Problem
Research questions and friction points this paper is trying to address.
Control Barrier Functions
Safety-critical Control
Model Perturbations
Robust Adaptive Control
Robotics
Neural Barriers: An Online Certifiable Learning-enhanced Adaptive High Order Safety Critical Control
🤖 AI Summary
This study addresses the failure of safety guarantees in Control Barrier Functions (CBFs) for real-world robots caused by model perturbations. We propose an online certifiable, robust adaptive CBF framework that integrates Neural Ordinary Differential Equations (Neural ODEs) with conformal prediction. Specifically, Neural ODEs are employed to adapt to unknown time-varying disturbances in real time, while conformal prediction quantifies uncertainty to dynamically balance safety and control efficiency. Under Lipschitz assumptions, we provide rigorous probabilistic safety guarantees. Experimental results demonstrate that the proposed method effectively handles model mismatch scenarios, including payload variations and wind disturbances, significantly reducing controller conservatism. Ultimately, this work achieves real-time adaptive safe control with formal theoretical assurances.
📝 Abstract
Control barrier functions are an effective model-based tool to formally certify the safety of a system. However, transferring their theoretical guarantees to real-world robotics systems requires high model fidelity. For example, payloads or wind disturbances can cause significant model perturbations to an aerial vehicle, leading to safety compromises. In this work, we propose a certifiable online learning-enhanced robust adaptive control barrier function, which adapts to disturbances using a Neural ODE and quantifies its adaptation uncertainty with conformal prediction. Our approach guarantees safety at all time under unknown time-varying model disturbances. It adopts a conservative strategy when the adaptation uncertainty is high; and efficiently adapts to reduce controller conservativeness as it receives more data. Our approach provides a provable safety guarantee with a probability bound under suitable Lipschitz smoothness assumptions on the underlying model and trajectory. These results demonstrate the potential of our method as a practical safety controller for robotics system operating under model perturbations.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Control Barrier Functions
Neural ODE
Conformal Prediction
Online Learning
Adaptive Control
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
