TY - GEN
T1 - Physics-informed supervised residual learning for electromagnetic modeling
AU - Shan, Tao
AU - Song, Xiaoqian
AU - Guo, Rui
AU - Li, Maokun
AU - Yang, Fan
AU - Xu, Shenheng
N1 - Publisher Copyright:
© 2021 Applied Computational Electromagnetics Society.
PY - 2021/8/1
Y1 - 2021/8/1
N2 - In this paper, we propose a non-stationary iterative physics-informed supervised residual learning scheme (NSIP-ISRL) as a general framework for modeling electromagnetic wave propagation in inhomogeneous medium. NSIPISRL is based on the residual neural network (ResNet) that maps the residuals of matrix equation to the update of solutions [1]. It incorporates the concept of non-stationary iterative method, in which physical principles are embedded in the solution process through matrix-vector multiplication. NSIPISRL is applied to solve 2D volume integral equations in order to model electromagnetic wave interaction with lossy scatterers. The results show that NSIPISRL has a good accuracy and a strong generalization ability. The trained network can be applied to various scenarios with different scatterers, different incident angles, and different frequencies and still maintain a good accuracy.
AB - In this paper, we propose a non-stationary iterative physics-informed supervised residual learning scheme (NSIP-ISRL) as a general framework for modeling electromagnetic wave propagation in inhomogeneous medium. NSIPISRL is based on the residual neural network (ResNet) that maps the residuals of matrix equation to the update of solutions [1]. It incorporates the concept of non-stationary iterative method, in which physical principles are embedded in the solution process through matrix-vector multiplication. NSIPISRL is applied to solve 2D volume integral equations in order to model electromagnetic wave interaction with lossy scatterers. The results show that NSIPISRL has a good accuracy and a strong generalization ability. The trained network can be applied to various scenarios with different scatterers, different incident angles, and different frequencies and still maintain a good accuracy.
KW - Generalization
KW - Methods of Moments
KW - Non-stationary Iterative Method
KW - Non-stationary Physics-informed Supervised Residual Learning
KW - Residual Neural Network
KW - Volume Integral Equations
UR - https://www.scopus.com/pages/publications/85115919077
U2 - 10.1109/ACES53325.2021.00120
DO - 10.1109/ACES53325.2021.00120
M3 - 会议稿件
AN - SCOPUS:85115919077
T3 - 2021 International Applied Computational Electromagnetics Society Symposium, ACES 2021
BT - 2021 International Applied Computational Electromagnetics Society Symposium, ACES 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2021 International Applied Computational Electromagnetics Society Symposium, ACES 2021
Y2 - 1 August 2021 through 5 August 2021
ER -