ASCL.net

Astrophysics Source Code Library

Making codes discoverable since 1999

ASCL Code Record

[ascl:2406.024] GRINN: Gravity Informed Neural Network for studying hydrodynamical systems

GRINN (Gravity Informed Neural Network) solves the coupled set of time-dependent partial differential equations describing the evolution of self-gravitating flows in one, two, and three spatial dimensions. It is based on physics informed neural networks (PINNs), which are mesh-free and offer a fundamentally different approach to solving such partial differential equations. GRINN has solved for the evolution of self-gravitating, small-amplitude perturbations and long-wavelength perturbations and, when modeling 3D astrophysical flows, provides accuracy on par with finite difference (FD) codes with an improvement in computational speed.

Code site:
https://github.com/sauddy/GRINN
Described in:
https://ui.adsabs.harvard.edu/abs/2024MLS%26T...5b5014A
Bibcode:
2024ascl.soft06024A
Keywords: NASA

Views: 224

ascl:2406.024
Add this shield to your page
Copy the above HTML to add this shield to your code's website.