TrafficReconstructionDL

Code for the paper Learning-based State Reconstruction for a Scalar Hyperbolic PDE under Lagrangian Sensing -> PDF

Description of the files

• godunov.py simulates the problem using a godunov numerical scheme.
• main.py is the main file, the one to launch.
• neural_network.py is a class for general neural network purposes (using Tensorflow).
• reconstruction_neural_network.py is a class that uses neural_network.py to create a reconstruction neural network.

Requirements

• Python 3.7.7
• Tensorflow 1.15
• matplotlib 3.3.1
• numpy 1.19.1
• pyDOE 0.3.8

Start a reconstruction

To start the file run main.py.