Modeling the Acoustic Effect of Porous Materials with Physics-Informed Neural Networks

In recent years, physics-informed neural networks (PINNs) have become increasingly popular due to their ability to incorporate physics knowledge into their training process. PINNs can be interpreted as a method to approximate solutions of partial differential equations (PDEs) by minimizing the loss function, which contains the residual of the PDE. In the proposed paper, this approach will be applied to approximate solutions of the Helmholtz equation in two dimensions with (i) a homogeneous medium and (ii) an inhomogeneous medium modeling the behavior of a porous material with the equivalent fluid model (EFM). The PINN results are compared to reference simulations obtained by the Finite Element Method.