When designing controllers for safety-critical systems, practitioners often face a challenging tradeoff between robustness and performance. While robust control methods provide …
S Liu, C Liu, J Dolan - Conference on Robot Learning, 2023 - proceedings.mlr.press
We propose new methods to synthesize control barrier function (CBF) based safe controllers that avoid input saturation, which can cause safety violations. In particular, our method is …
Y Yang, Y Jiang, Y Liu, J Chen… - IEEE Robotics and …, 2023 - ieeexplore.ieee.org
Safety is a critical concern when applying reinforcement learning (RL) to real-world control tasks. However, existing safe RL works either only consider expected safety constraint …
Modern nonlinear control theory seeks to endow systems with properties of stability and safety, and have been deployed successfully in multiple domains. Despite this success …
S Li, O Bastani - 2020 IEEE International Conference on …, 2020 - ieeexplore.ieee.org
We propose a framework for safe reinforcement learning that can handle stochastic nonlinear dynamical systems. We focus on the setting where the nominal dynamics are …
O Bastani - 2021 American control conference (ACC), 2021 - ieeexplore.ieee.org
Reinforcement learning is a promising approach to synthesizing policies for challenging robotics tasks. A key problem is how to ensure safety of the learned policy-eg, that a walking …
YC Chang, S Gao - 2021 IEEE International Conference on …, 2021 - ieeexplore.ieee.org
The lack of stability guarantee restricts the practical use of learning-based methods in core control problems in robotics. We develop new methods for learning neural control policies …
R Cheng, G Orosz, RM Murray, JW Burdick - Proceedings of the AAAI …, 2019 - aaai.org
Reinforcement Learning (RL) algorithms have found limited success beyond simulated applications, and one main reason is the absence of safety guarantees during the learning …
We provide a framework for incorporating robustness--to perturbations in the transition dynamics which we refer to as model misspecification--into continuous control …