|Abstract:||This publication describes the further development of a high accuracy navigation filter especially designed for maritime applications. It is part of the developments of the maritime joint project GALILEOnautic 2, which develops a system for autonomous vessel maneuvering and docking in harbors. The proposed navigation filter is tightly-coupled and bases on quantities of an inertial measurement unit (IMU), is aided by a Doppler velocity log (DVL), and signals of Global Navigation Satellite System (GNSS). Particularly, the navigation filter processes GNSS observables of two systems, the Global Positioning System (GPS) and the European satellite system Galileo. Proper filter initialization is crucial for fast and reliable convergence of the filter states. This paper addresses challenges of maritime navigation filter initialization and the influence of erroneous filter state initialization, e.g. disturbed magnetometer measurements on state estimation. To achieve reliable convergence behavior after initialization, even with heavy erroneous initial state calculation, this paper introduces a new dual constellation double differencing update within a tightly-coupled unscented Kalman filter (UKF). A dual antenna approach is realized, where the UKF predicts the double differenced measurement by means of the known baseline of the two antennas. A structure with two GNSS UKF filter updates is introduced. The first update processes dual frequency corrected, dual constellation pseudo- and deltarange observables from a single antenna only. For the synchronization of the GPS and Galileo observables it utilizes the reliable but still slightly inaccurate Galileo GGTO message. The second update processes single frequency, double differenced pseudoranges from two antennas. Here, the synchronization is performed by another, more accurate new method: The double differencing chooses two reference satellites, one from the GPS- and one from the Galileo observables. This is practicable thanks to the tightly-coupled structure, which makes the filter independent from a minimum set of observables of each system. The navigation filter is evaluated in extensive test drives with a car in Campus Melaten, Aachen, with maritime filter initialization assumptions and methods. Results are compared to those of a dual constellation aided filter with a single antenna and a corresponding extended Kalman filter (EKF). A discussion investigates the influence of the dual antenna double differencing on the state estimation. All configurations are compared to an RTK reference measurement. A significantly improved orientation estimation right after initialization, which is accompanied with a more precise position estimation, is achieved. The dual antenna, dual constellation, double differencing filter achieves an average error of 51 cm, with a standard deviation of 6 cm, whereas the single antenna implementation achieves 63 cm and a standard deviation of 15 cm.|
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:||2186 - 2197|
|Cite this article:||
Gehrt, Jan-Jöran, Nitsch, Maximilian, Abel, Dirk, Zweigel, René, "High Accuracy Navigation Filter with Dual Antenna enabling Double-Differencing with Dual-Constellation," Proceedings of the 32nd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2019), Miami, Florida, September 2019, pp. 2186-2197.
ION Members/Non-Members: 1 Download Credit