GNSS-based high integrity positioning systems for railway applications

Abstract

The signaling system is one of the most important component of the railway world. In fact, its main tasks are the mitigation of issues due to human error and the allocation of the railway resource [1]. In Europe, to answer such a need, the standard ERTMS/ETCS (European Railway Traffic Management System/European Train Control System) has been deployed [2]. In the ETCS framework, three main levels, distinguishable among themselves in terms of system automation, can be individuated: L1, L2 and L3. A forth situation referred as ETCS Level 0 arises when an ETCS compliant train is moving on a non ETCS line. Up to now, only L1 and L2 are in operation while L3 has been only conceptualized. The main difference between L1 and L2 is in the communication link between the train control system and the train fleet. Concerning the traffic management, the line is divided in sections called blocks with the assumption that one block may be occupied by no more than one train at once. Both L1 and L2 rely on the “fixed block” approach, a strategy that foresees blocks of fixed length. To determine if a block is occupied by a train, the system makes use of track circuits. The main novelty that will be introduced by L3 is a more efficient traffic management strategy called “moving block”. This approach foresees the dynamical tuning of the block lengths. This work of thesis deals with the study on how ERTMS can benefit from satellite navigation. Thank to this technology, it is possible to reduce maintenance and operational costs. The overall system architecture foresees three segments: 1. Space segment 2. Ground segment 3. User segment The space segment includes all the GNSS and the SBAS satellites. The Ground segment is constituted by the augmentation network and the User segment is represented by the train fleet. This work of thesis foresees the study of theoretical aspects related to satellite navigation as well as the analysis of typical issues related to railway environment and applications. The theoretical track has been mostly focused on the ground segment of the train control system and, particularly, on augmentation and integrity monitoring. In fact, due to the stringent requirements that must be fulfilled, a standalone positioning could be not enough to guarantee the required performance. The augmentation and the integrity monitoring are therefore key modules of the system. During this research period, network architectures and integrity monitoring techniques have been studied. The study of typical issues related to the railway field has been mostly related to the user segment. The first use of the GNSS based train control system is the determination of the train position along the line (progressive mileage). This operation, according to [1], can be done explicitly accounting for the fact that the train is constrained to lie on a line (i.e. the track). In presence of multiple tracks, the track discrimination process can be separated from the mileage estimation process. In this framework, the main focus of the research is the study of techniques to determine the track occupied by the train. Another important aspect described in this thesis is the train integrity assessment, an important issue that arises considering the introduction of ETCS L3. Particularly, if for any reason a portion of a convoy is decoupled in a line equipped with the fixed block approach, the track circuits will continue to detect the presence of rolling stokes on the block, labelling it as occupied. In this way, the train that comes after will not enter in the block and there will be no collision. If the same situation happened in a line equipped with the moving block approach, there would be no track circuits to detect the parted section of the train. For this reason, the dynamical tuning of the block length foresees that the train is able to assess its own integrity. The study of the solution for the train integrity assessment has been carried out for both single constellation and multi-constellation framework. The presence of a multi-constellation environment has been considered for both coherent and incoherent integration, highlighting the performance achievable with each technique. One of the key problems of the railway environment is the presence of areas where the GNSS signals are either not present or too degraded to be used to provide a reliable solution. To operate in such areas, alternative system must be used. In the final section of this thesis, it is depicted a theoretical background on how it is possible to exploit external sensors like inertial units to provide system continuity.