Runtime efficient path planning for a fixed-wing UAV considering flight performance limits

Abstract

In this thesis, a runtime efficient roadmap-based path planning approach for fixed-wing unmanned aerial vehicles (UAVs) in a static 3D-environment is presented. The approach considers kinematic constraints and trimmed flight performance limits of a fixed-wing UAV as well as local prevailing wind. The planner is capable to plan with multiple air-speeds. For this purpose, a 3D-roadmap representing collision free flight corridors is created. Based on the 3D-roadmap, a pseudo-dual graph, in which the path is planned, is created to introduce the aircraft specific constraints. The planned path is a progression of straight flight and turning flight segments. The time parametrized trajectories are super positioned of horizontal and vertical aircraft motion trajectories, described by trochoids in horizontal and 5th degree polynomials in vertical direction. The approach is evaluated in simple and real world scenarios.