Authors:
Mario Cimino
1
;
Federico Galatolo
1
;
Marco Parola
1
;
Nicola Perilli
2
and
Nunziante Squeglia
2
Affiliations:
1
Dept. of Information Engineering, University of Pisa, 56122 Pisa, Italy
;
2
Dept. of Civil and Industrial Engineering, University of Pisa, 56122 Pisa, Italy
Keyword(s):
Structural Health Monitoring, Multi-sensor System, Transformer, LSTM, Leaning Tower of Pisa.
Abstract:
Structural health monitoring of buildings via agnostic approaches is a research challenge. However, due to the recent advent of pervasive multi-sensor systems, historical data samples are still limited. Consequently, data-driven methods are often unfeasible for long-term assessment. Nevertheless, some famous historical buildings have been subject to monitoring for decades, before the development of smart sensors and Deep Learning (DL). This paper presents a DL approach for the agnostic assessment of structural changes. The proposed approach has been experimented to the stabilizing intervention carried out in 2000-2002 on the leaning tower of Pisa (Italy). The data set is made by operational and environmental measures collected from 1993 to 2006. Both conventional and recent approaches are compared: Multiple Linear regression, LSTM and Tansformer. Experimental results are promising, and clearly shows a better change sensitivity of the LSTM, as well as a better modeling accuracy of the
Transformer.
(More)