no code implementations • 26 Mar 2024 • Lawrence A. Bull, Chiho Jeon, Mark Girolami, Andrew Duncan, Jennifer Schooling, Miguel Bravo Haro
We formulate a combined model from simple units, representing strain envelopes (of each train passing) for two types of commuter train.
no code implementations • 9 Oct 2023 • Daniel R. Clarkson, Lawrence A. Bull, Tina A. Dardeno, Chandula T. Wickramarachchi, Elizabeth J. Cross, Timothy J. Rogers, Keith Worden, Nikolaos Dervilis, Aidan J. Hughes
At present, most surface-quality prediction methods can only perform single-task prediction which results in under-utilised datasets, repetitive work and increased experimental costs.
no code implementations • 15 May 2023 • Lawrence A. Bull, Matthew R. Jones, Elizabeth J. Cross, Andrew Duncan, Mark Girolami
Most interestingly, domain expertise and knowledge of the underlying physics can be encoded in the model at the system, subgroup, or population level.
no code implementations • 25 Jun 2022 • Aidan J. Hughes, Lawrence A. Bull, Paul Gardner, Nikolaos Dervilis, Keith Worden
For SHM applications, the value of information is evaluated with respect to a maintenance decision process, and the data-label querying corresponds to the inspection of a structure to determine its health state.
no code implementations • 23 Jun 2022 • Tina A. Dardeno, Lawrence A. Bull, Nikolaos Dervilis, Keith Worden
In this paper, an overlapping mixture of Gaussian processes (OMGP), was used to generate labels and quantify the uncertainty of normal-condition frequency response data from the helicopter blades.
no code implementations • 23 Jun 2022 • Aidan J. Hughes, Paul Gardner, Lawrence A. Bull, Nikolaos Dervilis, Keith Worden
For risk-based active learning in SHM, the value of information is evaluated with respect to a maintenance decision process, and the data-label querying corresponds to the inspection of a structure to determine its health state.
no code implementations • 7 Jan 2022 • Aidan J. Hughes, Lawrence A. Bull, Paul Gardner, Nikolaos Dervilis, Keith Worden
In contrast, the discriminative classifiers are shown to have excellent robustness to the effects of sampling bias.
no code implementations • 2 Mar 2021 • Lawrence A. Bull, Paul Gardner, Timothy J. Rogers, Elizabeth J. Cross, Nikolaos Dervilis, Keith Worden
In data-driven SHM, the signals recorded from systems in operation can be noisy and incomplete.