Abstract:This paper discusses the design and software-in-theloop implementation of adaptive formation controllers for fixedwing unmanned aerial vehicles（UAVs） with parametric uncertainty in their structure, namely uncertain mass and inertia. In fact, when aiming at autonomous flight, such parameters cannot assumed to be known as they might vary during the mission（e.g.depending on the payload）. Modeling and autopilot design for such autonomous fixed-wing UAVs are presented. The modeling is implemented in Matlab, while the autopilot is based on ArduPilot, a popular open-source autopilot suite. Specifically, the ArduP ilot functionalities are emulated in Matlab according to the Ardupilot documentation and code, which allows us to perform software-in-the-loop simulations of teams of UAVs embedded with actual autopilot protocols. An overview of realtime path planning, trajectory tracking and formation control resulting from the proposed platform is given. The software-inthe-loop simulations show the capability of achieving different UAV formations while handling uncertain mass and inertia.

A Software-in-the-loop Implementation of Adaptive Formation Control for Fixed-Wing UAVs