Development of artificial intelligence logic for the strategic management of multi-UAVs systems
keywords CONTROL, DRONES, FLIGHT MECHANICS, MECHATRONICS
Reference persons GIORGIO GUGLIERI
External reference persons Simone Godio (firstname.lastname@example.org)
Research Groups 08- Dinamica, controllo e simulazione del volo
Thesis type SIMULATION AND DESIGN
Description The objective is to plan and develop an algorithm based on reinforcement learning or imitation learning to control a fleet of UAVs, with a variable number of units (between 2 and 10). We want to optimize the behavior of the fleet in exploring a critical area (with a high density of obstacles), minimizing the consumption, covering the whole area in the shortest time possible, and maintaining a strategic distribution of the aircraft in space. The algorithm must be validated on similar state of the art logics and tested in a realistic ROS/Gazebo simulation environment, through SITL (software in the loop) approach.
Required skills Meccanica del volo (ala fissa e rotante). Controlli automatici.
Notes Si richiede media superiore a 27/30 e debito massimo pari a 2-3 esami. Exams left less than 3, average exams higher than 27.0, and programming skills (C/C++, basic python)
Deadline 30/08/2022 PROPONI LA TUA CANDIDATURA