Abstract
Uncertainty quantification has become an essential component of modern computational science, particularly in electromagnetic modeling where material properties, geometrical tolerances, excitation conditions, and environmental parameters are often imperfectly known. This talk presents the use of Polynomial Chaos methods as an efficient framework for propagating input uncertainties through deterministic electromagnetic solvers and quantifying their impact on system responses. By representing uncertain inputs and outputs through orthogonal polynomial expansions, Polynomial Chaos enables the estimation of statistical moments, sensitivities, and probability distributions with significantly fewer simulations than standard Monte Carlo approaches. The discussion will introduce the basic theoretical principles of Polynomial Chaos expansions and demonstrate how they can improve prediction reliability, support robust design, and provide deeper physical insight into electromagnetic systems affected by unavoidable variability.
About the Speaker
Aristeides D. Papadopoulos received his B.Sc. in Physics from the National and Kapodistrian University of Athens (NKUA), Athens, Greece, in 1999, his M.Sc in Physics from the University of Florida (UF), FL, USA in 2001, his M.Sc in Electrical Engineering from the Technical University of Delft (TUDelft) in 2003, his PhD in Electrical Engineering, from the School of Electrical and Computer Engineering, National Technical University of Athens (NTUA) in 2010 and his M.Sc. in Informatics, from the Department of Informatics and Telecommunications, NKUA in 2024. He is a senior post-doctoral researcher affiliated with the School of Electrical and Computer engineering-NTUA, the Center for Research and Technology Hellas / Informatics and Telematics Institute (CERTH/ITI) in Thessaloniki, Greece, the Department of Informatics and Communications (NKUA) and Eulambia Advanced Technologies in Athens-Greece. His research interests include Computational Electromagnetrics with applications in Optics and Fusion Plasma, Signal Processing for Communications and Machine Learning-AI applications in various fields. He has co-authored over 80 journal (42) and conference papers. Dr. Papadopoulos is an IEEE member, Senior Optica (formerly OSA) member and member of Euratom-EUROfusion Hellenic Republic. In addition he's teaching Electronics, Optoelectronics, and Communications at the Hellenic Air Force Academy.
