Robust Tightly Coupled GNSS/INS Experimental Assessment for Autonomous Aircraft Inspection

Yi Ding, Gaël Pagès, Philippe Asseman, Éric Chaumette

Abstract: The combination of the Global Navigation Satellite System (GNSS), commonly used as economic outdoor navigation solution, and the Inertial Navigation System (INS) is widely developed and has become the baseline for numerous autonomous applications. However, the classical tightly coupled GNSS/INS (TC-GNSS/INS) navigator is very vulnerable to multiple faulty measurements in challenging GNSS conditions, including multipaths, partial sky visibility and non-line-of-sight (NLOS). This article addresses the problem of achieving a robust low-cost TC-GNSS/INS solution of an autonomous robot for the aircraft ground inspection application. In such a challenging navigation scenario, the performances of the classical TC-GNSS/INS can be strongly degraded by contaminated GNSS measurements. In literature we can find a certain number of contributions on the design of estimation algorithms based on robust statistics, which show that the robust TC-GNSS/INS approach is capable of mitigating large positioning errors in simulations or some specific experiments. To the best of our knowledge, for the purpose of aircraft ground inspection by autonomous robot, there is no related evaluation work with experimental data. To fill this lacuna, we propose to assess the use of a robust statistical approach, based on an M-Estimator formulation in a TC-GNSS/INS framework, with inertial and GNSS measurements of a real aircraft ground inspection scenario, where the faulty measurements are no more subject to Gaussian law. The implemented M-estimation-based robust extended Kalman filter (M-EKF) is compared with the standard EKF (S-EKF) and the navigation solution provided by the commercial u-blox M8T GNSS receiver. A series of data analysis on the behaviour of the M-EKF along with the GNSS raw measurements is carried out and discussed in details. The results show that the M-EKF is capable of mitigating the unexpected large errors by applying appropriate weights on the measurements outliers. This work provides a deep analysis on the use of the robust estimator in the case of an autonomous robot for the aircraft ground inspection. We also provide references and perspectives on this subject.
Published in: Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021)
September 20 - 24, 2021
Union Station Hotel
St. Louis, Missouri
Pages: 3452 - 3463
Cite this article: Ding, Yi, Pagès, Gaël, Asseman, Philippe, Chaumette, Éric, "Robust Tightly Coupled GNSS/INS Experimental Assessment for Autonomous Aircraft Inspection," Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021), St. Louis, Missouri, September 2021, pp. 3452-3463.
https://doi.org/10.33012/2021.17881
Full Paper: ION Members/Non-Members: 1 Download Credit
Sign In