Snapshot residual and Kalman Filter based fault detection and exclusion schemes for robust railway navigation

Abstract

Integrating satellite based navigation into the railway standard can enable reliable and cost-efficient railway navigation everywhere. This makes is very attractive for railway. Thus its integration is strongly supported within the European railway evolution program. However, railway environments exhibit many challenges. Local threats are major issues for robust GNSS based railway navigation. They cannot be observed by any augmentation methods and can cause hazardous misleading information. Hence, they form an integrity risk, which needs to be detected and mitigated by the onboard system. We analyze three different approaches suitable for railway: two snapshot approaches exploiting track constraints during or after the GNSS position determination, and a sequential approach using an Extended Kalman Filter. We derive global fault detection and exclusion (FDE) schemes for all three positioning methods. We measure their performance in terms of along track position accuracy and position uncertainty. Additionally, we investigate each scheme's FDE quality in detail and clearly show that the innovation based FDE of the extended Kalman filter has the best performance in terms of along track position, fault detection capability and exclusion gain. All investigations are done via Monte-Carlo simulations. The considered scenario was extracted from data collected during a measurement campaign in Brunswick, Germany.