Reproduce “Continuous calibration of a digital twin: comparison of particle filter and Bayesian calibration approaches” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Rebecca Ward, Ruchi Choudhary, Alastair Gregory, Melanie Jans-Singh, Mark Girolami
Researchers compared Particle Filtering and a sequential implementation of Bayesian calibration for continuously updating digital twin models with streamed data, demonstrating that Particle Filtering offers faster real-time parameter tracking while sequential Bayesian calibration provides insights into model bias for refinement. The study showed both methods successfully track time-varying parameters, outperforming static calibration approaches.
Key insights
- Both Particle Filtering and sequential Bayesian calibration effectively track time-varying parameters in digital twin models, significantly improving upon static calibration methods for dynamic systems.
- Particle Filtering offers substantial computational efficiency, providing quicker parameter estimates suitable for real-time data assimilation in operational digital twins.
- The Bayesian calibration framework (KOH) yields valuable diagnostic information, such as model bias and error terms, which can inform necessary improvements to the underlying physics-based model.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.