Journal Article10.1109/usnc-ursi52151.2023.10237743
2D Eigenmode Analysis Based on Physics Informed Neural Networks
Md. Rayhan Khan,Constantinos L. Zekios,Shubhendu Bhardwaj,Stavros V. Georgakopoulos +3 more
- 23 Jul 2023
pp 1015-1016
TL;DR: A novel deep learning approach for identifying eigenmode distributions of closed waveguides based on physics informed neural networks is presented. The approach is capable of identifying all the eigenmode distributions with an error of less than −12 dB.
read more
Abstract: In this work, a novel deep learning-based approach for identifying the modal field distributions of closed waveguides is introduced. Specifically, physics informed neural networks are used to solve the Helmholtz partial differential equation (PDE) with the imposition of the appropriate boundary conditions. Based on our results, the proposed approach is capable of identifying all the eigenmode distributions of the studied waveguides with an error of less than −12 dB when compared with both analytical and full-wave simulation results.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
References
•Proceedings Article
On optimization methods for deep learning
Jiquan Ngiam,Adam Coates,Ahbik Lahiri,Bobby Prochnow,Quoc V. Le,Andrew Y. Ng +5 more
- 28 Jun 2011
TL;DR: It is shown that more sophisticated off-the-shelf optimization methods such as Limited memory BFGS (L-BFGS) and Conjugate gradient (CG) with line search can significantly simplify and speed up the process of pretraining deep algorithms.
Deep learning in computer vision: A critical review of emerging techniques and application scenarios
Junyi Chai,Junyi Chai,Hao Zeng,Anming Li,Eric W.T. Ngai +4 more
- 15 Dec 2021
TL;DR: Deep learning has been overwhelmingly successful in computer vision (CV), natural language processing, and video/speech recognition, and as mentioned in this paper provides a critical review of recent achievements in terms of techniques and applications.
556
A Maxwell's Equations Based Deep Learning Method for Time Domain Electromagnetic Simulations
Pan Zhang,Yanyan Hu,Yuchen Jin,Shaogui Deng,Xuqing Wu,Jiefu Chen +5 more
- 08 Feb 2021
TL;DR: In this article, an unsupervised deep learning (DL) method for solving time domain electromagnetic simulations is discussed, which encodes initial conditions, boundary conditions as well as Maxwell's equations as the constraints when training the network, turning an electromagnetic simulation problem into an optimization process.
64
A Maxwell's Equations Based Deep Learning Method for Time Domain Electromagnetic Simulations
Pan Zhang,Yanyan Hu,Yuchen Jin,Shaogui Deng,Xuqing Wu,Jiefu Chen +5 more
- 26 May 2020
TL;DR: An unsupervised deep learning approach for time-domain electromagnetic simulations, based on physics informed neural network, encodes initial conditions, boundary conditions as well as governing equations as the constraints when training the network, turning an electromagnetic simulation problem into an optimization process.
55
A Generalized Approach to Real-Time Performance Estimation of Antenna Types Using Deep Learning
10 Jul 2022
TL;DR: In this article , a deep neural network was used to predict antenna performance using near-field data using an artificial neural network (ANN) and achieved acceptable antenna performance estimation, e.g., a mean absolute percentage error of 0.27% for an arbitrary antenna configuration.
2