Learning to Run a Power Network

https://competitions.codalab.org/competitions/20767

Pypownet Installation

Requirements:

Python >= 3.6
Virtual Environment Recommended

cd L2RPN/
python setup.py install

Pypownet Introduction: https://github.com/MarvinLer/pypownet

Basic Usage

Runing Test

python Run_and_Submit_agent.py

You can modify the testing number of chronics and timesteps in the 'Run_and_Submit_agent.py' file.

Train Your Model

python -m pypownet.main -f train

To see all the options:

python -m pypownet.main --help

Key Files and Features

data
- Saved numpy files of action_space and generated train/val data
- Trained model
example_submission
- Sample submission to the L2RPN competition
parameters
- reward_signal, configuration, and training chronics of different power grids
public_data
- Extra data for IEEE-14 bus
pypownet
- agent.py: Defines the Dueling DQN agent
- analyze_action.py: Analyze the simulation results
- generate_action_space.py: Generate action space
- main.py: Main run file, including imitation, training, and test
- prepare_data.py: prepare data for imitation learning
- runner.py: key file controling the training process
Run_and_Submit_agent.py
- Test the trained model

License information

GEIRINA: https://www.geirina.net/
RTE: http://www.rte-france.com
INRIA: https://www.inria.fr/

This Source Code is subject to the terms of the GNU Lesser General Public License v3.0. If a copy of the LGPL-v3 was not distributed with this file, You can obtain one at https://www.gnu.org/licenses/lgpl-3.0.fr.html.

GuilinF/L2RPN