|Abstract:||Autonomous driving is playing more and more an important role, not only in automobile but also in maritime applications. Accurate and robust localization of the local system and the participants nearby is one of the key aspects in applying autonomous surface vehicles. The presented publication is part of the developments of the joint project GALILEOnautic 2, which develops a system for autonomous, cooperative maneuvering of networked vessels in harbors and safety critical areas. In this context, the paper on hand proposes a novel cooperative navigation strategy of a connected autonomous vessels system, which ensures an accurate and robust localization of the networked participants and detects the marine environment as basis of autonomous driving in the maritime field. Using environmental sensors, such as light detecting and ranging (LIDAR) and radio detection and ranging (radar), the objects nearby can be detected and clustered. These detections are then associated with the participants information accessed via network communication. Using the methods, such as similarity comparison and modified Dempster-Schaefer theory, the detections from different resources can be aligned with each other, and the environmental detections can then be classified in different categories. Subsequently, the information from networked participants who possess an accurate geodetic position can be used to enhance the robustness and accuracy of the own local positioning by combining relative distance and relative angle between the own vehicle and a networked participant with a known geodetic position. At this point, an extended Kalman filter is implemented: the information of the neighbor participants, e.g. distance and angle to other objects along with local GNSS information, and velocity measurements, are used to update the EKF states. Before a future maritime test, the proposal is validated through a smaller scenario with three experimental vehicles. A possible networked scenario with networked participants and obstacles is conducted. The results showed, that the proposed implementation can successfully detect, identify, and categorize the objects in the environment. Aided with the cooperative information, the navigation filter can achieve a robust positioning with an accuracy of under one meter, even in the case of an unavailability of GNSS measurements.|
Proceedings of the 32nd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2019)
September 16 - 20, 2019
Hyatt Regency Miami
|Pages:||1976 - 1990|
|Cite this article:||
Lin, Jiaying, Gehrt, Jan-Jöran, Zweigel, René, Abel, Dirk, "Cooperative Localization of Networked Multi-agent System," Proceedings of the 32nd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2019), Miami, Florida, September 2019, pp. 1976-1990.
ION Members/Non-Members: 1 Download Credit