Consensus-Based Algorithm for Multi-UAV Formation Flight

In recent years, technological advancements have driven increasing interest in autonomous multi-vehicle formations and multi-agent systems (MASs). Specifically, the field of Unmanned Aerial Vehicles (UAVs) has attracted considerable attention and is expected to grow significantly due to its extensive potential applications. UAVs present a viable solution for both military and civilian operations, reducing the risks associated with human presence in hazardous environments. Nevertheless, current systems have not yet achieved full autonomy and require a manned platform to accomplish mission objectives. Within this context, the concept of Manned-Unmanned Teaming (MUM-T) has emerged,which involves the collaboration of unmanned aircraft with manned aircraft to provide support in operational scenarios.This concept plays a central role in this work. This thesis was developed during an internship done at Leonardo company (Aircraft Division), as part of a collaboration between Politecnico di Torino the aforementioned company. The first part of this work focuses on the development of a fixed-wing UAV model, which was implemented in the Matlab/Simulink environment, including the flight control system and autopilots. Subsequently, a formation control algorithm was implemented to coordinate a five-aircraft formation using a leader-follower approach, within the context of manned-unmanned teaming. Specifically, a consensus algorithm was adopted, which aim is to drive the states of all aircraft in the formation to a common (or consensus) value to accomplish a global task. Both standard and novel consensus protocols, involving edge-weighted functions in the communication topology, were integrated to perform various tasks such as formation generation and maintenance, formation shape variation, collision avoidance and obstacle avoidance (including no-fly-zones). To demonstrate the functionality of the developed formation control algorithm, various flight scenarios were simulated and discussed extensively in this work.