Probabilistic Inference for Structural Health Monitoring: New Modes of Learning from Data

Abstract

In data-driven SHM, the signals recorded from systems in operation can be noisy and incomplete. Data corresponding to each of the operational, environmental, and damage states are rarely available a priori ; furthermore, labelling to describe the measurements is often unavailable. In consequence, the algorithms used to implement SHM should be robust and adaptive, while accommodating for missing information in the training-data -such that new information can be included if it becomes available.By reviewing novel techniques for statistical learning (introduced in previous work), it is argued that probabilistic algorithms offer a natural solution to the modelling of SHM data in practice. In three case-studies, probabilistic methods are adapted for applications to SHM signals -including semi-supervised learning, active learning, and multi-task learning.

An RUL-informed approach for life extension of high-value assets

“…In recent times, practices similar to those used in the field of datadriven PHM have been extensively applied to Structural Health Monitoring (SHM) to estimate the condition of structures and predict remaining fatigue life (Bull et al, 2021;Entezami et al, 2019;Entezami et al, 2021). Again, the data used for data-driven SHM and health condition assessment for asset structures are from sensors which typically log vibration and environmental condition data (Bhowmik, 2020).…”

A spectrum of physics-informed Gaussian processes for regression in engineering

“…GPs have been shown to be a powerful tool for regression tasks (Rasmussen and Williams, 2006), and their use in this context within engineering is becoming common (see, e.g., Kullaa, 2011;Avendaño-Valencia et al, 2017;Ni, 2018, 2019). The regular use of GP regression by the authors of this paper (e.g., Cross, 2012;Holmes et al, 2016;Bull et al, 2020;Rogers et al, 2020) is because of their simple ability to function given small datasets and, importantly, the Bayesian framework within which they naturally work; the predictive distribution provided allows the calculation of useful confidence intervals and the opportunity for uncertainty to be propagated forward into any following analysis (see, e.g., Gibson et al, 2020). Despite these advantages, their use in the provided citations remains entirely data-driven and thus open to the challenges/limitations discussed above.…”

Early Anomaly Warning of Environment-Induced Bridge Modal Variability through Localized Principal Component Differences

“…In this field, data-driven warning methods have been widely applied to discard environmental effects. There are two primary methods based on the use of environmental measurements: explicit and implicit (Soria et al 2016;Sun et al 2020;Bull et al 2021;Han et al 2021). The explicit method establishes an environmentfrequency regression model using a large amount of historical data to quantify the variability of the environment.…”