Physics-Informed Neural Networks for Estimation of Scattered Sound Fields with Boundary Condition

Sound field measurement helps us to understand the sound field, which is difficult to understand only by hearing. Recently, to reduce the measurement points, methods for estimating sound fields using physics-informed neural networks (PINNs) based on the wave equation or the Helmholtz equation have attracted much attention. In this study, we propose the sound field estimation method for the scattered sound field with PINNs by introducing the boundary condition into the loss function as additional physical laws. From the simulation experiments, the proposed method can improve the estimation accuracies in three-dimensional sound field with a scattering object with the acoustic impedance.