A PPP-based Multi-sensor Fusion Positioning Solution for Train Localization

GNSS-based train positioning is one of the most promising solutions for next generation train control systems, e.g. the European Train Control System (ETCS), Chinese Train Control System (CTCS). The key for implementing GNSS-based positioning solution in railways application are the critical RAMS (Reliability, Availability, Maintainability, and Safety) requirements. This paper proposes a real-time Precise Point Positioning (PPP)-based multisensor positioning system, which has property of high availability and safety. Depending on different train operation scenarios, the proposed fusion positioning and integrity check solution are intended to improve the RAMS requirements. The integrity will be always monitored, so that critical RAMS requirements are met seamlessly. The feasible system architectures including real time PPP data transmission and GNSS/Odometer/INS fusion positioning are presented in details. The results of the developed solution indicate that GNSS/Odometer/INS fusion results in better accuracy (0.76 m, 95%) in comparison with standalone INS navigation (15.3 m, 95%). In spite of the fact that Kalman Filter convergence process has an adverse impact on detection, identification and adaptation (DIA) quality control, this paper reveals DIA procedure could be used as an integrity monitoring solution of Kalman Filter-based multi-sensor fusion positioning.