Physics‐Informed Neural Networks (PINNs) for Wave Propagation and Full Waveform Inversions
نویسندگان
چکیده
We propose a new approach to the solution of wave propagation and full waveform inversions (FWIs) based on recent advance in deep learning called Physics-Informed Neural Networks (PINNs). In this study, we present an algorithm for PINNs applied 2D acoustic equation test model with both forward FWIs case studies. These synthetic studies are designed explore ability handle varying degrees structural complexity using teleseismic plane waves seismic point sources. meshless formalism allows flexible implementation different types boundary conditions. For instance, our models demonstrate that PINN automatically satisfies absorbing conditions, serious computational challenge common solvers. Furthermore, priori knowledge subsurface structure can be seamlessly encoded formulation. find current state-of-the-art provide good results model, even though spectral element or finite difference methods more efficient accurate. More importantly, yield excellent all cases considered limited complexity. Using as geophysical inversion solver offers exciting perspectives, not only inversions, but also when dealing other datasets (e.g., magnetotellurics, gravity) well joint because its robust framework simple implementation.
منابع مشابه
A Potential Method for Body and Surface Wave Propagation in Transversely Isotropic Half- and Full-Spaces
The problem of propagation of plane wave including body and surface waves propagating in a transversely isotropic half-space with a depth-wise axis of material symmetry is investigated in details. Using the advantage of representation of displacement fields in terms of two complete scalar potential functions, the coupled equations of motion are uncoupled and reduced to two independent equations...
متن کاملNonlinear nonoverlapping Schwarz waveform relaxation for semilinear wave propagation
We introduce a nonoverlapping variant of the Schwarz waveform relaxation algorithm for semilinear wave propagation in one dimension. Using the theory of absorbing boundary conditions, we derive a new nonlinear algorithm. We show that the algorithm is well-posed and we prove its convergence by energy estimates and a Galerkin method. We then introduce an explicit scheme. We prove the convergence ...
متن کاملa potential method for body and surface wave propagation in transversely isotropic half- and full-spaces
the problem of propagation of plane wave including body and surface waves propagating in a transversely isotropic half-space with a depth-wise axis of material symmetry is investigated in details. using the advantage of representation of displacement fields in terms of two complete scalar potential functions, the coupled equations of motion are uncoupled and reduced to two independent equations...
متن کاملNeural Networks for Radar Waveform Recognition
For passive radar detection system, radar waveform recognition is an important research area. In this paper, we explore an automatic radar waveform recognition system to detect, track and locate the low probability of intercept (LPI) radars. The system can classify (but not identify) 12 kinds of signals, including binary phase shift keying (BPSK) (barker codes modulated), linear frequency modul...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Journal Of Geophysical Research: Solid Earth
سال: 2022
ISSN: ['2169-9356', '2169-9313']
DOI: https://doi.org/10.1029/2021jb023120