Unmanned Aerial Vehicles (UAVs) are increasingly gaining traction due to their potential and major use case applications. The UAV is typically required to navigate autonomously in highly dynamic environments to deliver on its intended applications. Existing UAV positioning and navigation solutions face several challenges, particularly in dense outdoor settings. To this end, we present various technological approaches for autonomous UAV navigation outdoors. This chapter aims to provide an efficient real-time autonomous solution that enables the UAV to navigate through a dynamic urban or suburban environment. Particularly, we evaluate the performance of Machine Learning (ML)-based techniques in UAV navigation solutions. The computational complexity involved in standard optimization-based methods hinders its utilization for UAV navigation in dynamic environments. The use of ML-based approaches can potentially enable near-optimal UAV navigation, while providing a practical real-time calculation that is needed in such dynamic applications. We provide a comprehensive detailed analysis to evaluate the performance of each of the presented ML-based UAV navigation methods as compared to other existing navigation approaches that we also discussed in this chapter.
Autonomous UAV Outdoors Navigation—A Machine-Learning Perspective
Navigation: Science and Technology
18.09.2024
27 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Editorial: Visual navigation and mapping outdoors
British Library Online Contents | 2010
|Submersible motor for outdoors
Online Contents | 2010
Learning for Autonomous Navigation
NTRS | 2005
|Learning for autonomous navigation
NTRS | 2005
|