SeismicNet: Physics-informed neural networks for seismic wave modeling in semi-infinite domain

Pu Ren, Chengping Rao, Su Chen, Jian-Xun Wang, Hao Sun, Yang Liu

Abstract: Highlights•Proposed a PINN model for forward modeling of seismic wave propagation in semi-infinite domain.•Incorporated absorbing boundary conditions to eliminate boundary effect in a truncated domain.•Introduced a temporal domain decomposition training strategy to scale up learning.•Considered parametric surrogate modeling in the context of different loading locations.•Demonstrated method efficacy via several numerical examples.