engyasin/SLRL

Supervised and Reinforcement Learning Traffic Simulation Model

Traffic Simulation for mixed traffic based on supervised learning models guided by reinforcement learning.

This code is related to the paper titled:

Integrating Supervised and Reinforcement Learning for Heterogeneous Traffic Simulation

Installation:

pip install -r requirements.txt

Citation:

PDF of the paper is avaliable here

Will change later

@InProceedings{yousif2024,
    author="Yousif, Yasin and Müller, Jörg",
    title="Integrating Supervised and Reinforcement Learning for Heterogeneous Traffic Simulation",
    booktitle="Advances in Practical Applications of Agents, Multi-Agent Systems.",
    year="2024",
    publisher="Springer Nature Switzerland",
    address="Cham",
    note="To appear"
}

Videos of the results:

Intersection Case

Shared Space Case

To run the trained model, just cd to either unid_model or ind_model and run:

python trafficenv_D.py

Help

For discussing of issues and problems running the code, please consider creating an issue