Constructing and Navigating Connected Air Roads: A Safety-Critical Reinforcement Learning Approach for Multi-UAV Systems
Qihan Qi, Haojie Xia, Xinsong Yang, Jianquan Lu, Xingxing Ju
AI summary
Problem
Existing air-road designs are restricted to simple rectangular paths or suffer from local minima, while current safety controllers lack flexible road construction and efficient decentralized learning frameworks for complex multi-UAV systems.
Approach
The authors construct flexible connected air roads from arbitrary quadrilateral segments and use decentralized control barrier functions to filter a single-agent RL controller, ensuring collision avoidance and road adherence while drastically reducing training complexity.
Key results
- Modular quadrilateral-based connected air road construction framework
- Decentralized air-road and inter-UAV collision avoidance CBFs
- Single-agent RL training with CBF safety filter for scalable multi-UAV deployment
- Rigorous validation of safety, stability, and energy efficiency in simulations and real-world tests
Why it matters
Provides a scalable, computationally efficient safety framework for deploying multi-UAV fleets in complex, connected airspace for logistics and surveillance.
Abstract
This paper presents an integrated control method in air road navigation for multi-UAV systems, combining an efficient reinforcement learning (RL) controller with a control barrier function (CBF)-based filter that guarantees flight safety. First, an air road construction method based on arbitrary quadrilateral combinations is proposed, which enables flexible air road design. Second, two specific CBFs are designed: an air road CBF which keeps UAVs within designed air roads, and a collision avoidance CBF which prevents collisions between UAVs. Based on the CBF-based filter, the RL controller is allowed to be trained in a simple, single-agent environment, which reduces computational costs and enhances training effi- ciency. Furthermore, the RL reward is carefully designed, which considers both the stability during movement and the optimality of energy conservation. The performance, safety, and efficiency of the proposed approach are rigorously validated through comprehensive simulations and real-world experiments.