Machine Learning for Structural Monitoring and Anomaly Detection

Favarelli, Elia (2021) Machine Learning for Structural Monitoring and Anomaly Detection, [Dissertation thesis], Alma Mater Studiorum Università di Bologna. Dottorato di ricerca in Monitoraggio e gestione delle strutture e dell'ambiente - sehm2, 33 Ciclo. DOI 10.48676/unibo/amsdottorato/9796.
Documenti full-text disponibili:
[img] Documento PDF (English) - Richiede un lettore di PDF come Xpdf o Adobe Acrobat Reader
Disponibile con Licenza: Salvo eventuali più ampie autorizzazioni dell'autore, la tesi può essere liberamente consultata e può essere effettuato il salvataggio e la stampa di una copia per fini strettamente personali di studio, di ricerca e di insegnamento, con espresso divieto di qualunque utilizzo direttamente o indirettamente commerciale. Ogni altro diritto sul materiale è riservato.
Download (11MB)

Abstract

Autonomous structural health monitoring (SHM) of a large number of structures became a topic of paramount importance for maintenance purposes and safety reasons in the last few decades. Civil infrastructures are the backbone of modern society, and the assessment of their conditions is of renowned importance. This aspect is even more exacerbated because of the existing system that are fast approaching their service life. Since the replacement of those structures is functionally and economically demanding, maintenance and retrofitting operations must be planned wisely. Moreover, the increasing amount and variety of data generated by users and sensors interconnected to the future 6G network requires new strategies to manage several types of data with highly different characteristics and also requires solutions to power the wireless network with renewable energies. In this scenario, the adoption of artificial intelligence and in particular machine learning (ML) strategies represents a flexible and potentially powerful solution that must be investigated. To manage the big and widespread amount of data generated by the extensive usage of multiple types of sensors, several ML techniques can be investigated, with the aim to perform data fusion and reduce the amount of data. Furthermore, the usage of anomaly detection techniques to identify potentially critical situations in infrastructures and buildings represents a topic of particular interest that still needs a significant investigation effort. In this research activity, we provide the fundamental guidelines to perform automatic damage detection, which combines SHM strategies and ML algorithms capable of performing anomaly detection on a wide set of structures. In particular, several algorithms and strategies capable of extracting relevant features from large amounts of data generated by different types of sensors are investigated. Finally, to effectively manage such an amount of data in communication constraints, we obtained some design rules for the acquisition system for bridge monitoring.

Abstract
Tipologia del documento
Tesi di dottorato
Autore
Favarelli, Elia
Supervisore
Co-supervisore
Dottorato di ricerca
Ciclo
33
Coordinatore
Settore disciplinare
Settore concorsuale
Parole chiave
Machine Learning, Deep Learning, Neural Network, Anomaly Detection, Structural Health Monitoring
URN:NBN
DOI
10.48676/unibo/amsdottorato/9796
Data di discussione
21 Maggio 2021
URI

Altri metadati

Statistica sui download

Gestione del documento: Visualizza la tesi

^