Acting on the Tangent Space of the Constraint Manifold

Implementation of “Robot Reinforcement Learning on the Constraint Manifold”

[paper]
[website]

Install

pip install -e .

Run Examples

cd examples

CircularMotion Environment.

Environment options [A, E, T], algorithms options [TRPO, PPO, SAC, DDPG, TD3]

python circle_exp.py --render --env A --alg TRPO

PlanarAirHockey Environment.

Environment options [H, D, UH, UD], algorithms options [TRPO, PPO, SAC, DDPG, TD3]

python planar_air_hockey_exp.py --debug-gui --env H --alg SAC

IiwaAirHockey Environment.

Environment options [7H, RMP], algorithms options [TRPO, PPO, SAC, DDPG, TD3]

python iiwa_air_hockey_exp.py --debug-gui --env 7H --alg SAC

CollisionAvoidance Environment.

Environment options [C], algorithms options [TRPO, PPO, SAC, DDPG, TD3]

python collision_avoidance_exp.py --render --env C --alg SAC

Bibtex

@inproceedings{CORL_2021_Learning_on_the_Manifold,
  author =      "Liu, P. and  Tateo D. and  Bou-Ammar, H. and  Peters, J.",
  year =        "2021",
  title =       "Robot Reinforcement Learning on the Constraint Manifold",
  booktitle =   "Proceedings of the Conference on Robot Learning (CoRL)",
  key =	        "robot learning, constrained reinforcement learning, safe exploration",
}

GitHub

View Github