Motion Planning of Unmanned Aerial Vehicle Based on Rapid-exploration Random Tree Algorithm

Abstract

In the study of route planning problems in complex environments, in order to reduce the flight cost of unmanned aerial vehicles (UAVs), it is necessary to achieve a better balance between planning time and path quality. This paper utilizes the Rapid-exploration Random Tree (RRT) algorithm for motion planning of a fixed-wing UAV and a multi-rotor UAV (i.e., a quad-rotor UAV), and gives the origin and destination locations on a 3-D map. By following aerodynamic constraints such as maximum roll angle, flight path angle, and airspeed, a collision-free and flight-friendly path is found through simulation. In addition, this paper employs a path smoothing algorithm to simplify the 3-D Dubins path and generates the shortest trajectory. The simulation results show that the RRT and the optimization method have faster convergence speed and shorter search time, reduce redundant planning points, shorten the planning track, and improve the track planning efficiency.