Decentralized Formation Control for Multiple Quadrotors under Unidirectional Communication Constraints

Abstract

This paper addresses the problem of formation control for multiple Unmanned Aerial Vehicles (UAVs) with relative sensing capabilities and operating under input–state limitations. More specifically, we propose a novel distributed leader-follower architecture for a team of aerial robots to follow cooperatively a desired path while maintaining a predefined formation geometry. In the proposed control strategy, knowledge of the desired formation path is restricted to the leader UAV, which can also broadcast its state information to all followers. In this way, the leader UAV, by employing a Nonlinear Model Predictive Control (NMPC) law, tries to navigate the whole group of agents towards the desired path while ensuring the connectivity of the team. More specifically, in order to maintain the connectivity, the leading UAV estimates the followers motion, by employing a fast geometric propagation that exploits the knowledge of the desired formation and sensing capabilities. On the other hand, the followers estimate the motion of the leading UAV by receiving its local state information and implementing a NMPC law that achieves tracking of the desired formation and maintains the connectivity with respect to the leader UAV.