Abstract:
Digital twins are virtual counterparts of complex physical systems that continually refine their predictive models through sensor‑driven data assimilation, and then leverage those models for forecasting and optimization of the system’s future behavior. Optimizing the system’s future behavior can also target the data acquisition process itself, framing an optimal experimental design or sensor placement problem. Common challenges in solving these problems include high computational cost in evolving high‑fidelity models, curse‑of‑dimensionality of high‑dimensional uncertainty, observation data are sparse in space and/or time, and noisy. Several fast and scalable computational methods to address these challenges will be presented. These methods enable real‑time and robust digital twins of complex physical systems, with application to real‑time flood prediction, medium‑range weather forecasting, and tumor growth monitoring.
Speaker’s Bio:
Dr. Chen is an Assistant Professor at the School of Computational Science and Engineering at the Georgia Institute Technology. Previously, he was a Research Scientist at the University of Texas at Austin, a postdoc and lecturer at Eidgenössische Technische Hochschule Zurich, and obtained his PhD at École Polytechnique Fédérale de Lausanne. Dr. Chen’s research is in the multidisciplinary fields of computational mathematics, data science, scientific machine learning, and parallel computing, with various applications in materials, energy, health, and natural hazard. His research focuses on developing fast, scalable, and parallel computational methods for integrating data and models under high‑dimensional uncertainty to enable (1) statistical model learning via Bayesian inference, (2) reliable system prediction with uncertainty quantification, (3) efficient data acquisition through optimal experimental design, and (4) robust control and design via stochastic optimization.
