An Efficient Stochastic Game Framework for Energy Management in Microgrid Networks

This repository contains all the code for the paper "An Efficient Stochastic Game Framework for Energy Management in Microgrid Networks". The code is organized as follows:

• The 3-agents folder contains all the code and analysis files for Setup1 (both normal and stochastic ADL).
• The 8-agents folder contains all the code and analysis files for Setup2 and Setup3 (both normal and stochastic ADL).

The file requirements.txt contains all the requirements needed to run the code. You can also use install.sh to set up a virtual environment and install all the dependencies automatically.

To install all the dependencies follow the following steps:

pip install -r requirements.txt

or for creating a virtual environment and then installing the dependencies:

chmod +x install.sh
./install.sh

Make sure you have virtualenv package installed. Check this guide to install virtualenv.

Steps to run the code

• Go to any folder which contains code. Each of the folders will contain:
  • main.py (contains all the code to run the experiment).
  • DQN_agent.py (contains classes and functions for the agents).
  • analysis notebook (analysis of all the results).
• Create two folders, 'logs' and 'saved', where all the results of the experiments and the trained models will be saved
• In the terminal type the following to run the code:

python main.py

• All the results will be stored in the 'logs' folder and all the trained models will be stored in the 'saved' folder.
• All the logs and saved models for all the experiments that we have performed can be found in this link.