Online Learning-Enhanced High Order Adaptive Safety Control

2025-11-24

AI Summary

Conventional control barrier functions (CBFs) fail to guarantee safety under model uncertainty and complex time-varying disturbances—e.g., wind gusts and payload variations—due to reliance on precise system models. Method: This paper proposes an online learning–enhanced high-order adaptive safety control framework. Its core innovation is the first integration of neural ordinary differential equations (Neural ODEs) into the high-order CBF formalism, enabling real-time modeling, online estimation, and compensation of unknown dynamic disturbances without requiring exact system dynamics. The resulting hybrid architecture synergistically combines high-order CBF theory, adaptive control, and continuous-depth learning. Contribution/Results: Rigorous safety constraints are strictly enforced while robustness is significantly improved. Experiments on a 38-gram nano-quadrotor demonstrate stable maintenance of obstacle clearance under 18 km/h wind disturbances, markedly enhancing safety performance and generalization capability in realistic, complex environments.

Technology Category

• Natural Language Processing: Safety and Robustness
• Search and Optimization: Learning to Search
• Intelligent Robots: State Estimation

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Control barrier functions (CBFs) are an effective model-based tool to formally certify the safety of a system. With the growing complexity of modern control problems, CBFs have received increasing attention in both optimization-based and learning-based control communities as a safety filter, owing to their provable guarantees. However, success in transferring these guarantees to real-world systems is critically tied to model accuracy. For example, payloads or wind disturbances can significantly influence the dynamics of an aerial vehicle and invalidate the safety guarantee. In this work, we propose an efficient yet flexible online learning-enhanced high-order adaptive control barrier function using Neural ODEs. Our approach improves the safety of a CBF-certified system on the fly, even under complex time-varying model perturbations. In particular, we deploy our hybrid adaptive CBF controller on a 38g nano quadrotor, keeping a safe distance from the obstacle, against 18km/h wind.

Problem

• Addresses safety guarantee failures due to inaccurate system models
• Handles complex time-varying perturbations like payloads and wind
• Enables real-time safety adaptation for quadrotors under disturbances

Innovation

• Online learning-enhanced adaptive control barrier functions
• Neural ODEs for handling time-varying model perturbations
• Hybrid adaptive CBF controller tested on nano quadrotor