Formation control of multiple UAVs with connectivity preservation on a directed graph

This paper investigates the formation control problem of multiple unmanned aerial vehicles (UAVs) connected via a directed communication network, with the aim of preserving connectivity while ensuring collision avoidance. First, a second-order dynamics model for the UAVs is established, and the communication constraints are formulated under a directed graph. To tackle the varying connectivity links, an edge-based consensus algorithm is employed, which allows each UAV to adjust its velocity and position by processing only local edge information. A connectivity-preserving potential function is constructed to ensure all UAVs remain within communication distance bounds but outside collision boundaries. The proposed controller is then derived using backstepping and Lyapunov methods, resulting in stability guarantees that asymptotically drive the formation to the desired configuration. Furthermore, to address the requirement for rapid attitude stabilization, a finite-time attitude tracking controller is designed, thereby enhancing the robustness of the overall formation under external disturbances or model uncertainties. Numerical simulations conducted on both directed spanning trees and directed cycles confirm the effectiveness of the proposed strategy, illustrating that all UAVs maintain formation connectivity throughout the flight and converge to the desired configuration.