The aim of this work is to propose and analyze two functional ground motion mixed effects models, one calibrated for Italy and one for central Italy, by resorting to the theoretical framework developed by Scheipl, Staicu and Greven (2015). The addition of functional random effects to a functional estimate for the median collocates the model in a fully functional framework. The inclusion of random effects allows one to relax the ergodic assumption and, as a consequence, to remove some components of the ground motion variability from the aleatory component and transfer them to the epistemic uncertainty. In particular, the inclusion of two random effects in the proposed ground motion model for Italy makes it a partially non-ergodic model, while the ground motion model for central Italy, with the inclusion of four random effects, is a fully non-ergodic model. The analyses point out strengths and weaknesses of the proposed models, compared to their scalar and functional fixed effects counterparts.
Lo scopo di questo lavoro è quello di proporre ed analizzare due modelli funzionali ad effetti misti di movimento del suolo, uno calibrato per l’Italia e il secondo per il solo centro Italia, ricorrendo al framework sviluppato da Scheipl, Staicu e Greven (2015). L’aggiunta di effetti random funzionali ad una stima per la media funzionale consente di collocare i modelli in un contesto interamente funzionale. L’inclusione di effetti random permette di rilassare l’ipotesi di ergodicità e rimuovere alcune componenti della variabilità dalla componente aleatoria e di trasferirle nella componente epistemica. In particolare, l’inclusione di due effetti random nel modello di movimento del suolo per l’Italia lo rende un modello parzialmente non ergodico, mentre il modello per il centro Italia, con l’aggiunta di quattro effetti random, è un modello interamente non ergodico. Le analisi evidenziano i punti di forza e di debolezza dei modelli proposti, comparati con le loro controparti scalari e funzionali ad effetti fissi.
A functional additive mixed model approach for ground motion modelling : two case studies in Italy
Pirovano, Andrea
2021/2022
Abstract
The aim of this work is to propose and analyze two functional ground motion mixed effects models, one calibrated for Italy and one for central Italy, by resorting to the theoretical framework developed by Scheipl, Staicu and Greven (2015). The addition of functional random effects to a functional estimate for the median collocates the model in a fully functional framework. The inclusion of random effects allows one to relax the ergodic assumption and, as a consequence, to remove some components of the ground motion variability from the aleatory component and transfer them to the epistemic uncertainty. In particular, the inclusion of two random effects in the proposed ground motion model for Italy makes it a partially non-ergodic model, while the ground motion model for central Italy, with the inclusion of four random effects, is a fully non-ergodic model. The analyses point out strengths and weaknesses of the proposed models, compared to their scalar and functional fixed effects counterparts.File | Dimensione | Formato | |
---|---|---|---|
2022_12_Pirovano_01.pdf
accessibile in internet per tutti
Descrizione: Tesi
Dimensione
2.81 MB
Formato
Adobe PDF
|
2.81 MB | Adobe PDF | Visualizza/Apri |
2022_12_Pirovano_02.pdf
accessibile in internet per tutti
Descrizione: Executive Summary
Dimensione
455.58 kB
Formato
Adobe PDF
|
455.58 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/198690