A Study on Dynamic Inversion Based Nonlinear Guidance Scheme for Autonomous Formation Flying

Kent Chua

Abstract


Due to increased applications of multiple UAVs in diverse scenarios, there is now extensive research in the area of coordinated control in formation flight. This project aims to present a three-dimensional nonlinear guidance scheme for autonomous flying of unmanned aerial vehicles (UAV). In the guidance philosophy presented, the followers’ formation commands are generated in the leader’s velocity frame. In doing so, the followers’ desired position commands are tracked while its’ velocity vectors are simultaneously aligned with that of the leader. Simulation studies performed indicate that the algorithm is capable of bringing the UAVs into formation from arbitrary positions and maintaining the formation in flight. Another part of this project is to eliminate the singularity issue associated with the use of Euler angles in the theorem which requires that the flight path angle be bounded away from vertical flight. Simulations have also been performed to determine the robustness of the algorithm in wind gusts. Wind forces in the environment have been simulated and the simulation results have been analyzed to conclude the guidance robustness. Finally, this project also presents a three-dimensional tool for visualizing autonomous formation flying based on dynamic inversion which allows viewers to verify the dynamic system behavior in a virtual reality environment.

Full Text:

PDF