부산대학교 도서관

We investigate the distributed time-varying formation tracking of networked quadrotors with input saturation and moving obstacles. The quadrotors’ position and attitude models are subjected to unknown external disturbances. This study has two main contributions. First, we develop filter-based distributed desired profiles and extended state observers (ESOs) for state-transformed nonlinear quadrotors to achieve time-varying formation tracking without requiring the velocity and acceleration information of the leader, followers, and moving obstacles. This achievement holds even in the presence of input saturation. Second, the proposed approach guarantees collision avoidance with moving obstacles, including adjacent quadrotors and unknown objects, by using only the relative distance from the obstacle, irrespective of external disturbances. This is accomplished by designing error functions for avoiding moving obstacles and including auxiliary signals in the formation tracker design. Based on the distributed profiles and the signals estimated using the ESOs, we design a collision-free time-varying formation tracker. The Lyapunov stability theory is utilized to prove that all signals of the proposed closed-loop formation tracking system are bounded, and the tracking errors converge to an adjustable bound that includes the origin. Finally, simulation results are provided to demonstrate the effectiveness of the proposed method.