Learning Wave Propagation with Attention-Based Convolutional Recurrent Autoencoder Net [article]

Indu Kant Deo, Rajeev Jaiman
2022 arXiv   pre-print
In this paper, we present an end-to-end attention-based convolutional recurrent autoencoder network (AB-CRAN) for data-driven modeling of wave propagation phenomena. To construct the low-dimensional learning model, we employ a denoising-based convolutional autoencoder from the full-order snapshots of wave propagation generated by solving hyperbolic partial differential equations. The proposed deep neural network architecture relies on the attention-based recurrent neural network with long
more » ... term memory cells. We assess the proposed AB-CRAN framework against the recurrent neural network for the low-dimensional learning of wave propagation. To demonstrate the effectiveness of the AB-CRAN model, we consider three benchmark problems namely one-dimensional linear convection, nonlinear viscous Burgers, and a two-dimensional Saint-Venant shallow water system. Using the time-series datasets from the benchmark problems, our novel AB-CRAN architecture accurately captures the wave amplitude and preserves the wave characteristics of the solution for long time horizons. The attention-based sequence-to-sequence network increases the time-horizon of prediction by five times compared to the standard recurrent neural network with long short-term memory cells. Denoising autoencoder further reduces the mean squared error of prediction and improves the generalization capability in the parameter space.
arXiv:2201.06628v3 fatcat:gr74fjdqrfezfdhm5bt4ajlhjy