Skip to content

applying a deep CNN to find the pressure field for an uncertain geological scenario.

Notifications You must be signed in to change notification settings

amirdel/deepressure

Repository files navigation

Using Convolutional Neural Nets to model water contamination

What is this repository for?

This code is for using a CNN to solve the governing equations for contaminant transport in porous media. The neural net is trained to solve the pressure equation for a given geological scenario. The key idea used in this work was to use physical constraint (e.g. divergence free velocity field) in the loss function definition.

For details about the ideas and models used in this code please refer to the project poster and the full project report.

About

applying a deep CNN to find the pressure field for an uncertain geological scenario.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages