Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

University of Toronto

$83,000.00

Exploring New Frontiers for Computational Electromagnetics with Scientific Machine Learning

EF 2024-25

Résumé
Scientific machine learning is an area where machine learning algorithms and associated hardware are harnessed to develop new approaches to the solution of partial differential equations and inverse problems, optimization, and uncertainty quantification. The present proposal is aimed at exploring new possibilities for computational electromagnetics created by the rapid expansion of scientific machine learning techniques and the emergence of Physics-Informed Neural Networks (PINNs). This effort is driven by the ever-increasing need for high-fidelity and efficiency numerical tools for the solution of Maxwell's equations over a wide range of spatio-temporal scales and for the simulation of multi-physics interactions of electromagnetic fields including thermal and quantum effects. Such computations are involved in many cutting-edge technologies that flourish in Canada and contribute to the growth of the Canadian economy, from wireless communications to bio-engineering. The proposed project is organized in four interconnected foundational and application-oriented streams: (a) exploration of suitable PINN structures for Maxwell's equations and systematic derivation of their accuracy, stability and efficiency properties compared to standard solvers such as the Finite-Difference Time-Domain method for benchmark problems; (b) stable and accurate implementation of complex boundary conditions encountered in time-varying metasurfaces and electronic packaging geometries, and modeling of dispersive and nonlinear media that will extend the applicability of PINNs to challenging problems in photonics; (c) formulation of hybrid solvers, using PINNs as a platform for integrating heterogeneous models for extreme multiscale electromagnetics (for example, modeling wave propagation in "smart" electromagnetic environments with reconfigurable intelligent surfaces); (d) multiphysics computations, with emphasis on coupled electromagnetic and thermal/quantum phenomena. The long-term goal of this program is to develop a powerful simulation toolbox for electromagnetic field computations in advanced applications based on PINNs, resolving long-standing challenges of conventional techniques. We will disseminate the underlying theory to the research community and will facilitate its broad adoption in academia and industry, as an enabler of new discoveries of electromagnetic field based technologies. The proposed program is at the intersection of computational electromagnetics and machine learning, combining advanced electromagnetic concepts with new developments in machine learning. It is designed to create exciting and equitable training opportunities for undergraduate and graduate students. Targeting key problems from microelectronics to communications and photonics, this research is also bound to generate productive industry collaborations, leading to enriched student training and long-term benefits for relevant sectors of the Canadian economy.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
RGPIN-2024-04326
Name nom
Sarris, Costas
Department département
Electrical and Computer Engineering
Organization id
31
Institution établissement
University of Toronto
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
83000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1510
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et informatique
Area of application code
805
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Communications technologies (satellites, radar, etc.)
Area of application fr
Technologie des télécommunications (satellites, radar,etc.)
Research subject code
2502
Research subject group en
Electrical and electronic engineering
Research subject group fr
Génie électrique et électronique
Research subject en
Electromagnetics
Research subject fr
Électromagnétisme
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Exploring New Frontiers for Computational Electromagnetics with Scientific Machine Learning
Keywords
computational electromagnetics; finite difference time domain methods; scientific machine learning; neural networks; multiscale computational electromagnetics; multiphysics computational electromagnetics
Application summary
Scientific machine learning is an area where machine learning algorithms and associated hardware are harnessed to develop new approaches to the solution of partial differential equations and inverse problems, optimization, and uncertainty quantification. The present proposal is aimed at exploring new possibilities for computational electromagnetics created by the rapid expansion of scientific machine learning techniques and the emergence of Physics-Informed Neural Networks (PINNs). This effort is driven by the ever-increasing need for high-fidelity and efficiency numerical tools for the solution of Maxwell's equations over a wide range of spatio-temporal scales and for the simulation of multi-physics interactions of electromagnetic fields including thermal and quantum effects. Such computations are involved in many cutting-edge technologies that flourish in Canada and contribute to the growth of the Canadian economy, from wireless communications to bio-engineering. The proposed project is organized in four interconnected foundational and application-oriented streams: (a) exploration of suitable PINN structures for Maxwell's equations and systematic derivation of their accuracy, stability and efficiency properties compared to standard solvers such as the Finite-Difference Time-Domain method for benchmark problems; (b) stable and accurate implementation of complex boundary conditions encountered in time-varying metasurfaces and electronic packaging geometries, and modeling of dispersive and nonlinear media that will extend the applicability of PINNs to challenging problems in photonics; (c) formulation of hybrid solvers, using PINNs as a platform for integrating heterogeneous models for extreme multiscale electromagnetics (for example, modeling wave propagation in "smart" electromagnetic environments with reconfigurable intelligent surfaces); (d) multiphysics computations, with emphasis on coupled electromagnetic and thermal/quantum phenomena. The long-term goal of this program is to develop a powerful simulation toolbox for electromagnetic field computations in advanced applications based on PINNs, resolving long-standing challenges of conventional techniques. We will disseminate the underlying theory to the research community and will facilitate its broad adoption in academia and industry, as an enabler of new discoveries of electromagnetic field based technologies. The proposed program is at the intersection of computational electromagnetics and machine learning, combining advanced electromagnetic concepts with new developments in machine learning. It is designed to create exciting and equitable training opportunities for undergraduate and graduate students. Targeting key problems from microelectronics to communications and photonics, this research is also bound to generate productive industry collaborations, leading to enriched student training and long-term benefits for relevant sectors of the Canadian economy.