De Lellis, Ettore (2011) Autonomous Approach and Landing Algorithms for Unmanned Aerial Vehicles. [Tesi di dottorato] (Unpublished)
Download (2MB) | Preview
|Item Type:||Tesi di dottorato|
|Uncontrolled Keywords:||UAV, Autolanding, Sensor Fusion|
|Date Deposited:||09 Dec 2011 08:45|
|Last Modified:||30 Apr 2014 19:47|
In recent years, several research activities have been developed in order to increase the autonomy features in Unmanned Aerial Vehicles (UAVs), to substitute human pilots in dangerous missions or simply in order to execute specific tasks more efficiently and cheaply. In particular, a significant research effort has been devoted to achieve high automation in the landing phase, so as to allow the landing of an aircraft without human intervention, also in presence of severe environmental disturbances. The worldwide research community agrees with the opportunity of the dual use of UAVs (for both military and civil purposes), for this reason it is very important to make the UAVs and their autolanding systems compliant with the actual and future rules and with the procedures regarding autonomous flight in ATM (Air Traffic Management) airspace in addition to the typical military aims of minimizing fuel, space or other important parameters during each autonomous task. Developing autolanding systems with a desired level of reliability, accuracy and safety involves an evolution of all the subsystems related to the guide, navigation and control disciplines. The main drawbacks of the autolanding systems available at the state of art concern or the lack of adaptivity of the trajectory generation and tracking to unpredicted external events, such as varied environmental condition and unexpected threats to avoid, or the missed compliance with the guide lines imposed by certification authorities of the proposed technologies used to get the desired above mentioned adaptivity. During his PhD period the author contributed to the development of an autonomous approach and landing system considering all the indispensable functionalities like: mission automation logic, runway data managing, sensor fusion for optimal estimation of vehicle state, trajectory generation and tracking considering optimality criteria, health management algorithms. In particular the system addressed in this thesis is capable to perform a fully adaptive autonomous landing starting from any point of the three dimensional space. The main novel feature of this algorithm is that it generates on line, with a desired updating rate or at a specified event, the nominal trajectory for the aircraft, based on the actual state of the vehicle and on the desired state at touch down point. Main features of the autolanding system based on the implementation of the proposed algorithm are: on line trajectory re-planning in the landing phase, fully autonomy from remote pilot inputs, weakly instrumented landing runway (without ILS availability), ability to land starting from any point in the space and autonomous management of failures and/or adverse atmospheric conditions, decision-making logic evaluation for key-decisions regarding possible execution of altitude recovery manoeuvre based on the Differential GPS integrity signal and compatible with the functionalities made available by the future GNSS system. All the algorithms developed allow reducing computational tractability of trajectory generation and tracking problems so as to be suitable for real time implementation and to still obtain a feasible (for the vehicle) robust and adaptive trajectory for the UAV. All the activities related to the current study have been conducted at CIRA (Italian Aerospace Research Center) in the framework of the aeronautical TECVOL project whose aim is to develop innovative technologies for the autonomous flight. The autolanding system was developed by the TECVOL team and the author’s contribution to it will be outlined in the thesis. Effectiveness of proposed algorithms has been then evaluated in real flight experiments, using the aeronautical flying demonstrator available at CIRA.
Actions (login required)