POLICEd RL: Learning Closed-Loop Robot Control Policies with Provable Satisfaction of Hard Constraints

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Bouvier, Jean-Baptiste, Nagpal, Kartik, Mehr, Negar
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916468029915136
author Bouvier, Jean-Baptiste
Nagpal, Kartik
Mehr, Negar
author_facet Bouvier, Jean-Baptiste
Nagpal, Kartik
Mehr, Negar
contents In this paper, we seek to learn a robot policy guaranteed to satisfy state constraints. To encourage constraint satisfaction, existing RL algorithms typically rely on Constrained Markov Decision Processes and discourage constraint violations through reward shaping. However, such soft constraints cannot offer verifiable safety guarantees. To address this gap, we propose POLICEd RL, a novel RL algorithm explicitly designed to enforce affine hard constraints in closed-loop with a black-box environment. Our key insight is to force the learned policy to be affine around the unsafe set and use this affine region as a repulsive buffer to prevent trajectories from violating the constraint. We prove that such policies exist and guarantee constraint satisfaction. Our proposed framework is applicable to both systems with continuous and discrete state and action spaces and is agnostic to the choice of the RL training algorithm. Our results demonstrate the capacity of POLICEd RL to enforce hard constraints in robotic tasks while significantly outperforming existing methods.
format Preprint
id arxiv_https___arxiv_org_abs_2403_13297
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle POLICEd RL: Learning Closed-Loop Robot Control Policies with Provable Satisfaction of Hard Constraints
Bouvier, Jean-Baptiste
Nagpal, Kartik
Mehr, Negar
Robotics
In this paper, we seek to learn a robot policy guaranteed to satisfy state constraints. To encourage constraint satisfaction, existing RL algorithms typically rely on Constrained Markov Decision Processes and discourage constraint violations through reward shaping. However, such soft constraints cannot offer verifiable safety guarantees. To address this gap, we propose POLICEd RL, a novel RL algorithm explicitly designed to enforce affine hard constraints in closed-loop with a black-box environment. Our key insight is to force the learned policy to be affine around the unsafe set and use this affine region as a repulsive buffer to prevent trajectories from violating the constraint. We prove that such policies exist and guarantee constraint satisfaction. Our proposed framework is applicable to both systems with continuous and discrete state and action spaces and is agnostic to the choice of the RL training algorithm. Our results demonstrate the capacity of POLICEd RL to enforce hard constraints in robotic tasks while significantly outperforming existing methods.
title POLICEd RL: Learning Closed-Loop Robot Control Policies with Provable Satisfaction of Hard Constraints
topic Robotics
url https://arxiv.org/abs/2403.13297