Electromagnetic-Thermal Analysis With FDTD and Physics-Informed Neural Networks

• Published in: IEEE journal on multiscale and multiphysics computational techniques Vol. 8; pp. 1 - 11
• Main Authors: Qi, Shutong, Sarris, Costas D.
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.01.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Boundary conditions; Electromagnetic fields; Electromagnetic heating; electrothermal problems; FDTD; Finite difference methods; Finite difference time domain method; Heating systems; Mathematical analysis; Mathematical models; Neural networks; Numerical methods; Solvers; Steady-state; Thermal analysis; Thermal stability; Thermodynamics; Time-domain analysis; unsupervised learning
• ISSN: 2379-8815; 2379-8815
• DOI: 10.1109/JMMCT.2023.3236946

This paper presents the coupling of the finite-difference time-domain (FDTD) method for electromagnetic field simulation, with a physics-informed neural network based solver for the heat equation. To this end, we employ a physics-informed U-Net instead of a numerical method to solve the heat equation. This approach enables the solution of general multiphysics problems with a single-physics numerical solver coupled with a neural network, overcoming the questions of accuracy and efficiency that are associated with interfacing multiphysics equations. By embedding the heat equation and its boundary conditions in the U-Net, we implement an unsupervised training methodology, which does not require the generation of ground-truth data. We test the proposed method with general 2-D coupled electromagnetic-thermal problems, demonstrating its accuracy and efficiency compared to standard finite-difference based alternatives.