Saved in:
Bibliographic Details
Main Authors: Cao, Hongpeng, Mao, Yanbing, Sha, Lui, Caccamo, Marco
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.13224
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915069698244608
author Cao, Hongpeng
Mao, Yanbing
Sha, Lui
Caccamo, Marco
author_facet Cao, Hongpeng
Mao, Yanbing
Sha, Lui
Caccamo, Marco
contents Real-world accidents in learning-enabled CPS frequently occur in challenging corner cases. During the training of deep reinforcement learning (DRL) policy, the standard setup for training conditions is either fixed at a single initial condition or uniformly sampled from the admissible state space. This setup often overlooks the challenging but safety-critical corner cases. To bridge this gap, this paper proposes a physics-model-guided worst-case sampling strategy for training safe policies that can handle safety-critical cases toward guaranteed safety. Furthermore, we integrate the proposed worst-case sampling strategy into the physics-regulated deep reinforcement learning (Phy-DRL) framework to build a more data-efficient and safe learning algorithm for safety-critical CPS. We validate the proposed training strategy with Phy-DRL through extensive experiments on a simulated cart-pole system, a 2D quadrotor, a simulated and a real quadruped robot, showing remarkably improved sampling efficiency to learn more robust safe policies.
format Preprint
id arxiv_https___arxiv_org_abs_2412_13224
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Physics-model-guided Worst-case Sampling for Safe Reinforcement Learning
Cao, Hongpeng
Mao, Yanbing
Sha, Lui
Caccamo, Marco
Robotics
Artificial Intelligence
Machine Learning
Real-world accidents in learning-enabled CPS frequently occur in challenging corner cases. During the training of deep reinforcement learning (DRL) policy, the standard setup for training conditions is either fixed at a single initial condition or uniformly sampled from the admissible state space. This setup often overlooks the challenging but safety-critical corner cases. To bridge this gap, this paper proposes a physics-model-guided worst-case sampling strategy for training safe policies that can handle safety-critical cases toward guaranteed safety. Furthermore, we integrate the proposed worst-case sampling strategy into the physics-regulated deep reinforcement learning (Phy-DRL) framework to build a more data-efficient and safe learning algorithm for safety-critical CPS. We validate the proposed training strategy with Phy-DRL through extensive experiments on a simulated cart-pole system, a 2D quadrotor, a simulated and a real quadruped robot, showing remarkably improved sampling efficiency to learn more robust safe policies.
title Physics-model-guided Worst-case Sampling for Safe Reinforcement Learning
topic Robotics
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2412.13224