Publications

* means equal contributions.

Conference papers

1. [ICML’22] Dongjie Yu*, Haitong Ma*, Shengbo Eben Li and Jianyu Chen. “Reachability Constrained Reinforcement Learning”. Proceedings of the 39th International Conference on Machine Learning, PMLR 162:25636-25655, 2022.

tldr: We characterize feasible sets (persistently safe states) with Hamilton-Jacobi reachability analysis in constrained RL. The algorithm reachs a zero-violation policy with competitive performance.

[Paper] [Code_learning] [Code_env]

Journal papers

1. [T-ASE] Dongjie Yu*, Wenjun Zou, Yujie Yang, Haitong Ma, Shengbo Eben Li, Yuming Yin, Jianyu Chen and Jingliang Duan. “Safe Model-Based Reinforcement Learning with an Uncertainty-Aware Reachability Certificate”. IEEE Transactions on Automation Science and Engineering, DOI: 10.1109/TASE.2023.3292388, 2023.

tldr: We propose a distributional form of reachability certificate to consider the uncertaity of the system dynamics in safe RL. The algorithm is able to learn a zero-violation policy with competitive performance, significantly reducing violations during exploration.

[Preprint] [Code]

Preprints

1. TBD.