TrafQ is a visual framework based on Open AI Gym environments and baseline algorithms for Traffic Light control optimization with Reinforcement Learning. Environments simulate real world road networks and traffic.
- A collection of RL environments for traffic light control of real road network, in the format of Open AI Gym environments
- A set of baseline algorithms to run on the network environments in the style of Open AI Baselines
- A webapp dashboard to provide an interface for training, testing and visualizing RL agents on road networks
Main features of the dashboard are:
- Listing and loading Open AI Gym road network environments
- Visualizing current simulation
- Listing and loading RL algorithms