Environmental Feature Enabled Digital Track Map for GNSS-based Train Localization Safety Boundary Estimation

Abstract

Railway signalling systems requires the perception of train location for train operation and control. Mainstream train control systems as CTCS-3 (Chinese Train Control System, CTCS) and ETCS-2 (European Train Control System, ETCS) invest heavily on Balise installed along the railway track to provide discrete train location alignment. The accuracy of the train location perception relies on the density of the Balise allocations. Modern train control systems as CTCS-N (Chinse Train Control System - Next Generation) and NGTC (Next Generation of Train Control) seeks the adoption of GNSS and other trainborne localization sensors to provide accurate and continuous train locations, thus providing better location service for train control and other purposes for the train. Train location determination through GNSS and DTM (digital track map), migrating train localization method from track side to trainborne side, minimizes the Balise amount installed on the track and brings flexibility and continuity for train locations. However, GNSS localization performance is affected by the signal propagation environment. The GNSS signal quality received at the antenna suffers from multipath, shadowing, blockage and other affects. The GNSS signal in space (SIS) performance only indicates the performance of the signal from the satellite, the signal in ground environment (SIE) quality needs to be investigated to ensure the performance for railway train control purposes. Thanks to the characteristics of the railway train operation, the locomotives travel along the railway track according to the timetable, which means the locomotive movement is constrained by space and time. The pre-surveyed digital track map enables possibilities for improving the GNSS localization performance and further safety margin of the location estimation. This paper presents the safety boundary estimation method for GNSS-based train localization. The digital track map includes track information, POI (point of interest, including switch, track boundaries) and the pre-categorized environmental features (open area, half-sky, cutting, urban-canyon, tunnel). The environmental feature/track/POI information are surveyed, validated to ensure the DTM integrity. The multi-hypothesis considers the environmental feature to estimate the horizontal protection level using the acceptable integrity risk as the foundation. The data collected along the Beijing-Shenyang High Speed railway line is analyzed. In both open sky and cutting environment scenarios, the accuracy and safety margin estimated for each GNSS location is calculated. The safety boundary estimation includes location confidence interval, the result showed that the safety boundary estimations are consistent to the error propagation.