Ritz Neural Network (RitzNN) Method for $H$ (curl) Problems

In this paper, we propose to use deep neural network method for numerically solving
$H$
(curl) problems through the Ritz loss functional formulation, where discretization of a PDE is set up as an optimization problem through the natural minimization principles. A two-dimensional test problem shows that the resulting RitzNN method is capable of approximating
$H$
(curl) problem with corner singularity accurately.