Unmanned Cargo Operation: Decision Making for Collision Avoidance Based on the Rules of the Air

Abstract

[EN] The integration of a cargo UAV into the non-segregated airspace is studied in this thesis in terms of its normal and emergency operating regimes. Several algorithms have been developed in order to ensure that the Rules of the Air concerning the cruise flight phase are followed. The UAV mission consists of flying over a path generated between a given set of waypoints and avoiding collision with intruders respecting the Rules of the Air. Path planning and path-following algorithms are based on simple approaches which result in guidance command performance limitations. In particular, Dubins curves approach has been used to obtain the flyable paths, hence discontinuous guidance commands are demanded. Path-following approach is based on Pure Pursuit guidance, therefore the guidance commands are generated by this algorithm and affected by the path planning phase. A decision making algorithm has been developed to manage the UAV operations. This decision making algorithm requires information about the number of intruders, hence a conflict detection algorithm has been programmed as well. This together with the type of intruder, which is determined based on its velocity, determines the required manoeuvre to ensure the separation with the intruders while flying along the given waypoints. Simulations have shown that the safety distance is maintained for distinct types of intruders and different types of approaches in the case of single conflicts. Multiple conflicts resolution can be solved, but an improvement of this approach should be developed