Analysis of autonomous UAV navigation methods in GPS dead zones
The article presents a systematic review and comparative analysis of modern methods of autonomous navigation of unmanned aerial vehicles (UAVs) in the absence of a GPS signal. The main methods of autonomous navigation are considered: visual (optical flow, visual odometry, visual SLAM), inertial (IMU...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Zhytomyr Polytechnic State University
2025-07-01
|
| Series: | Технічна інженерія |
| Subjects: | |
| Online Access: | https://ten.ztu.edu.ua/article/view/334764 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The article presents a systematic review and comparative analysis of modern methods of autonomous navigation of unmanned aerial vehicles (UAVs) in the absence of a GPS signal. The main methods of autonomous navigation are considered: visual (optical flow, visual odometry, visual SLAM), inertial (IMU-based), lidar (LiDAR SLAM), radar, radio frequency (UWB, Wi-Fi, BLE), as well as hybrid multi-sensor systems. Particular attention is paid to the characteristics of navigation systems: positioning accuracy, processing latency, requirements for computing resources, reliability in different conditions and energy consumption level. Various application options for each method are considered, and the possibilities of use are also determined taking into account the technical limitations of the UAV. The results of a comparative analysis are presented, which allows systematizing the advantages and disadvantages of individual solutions depending on the type of environment, computing budget and target problem. Separately, modern trends are analyzed, namely the growing role of adaptive systems with sensor fusion and the involvement of machine learning methods. The main challenges that slow down the widespread implementation of the analyzed systems in practical applications are identified, namely the lack of unified testing methods, energy consumption of complex systems, lack of realistic datasets, the problem of integrating algorithms into various control systems. Promising directions for further research are identified, which include the development of energy-efficient navigation solutions, the integration of deep learning, and the creation of specialized hardware. The results obtained are of practical importance for engineers, developers, and researchers working to increase the autonomy of UAVs in the absence of GPS. |
|---|---|
| ISSN: | 2706-5847 2707-9619 |