/smart-cities-drl

Urban mobility simulations with Python3, RLlib (Deep Reinforcement Learning), Mesa (Agent-based modeling) and OpenAI Gym.

Primary LanguageJupyter Notebook

Deep Reinforcement Learning for Smart Cities

Documentation

RLlib: https://docs.ray.io/en/master/rllib.html

Mesa: https://mesa.readthedocs.io/en/stable/

Installation

Clone repository and install dependencies.

git clone git@github.com:eescriba/smart-cities-drl.git
cd smart-cities-drl
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
pip install -e .

Training

Train environments in Jupyter notebooks with RLlib.

SmartCab

Open In Colab

WasteNet

Open In Colab

smart-cities-drl

Simulations

Run and visualize environments with Mesa.

mesa runserver src/[env_name]

smart-cities-drl