Design and Performance Validation of Integrated Navigation System Based on Geometric Range Measurements and GIS Map for
- PDF / 8,612,609 Bytes
- 13 Pages / 594.77 x 793.026 pts Page_size
- 88 Downloads / 176 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
Design and Performance Validation of Integrated Navigation System Based on Geometric Range Measurements and GIS Map for Urban Aerial Navigation Gwangsoo Park, Byungjin Lee, Dong Gyun Kim, Young Jae Lee, and Sangkyung Sung* Abstract: This article proposes an efficient integrated navigation algorithm to secure reliable navigation solutions in urban flight environments where satellite navigation is not available. Also, this study investigates a new sensor deployment and filter configuration that can perform real-time navigation onboard small drones by reducing the complexity and computational burden of the conventional point cloud-based pose estimation. The proposed method first derives the geometric relationship between the ranging vector and the known three-dimensional map, where the number of range sensors and deployment structure is refined. Then, we designed the inertial navigation filter structure combining the derived measurement model and evaluated the estimation performance in a realistic urban flight environment. The validity of the proposed algorithm is verified through both error analysis using a simulator with a high-fidelity model and real navigation error analysis from onboard flight test adjacent to urban buildings. In conclusion, this paper presents a distinctive navigation method from the existing point cloud-based approaches and the performance of real-time three-dimensional navigation with a position error of about 1-2 m in satellite unavailability environment. Keywords: Flight test, GIS map, integrated navigation, mode transition, range sensor, satellite unavailability, urban navigation.
1.
INTRODUCTION
Detection and recognition of their surroundings are essential for unmanned vehicles to operate autonomously in real dynamic environments. Autonomy under sophisticated urban surroundings depends on precise dynamic control considering the environment, and navigation systems combining a variety of complementary sensors have been developed to this end. Above all, the reliability and continuity of navigation are essential for operating an unmanned vehicle in a real flight environment, in which the global navigation satellite system (GNSS) visibility and outage may alternate during the flight periods. To resolve this, integrated systems have been proposed using heterogeneous environmental recognition sensor modalities such as vision, ultrasound, radar, and light detection and ranging (LiDAR) along with existing navigation sensors [1–14]. In general, vision sensors are affordable and can provide abundant information for pose es-
timation, but they are sensitive to illumination, and image processing can be computationally expensive. Although ultrasonic sensors can be miniaturized due to their lightweight and small size, they have a low measurement rate, relatively low resolution, and small sensing coverage. In contrast, laser sensors allow long measuring distances at high measurement rates, but the laser may penetrate windows without detecti
Data Loading...
