Formation flight control of MAVs (mini aerial vehicles) in non- structured outdoor environments

Abstract

Today, autonomous MAVs (Mini Aerial Vehicles) are well-known as one of the most important topics in the field of mobile robotics. They have interesting applications on search-and-rescue, aerial surveillance, etc. When a mission involves risk of human lives, the importance and usefulness of this kind of vehicle is highlighted, since it does not compromise human lives. In this scenario, this project focus on the development of a full autonomous navigation system that can be applied on a group of MAV that can perform formation flight in outdoor and unstructured environment based on laser sensors, IMU, and GPS data and using visual odometry. Firstly, we will design and implement a dynamical model of the MAV (a mini aerial quadrotor robot Pelican) using the Microsoft Flight Simulator, C#, and a MatLab-Simulink toolbox. We will also use this software to evaluate different flight stabilization controller (for instance: PID, Fuzzy, and Hinf), the autonomous navigator, and the formation flight controller. Next, the routines will be embedded in the real robot in order to evaluate their performance. In the final phase of the project, we will implement and test the formation flight procedures based only in the embedded sensors data and using the three MAVs. (AU)