Motion Planning for Unmanned Aerial Vehicles

Sammanfattning: This project presents a motion planning algorithm for an unmanned aerial vehicle (UAV).UAVs have become more popular in recent years because of their maneuverability. This has causedthe need for motion planning algorithms to increase. The objective of this project is to develop amotion planning algorithm that generates collision-free high-level trajectories while also consideringthe dynamics of the UAV, input constraints, and optimizing the trajectory based on a cost function interms of total jerk. The presented motion planner uses the Rapidly Exploring Random Tree Star(RRT') algorithm. The RRT' algorithm utilizes motion primitives which are then examined forcollision with the Gilbert–Johnson–Keerthi (GJK) algorithm. The algorithm also considers dynamicobstacles to better simulate the real world. The motion planner is subjected to various tests toanalyze its performance in different scenarios. These tests include examining the impact of thenumber of nodes used in the RRT' algorithm, evaluating its capability in environments with dynamicobstacles, and assessing the ability to track the generated trajectories with a position-based flightcontroller. The motion planner succeeded in all tests and was able to find trajectories in obstacle-cluttered environments, but there are still improvements that can be made which is discussed at theend of this paper.