Physics-informed neural networks for acoustic wave propagation

Ružička, Marek and Štancel, Martin and Imrich, Miroslav and Havrysh, Dmytro (2025) Physics-informed neural networks for acoustic wave propagation. In: Proceedings of the International Conference on Formal Methods and Foundations of Artificial Intelligence. Eszterházy Károly Katolikus Egyetem Líceum Kiadó, Eger, pp. 174-187. ISBN 9789634963035

Preview

Text fmfai2025_pp174-187.pdf - Published Version Download (1MB) | Preview

Abstract

Acoustic wave propagation plays a fundamental role in various scientific and engineering disciplines, including medical imaging, seismology, and acoustics. Traditional numerical methods such as the Finite Element Method (FEM) and Finite Difference Method (FDM) are widely used to model these waves [5, 24], but they often suffer from computational inefficiencies, especially for high-dimensional problems or complex geometries. This work explores the application of Physics-Informed Neural Networks (PINNs) as an alternative approach, leveraging deep learning to solve wave equations efficiently [14]. PINNs integrate physical laws directly into the neural network’s loss function, enabling solutions that adhere to the governing differential equations. We present a comparative analysis of PINNs with traditional numerical solvers, highlighting advantages, limitations, and potential improvements. Our experiments demonstrate that PINNs can effectively model wave propagation with comparable accuracy while reducing computational cost in certain scenarios.

Actions (login required)

Edit Item