An Effective Integrity Monitoring Scheme for GNSS/INS/Vision Integration Based on Error State EKF Model

Haitao Jiang,Tuan Li,Dan Song,Chuang Shi
DOI: https://doi.org/10.1109/jsen.2022.3154054
IF: 4.3
2022-04-01
IEEE Sensors Journal
Abstract:It is well-known that the Global Navigation Satellite System/Inertial Navigation System (GNSS/INS) /Vision integration has been increasingly used for safety-critical applications like self-driving cars. However, the performance of the GNSS/INS/Vision integration degrades due to unknown faults. Therefore, it is essential to develop integrity monitoring algorithms for the integrated navigation system. A new protection level (PL) formula to ensure the integrity of the state estimation of the GNSS/INS/Vision integration is proposed, which can be calculated by the relation between the filter estimation error and faults. An integrity risk allocation tree is developed for each sensor fault hypothesis including the nominal hypothesis, and the specific PL is given in the case of GNSS, INS and visual measurement faults. In addition, the performance of the proposed PL is comprehensively evaluated. A field vehicular test was conducted to evaluate the performance of the integrity monitoring algorithm for the GNSS/INS/Vision integration. The results demonstrate that the proposed PL for the GNSS/INS/Vision integration can fit the position error against six different fault modes.
engineering, electrical & electronic,instruments & instrumentation,physics, applied
What problem does this paper attempt to address?