Evolutionary Constrained Reinforcement Learning

This is the implementation of the paper "Evolving Constrained Reinforcement Learning Policy" accepted by the International Joint Conference on Neural Networks (IJCNN) in 2023. The algorithm is adapted from Tianshou and original ERL.

Please use this bibtex if you use this repository in your work:

@inproceedings{hu2023ecrl,
  title={Evolving Constrained Reinforcement Learning Policy},
  author={Hu, Chengpeng and Pei, Jiyuan and Liu, Jialin and Yao, Xin},
  booktitle={2023 International Joint Conference on Neural Networks (IJCNN)},
  pages={accepted},
  year={2023},
  publisher={IEEE}
}

Requirments

Python=3.7
torch=1.8.1
gym==0.23.1
numpy==1.21.5
tianshou=0.4.6
MuJoCo150

Run

python main.py

For any problems, feel free to contact me (hucp2021@mail.sustech.edu.cn)