Electromagnetic Inverse Design Based on Physics-Informed Neural Network With Generalization Across 1-D PEC Geometries
IEEE Transactions on Antennas and Propagation · 2026-07 · 74 (7) · 6634-6643
DOI: 10.1109/tap.2026.3675308 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Antennas and Propagation, 2026-07. Verfasst von Hongyuan Chang und 3 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Sichuan Province, Aeronautical Science Foundation of China.
Abstract (Originalsprache)
This work introduces a physics-informed neural operator (PINO) that integrates the equivalence principle to enable electromagnetic computation with generalization across 1-D perfect electric conductor (PEC) geometries and inverse design. Unlike traditional physics-informed neural networks (PINNs), which perform instance-specific regression for a single, fixed geometry, our framework learns a geometry-agnostic operator that maps any structure from a given parametric class (e.g., 1-D surface structure) to its electromagnetic responses by encoding Maxwell’s equations as physical constraints. We demonstrate this through 1-D metasurface design, where the operator—trained on procedurally generated geometries—uses an attention-driven architecture to model integral equation correlations inspired by the method of moments. The resulting network serves as a universal function approximator for Maxwell’s equations, bridging the gap between PINNs and neural operators, and facilitates inference-driven inverse design.
Förderung
- National Natural Science Foundation of China
- Natural Science Foundation of Sichuan Province
- Aeronautical Science Foundation of China
Zitieren
APA
Chang, H., Fan, J., Wang, R., & Wang, B. (2026). Electromagnetic Inverse Design Based on Physics-Informed Neural Network With Generalization Across 1-D PEC Geometries. IEEE Transactions on Antennas and Propagation, 74(7), 6634-6643. https://doi.org/10.1109/tap.2026.3675308
Vancouver
Chang H, Fan J, Wang R, Wang B. Electromagnetic Inverse Design Based on Physics-Informed Neural Network With Generalization Across 1-D PEC Geometries. IEEE Transactions on Antennas and Propagation. 2026;74(7):6634-6643. doi:10.1109/tap.2026.3675308
BibTeX
@article{Chang2026,
author = {Chang, Hongyuan and Fan, Jinsong and Wang, Ren and Wang, Bing-Zhong},
title = {Electromagnetic Inverse Design Based on Physics-Informed Neural Network With Generalization Across 1-D PEC Geometries},
journal = {IEEE Transactions on Antennas and Propagation},
year = {2026},
volume = {74},
number = {7},
pages = {6634--6643},
doi = {10.1109/tap.2026.3675308}
}
EUVIMED