tiago Silveira Gontijo
Matteo Bodini
Dmitriy A. Martyushev
Camila Martins Saporetti
Alexandre Cury
Flávio Barbosa
Leonardo Goliatt
author_facet Tales Boratto
Heder Soares Bernardino
Alex Borges Vieira
Tiago Silveira Gontijo
Matteo Bodini
Dmitriy A. Martyushev
Camila Martins Saporetti
Alexandre Cury
Flávio Barbosa
Leonardo Goliatt
author_sort Tales Boratto
collection DOAJ
description Structural health monitoring (SHM) is critical for ensuring the safety and longevity of structures, yet existing methodologies often face challenges such as high data dimensionality, lack of interpretability, and reliance on extensive label datasets. Current research in SHM has primarily focused on supervised approaches, which require significant manual effort for data labeling and are less adaptable to new environments. Additionally, the large volume of data generated from dynamic structural monitoring campaigns often includes irrelevant or redundant features, further complicating the analysis and reducing computational efficiency. This study addresses these issues by introducing an unsupervised learning approach for SHM, employing an agglomerative clustering model alongside an unsupervised feature selection technique utilizing box-plot statistics. The proposed method is assessed through raw acceleration signals obtained from four dynamic structural monitoring campaigns, including 44 features with temporal, statistical, and spectral information. In addition, these features are also evaluated in terms of their relevance, and the most important ones are selected for a new execution of the computational procedure. The proposed feature selection not only reduces data dimensionality but also enhances model interpretability, improving the clustering performance in terms of homogeneity, completeness, V-measure, and adjusted Rand score. The results obtained for the four analyzed cases provide clear insights into the patterns of behavior and structural anomalies.
format Article
id doaj-art-515786655eef4b00a733021ef673d794
institution DOAJ
issn 2412-3811
language English
publishDate 2025-01-01
publisher MDPI AG
record_format Article
series Infrastructures
spelling doaj-art-515786655eef4b00a733021ef673d7942025-08-20T03:12:15ZengMDPI AGInfrastructures2412-38112025-01-011023210.3390/infrastructures10020032An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health MonitoringTales Boratto0Heder Soares Bernardino1Alex Borges Vieira2Tiago Silveira Gontijo3Matteo Bodini4Dmitriy A. Martyushev5Camila Martins Saporetti6Alexandre Cury7Flávio Barbosa8Leonardo Goliatt9Graduate Program in Computational Modeling, Federal University of Juiz de Fora, Juiz de Fora 36036-900, Minas Gerais, BrazilDepartment of Computer Science, Federal University of Juiz de Fora, Juiz de Fora 36036-900, Minas Gerais, BrazilDepartment of Computer Science, Federal University of Juiz de Fora, Juiz de Fora 36036-900, Minas Gerais, BrazilCampus Centro Oeste, Federal University of São João del-Rei, Divinópolis 355901-296, Minas Gerais, BrazilDipartimento di Economia, Management e Metodi Quantitativi, Università degli Studi di Milano, Via Conservatorio 7, 20122 Milano, ItalyDepartment of Oil and Gas Technologies, Perm National Research Polytechnic University, 614990 Perm, RussiaDepartment of Computational Modeling, Polytechnic Institute, Rio de Janeiro State University, Nova Friburgo 22000-900, Rio de Janeiro, BrazilDepartment of Computational and Applied Mechanics, Federal University of Juiz de Fora, Juiz de Fora 36036-900, Minas Gerais, BrazilDepartment of Computational and Applied Mechanics, Federal University of Juiz de Fora, Juiz de Fora 36036-900, Minas Gerais, BrazilDepartment of Computational and Applied Mechanics, Federal University of Juiz de Fora, Juiz de Fora 36036-900, Minas Gerais, BrazilStructural health monitoring (SHM) is critical for ensuring the safety and longevity of structures, yet existing methodologies often face challenges such as high data dimensionality, lack of interpretability, and reliance on extensive label datasets. Current research in SHM has primarily focused on supervised approaches, which require significant manual effort for data labeling and are less adaptable to new environments. Additionally, the large volume of data generated from dynamic structural monitoring campaigns often includes irrelevant or redundant features, further complicating the analysis and reducing computational efficiency. This study addresses these issues by introducing an unsupervised learning approach for SHM, employing an agglomerative clustering model alongside an unsupervised feature selection technique utilizing box-plot statistics. The proposed method is assessed through raw acceleration signals obtained from four dynamic structural monitoring campaigns, including 44 features with temporal, statistical, and spectral information. In addition, these features are also evaluated in terms of their relevance, and the most important ones are selected for a new execution of the computational procedure. The proposed feature selection not only reduces data dimensionality but also enhances model interpretability, improving the clustering performance in terms of homogeneity, completeness, V-measure, and adjusted Rand score. The results obtained for the four analyzed cases provide clear insights into the patterns of behavior and structural anomalies.https://www.mdpi.com/2412-3811/10/2/32unsupervised learninganomaly detectionstructural health monitoringvibrationfeature selection
spellingShingle Tales Boratto
Heder Soares Bernardino
Alex Borges Vieira
Tiago Silveira Gontijo
Matteo Bodini
Dmitriy A. Martyushev
Camila Martins Saporetti
Alexandre Cury
Flávio Barbosa
Leonardo Goliatt
An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring
Infrastructures
unsupervised learning
anomaly detection
structural health monitoring
vibration
feature selection
title An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring
title_full An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring
title_fullStr An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring
title_full_unstemmed An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring
title_short An Agglomerative Clustering Combined with an Unsupervised Feature Selection Approach for Structural Health Monitoring
title_sort agglomerative clustering combined with an unsupervised feature selection approach for structural health monitoring
topic unsupervised learning
anomaly detection
structural health monitoring
vibration
feature selection
url https://www.mdpi.com/2412-3811/10/2/32
work_keys_str_mv AT talesboratto anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT hedersoaresbernardino anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT alexborgesvieira anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT tiagosilveiragontijo anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT matteobodini anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT dmitriyamartyushev anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT camilamartinssaporetti anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT alexandrecury anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT flaviobarbosa anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT leonardogoliatt anagglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT talesboratto agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT hedersoaresbernardino agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT alexborgesvieira agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT tiagosilveiragontijo agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT matteobodini agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT dmitriyamartyushev agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT camilamartinssaporetti agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT alexandrecury agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT flaviobarbosa agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring
AT leonardogoliatt agglomerativeclusteringcombinedwithanunsupervisedfeatureselectionapproachforstructuralhealthmonitoring