Civil infrastructures always meet the problem of material damage and deterioration during their service lives. Structural Health Monitoring (SHM) based on Artificial Neural Networks (ANNs) is a modern technology that has been widely applied to detect damages of civil structures in recent years. In this work, feedforward ANN and Convolutional Neural Network (CNN) are employed for vibration-based and image-based damage detection, respectively. The Component-Wise finite element method based on the Carrera Unified Formulation (CUF) is utilized for dynamic and static analysis of civil structures, which further helps in creating a dataset of damage scenarios and ensure the success of the ANN training process. Once the dynamic parameters or the image of displacement field from an unknown structure subjected to the same boundary condition is given, the corresponding trained ANNs can be able to detect damage locations and intensities of the structure accurately.

Damage Detection of Civil Structures using Artificial Neural Networks and Refined Component-Wise Finite Element Models / Enea, M.; Shen, J.; Pagani, A.; Carrera, E.. - (2022). ((Intervento presentato al convegno 25th International Conference on Composite Structures (ICCS25) tenutosi a Porto, Portugal nel 19-22 July 2022.

Damage Detection of Civil Structures using Artificial Neural Networks and Refined Component-Wise Finite Element Models

M. Enea;J. Shen;A. Pagani;E. Carrera
2022

Abstract

Civil infrastructures always meet the problem of material damage and deterioration during their service lives. Structural Health Monitoring (SHM) based on Artificial Neural Networks (ANNs) is a modern technology that has been widely applied to detect damages of civil structures in recent years. In this work, feedforward ANN and Convolutional Neural Network (CNN) are employed for vibration-based and image-based damage detection, respectively. The Component-Wise finite element method based on the Carrera Unified Formulation (CUF) is utilized for dynamic and static analysis of civil structures, which further helps in creating a dataset of damage scenarios and ensure the success of the ANN training process. Once the dynamic parameters or the image of displacement field from an unknown structure subjected to the same boundary condition is given, the corresponding trained ANNs can be able to detect damage locations and intensities of the structure accurately.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2970431