🤖 AI Summary
This paper addresses cooperative localization and dynamic formation maintenance for fixed-wing UAV swarms in GPS-denied environments. Method: We propose a distributed affine formation control framework relying solely on onboard bearing-only measurements. For the first time, affine configuration control is unified with bearing-based cooperative localization—eliminating reliance on range measurements or global navigation systems. The approach employs SE(2) Lie group state representation and a fused extended Kalman filter (EKF) to construct a relative bearing observation model. Contribution/Results: The method enables autonomous reconfiguration and stable tracking of non-rigid, time-varying formations. Simulation and hardware-in-the-loop experiments demonstrate centimeter-level relative positioning accuracy, bearing tracking errors under 1°, and a 40% reduction in formation convergence time. The framework significantly enhances robustness and scalability under weak communication and GNSS-denied conditions.