Abstract
Modeling periodic phenomena with accuracy is a key aspect to detect abnormal behaviour in time series for the context of Structural Health Monitoring. Modeling complex non-harmonic periodic pattern currently requires sophisticated techniques and significant computational resources. To overcome these limitations, this paper proposes a novel approach that combines the existing Bayesian Dynamic Linear Models with a kernel-based method for handling periodic patterns in time series. The approach is applied to model the traffic load on the Tamar Bridge and the piezometric pressure under a dam. The results show that the proposed method succeeds in modeling the stationary and non-stationary periodic patterns for both case-studies. Also, it is computationally efficient, versatile, self-adaptive to changing conditions, and capable of handling observations collected at irregular time intervals.
Highlights
In civil engineering, Structural Health Monitoring (SHM) is a field of study that aims at monitoring the state of a civil structure based on time series data
This study proposes a novel approach for modeling the nonharmonic periodic patterns in time-series for the context of Structural Health Monitoring
The technique combines the formulation of the existing Bayesian dynamic linear models with a kernel-based method
Summary
Structural Health Monitoring (SHM) is a field of study that aims at monitoring the state of a civil structure based on time series data. The effects related to traffic load variations generally display a non-harmonic weekly periodic pattern due to the day/night as well as the weekdays/weekend alternance (Westgate et al, 2015; Goulet and Koo, 2018) Regression techniques, such as multiple linear regression (Léger and Leclerc, 2007; Tatin et al, 2015; Gamse and Oberguggenberger, 2017) and artificial neural networks (Mata, 2011), are widely used to perform time series decomposition in SHM. This study proposes to combine the periodic kernel regression and the BDLM approaches to model complex non-harmonic periodic phenomena in SHM time series data.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
