Consensus-based reconfigurable controller design for unmanned aerial vehicle formation flight

Abstract

This article presents a distributed controller design methodology with a failure detection logic for formation flight of unmanned aerial vehicles (UAVs). An output feedback linearization method with a consensus protocol is used to maintain a specified time-varying geometric configuration of multiple UAVs. In this approach, any explicit leader does not exist in the team, and the control strategy requires only the local neighbour-to-neighbour information between the vehicles. A graph Laplacian matrix is used to define the information flow topologies between the vehicles. The Failure detection algorithm is also proposed in order to detect a communication loss or an aircraft system fault and to isolate the failed one. The detection logic proposed in this study considers the case in which one of the vehicles totally broke down. Once the failure is detected, the failed vehicle is isolated. Reforming the interconnection topology using the failure detection logic can guarantee the stable formation flight with the proposed controller. Numerical simulations are performed to validate the performance of the proposed controller with which the formation geometry is dynamically reconfigured using failure detection information.