Presentation title

Simulation Framework for Enhanced Train Localization

Authors

Gianluca D'Amico, Federico Nesti, Mauro Marinoni, Giorgio Buttazzo, Gianluigi Lauro and Salvatore Sabina

Institution(s)

Scuola Superiore Sant'Anna, Pisa

Presentation type

Technical presentation

Abstract

Precise and reliable train localization is a difficult as well as crucial task in current railway applications, since it significantly affects the exploitation efficiency of railway resources in terms of number of trains that can run on the railway network.

This presentation proposes a simulation framework to generate different scenarios to test novel localization algorithms and sensor-fusion approaches in different settings and operating conditions.

The proposed methodology integrates different types of sensors, as Lidar, IMU, GNSS and wheel encoders to refine the 3D position of the train (3D localization), enhance the estimation of the covered distance on the route (1D localization) and discriminate the track where the train is running.

The method also uses Lidar odometry and feature detection to further enhance the localization through a map matching algorithm.

---

Additional material

• Presentation slides: [pdf]