Integration of physics-informed operator learning and finite element method for parametric learning of partial differential equations

We present a method that employs physics-informed deep learning techniques for parametrically solving partial differential equations. The focus is on the steady-state heat equations within heterogeneous solids exhibiting significant phase contrast.