Sie können Operatoren mit Ihrer Suchanfrage kombinieren, um diese noch präziser einzugrenzen. Klicken Sie auf den Suchoperator, um eine Erklärung seiner Funktionsweise anzuzeigen.
Findet Dokumente, in denen beide Begriffe in beliebiger Reihenfolge innerhalb von maximal n Worten zueinander stehen. Empfehlung: Wählen Sie zwischen 15 und 30 als maximale Wortanzahl (z.B. NEAR(hybrid, antrieb, 20)).
Findet Dokumente, in denen der Begriff in Wortvarianten vorkommt, wobei diese VOR, HINTER oder VOR und HINTER dem Suchbegriff anschließen können (z.B., leichtbau*, *leichtbau, *leichtbau*).
Dieses Kapitel befasst sich mit dem innovativen Einsatz globaler Satellitennavigationssysteme (GNSS) und Multi-Sensor-Daten zur Entwicklung präziser Bodenrichtwerte und digitaler Karten für Eisenbahnanwendungen. Das RAILGAP-Projekt, das von EUSPA im Rahmen des Programms H2020 gefördert wird, zielt darauf ab, die Lücken in digitalen Karten und verlässlichen Bodenrichtwerten für Zugsicherungs- und -überwachungssysteme zu schließen. Der Text diskutiert die Herausforderungen und Möglichkeiten der Nutzung von Satellitensignalen in verschiedenen Eisenbahnumgebungen, wie etwa in städtischen Schluchten und Tunneln. Es hebt die Integration von EGNSS mit anderen Sensoren hervor, um eine kontinuierliche Zugortung und die Entwicklung eines umfassenden digitalen Kartenwerkzeugs sicherzustellen. Das Kapitel skizziert auch die funktionale Architektur und Methodik für die Erstellung eines zuverlässigen Toolsets für Grundwahrheiten und digitale Karten, wobei der Schwerpunkt auf der Verwendung von Algorithmen zur Datenfusion und der Validierung der Ergebnisse liegt. Die Schlussfolgerung unterstreicht das Potenzial dieses Ansatzes, die wirtschaftliche Nachhaltigkeit des ERTMS-Signalsystems zu verbessern und den Weg für fortschrittliche Zugsteuerungstechnologien zu ebnen.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
Satellite-based localization solutions are expected to boost railway digitalization and in particular, they will enhance evolution and efficiency of railway signaling systems. The development of multi-sensor solutions is ongoing, but some gaps remain. This paper addresses two of them: the need for innovative high accuracy and precision Ground Truth and Digital Maps, essential elements of a EGNSS train positioning system and a V&V environment. These two objectives are focused in the RAILGAP EU project. For each of these tools, this paper presents the main high-level requirements and the selected architectural design exploiting specific data fusion algorithms. The novelty of the EGNSS multi-sensor solution proposed is that it does not require to install or modify any equipment on the track. It is based on datasets acquired through commercial runs in Italy and Spain, leveraging on regular train trips in different operational scenarios and time.
1 Introduction
The increasing demand for rail transportation is accelerating the adoption of digitalization technologies with the advantage of new products being developed for automated driving in the automotive sector. Some of these technologies are also foreseen for train control and monitoring systems and for improving the economic sustainability of the ERTMS signaling system, particularly for ETCS level 2, paving the way for the hybrid level 3 up to the full level 3 moving block for which the satellite train positioning is recognized as a game-changing technology. The use of Global Navigation Satellite Systems (GNSS) for train localization added a new tool providing high accuracy, confidence and integrity level, not affected neither by the train speed nor the rail-wheel adhesion. However, the railway environment presents its own challenges and opportunities when it comes to infrastructure and its fundamental elements. Railway tracks are often installed along challenging and harsh environments for the use of satellite signals like urban canyons, tunnels, or forested areas. On the other hand, the tracks constrain the vehicle dynamics to two dimensions. The combination of European GNSS (EGNSS) with other sensors like inertial sensors, cameras and LIDAR allows to compensate some of the sensor drawbacks and can guarantee continuous train localization in any operational scenario together with the availability of digital railway maps. RAILGAP project has been awarded by EUSPA within the H2020 program in order to contribute to the roadmap for their introduction in the railway sector. Two gaps are targeted here: first, the availability of Digital Maps (DM); second, the need for a reliable Ground Truth (GT) that will allow the quantitative evaluation of the new developed solutions and for the validation of Control-Command and Signaling (CCS) systems i.e., ATO, Fail-Safe Train Positioning, Hybrid or full Level 3. Conventional survey techniques are not as efficient as desirable for rail applications in terms of time and cost. Therefore, we propose a novel EGNSS multi-sensor solution and to develop these two tools (DM and GT) without any dependence or modification of trackside equipment. In this project, a “big dataset” collected with different sensors through several runs in Italy and Spain is analyzed and processed to yield the mapping information as well as ground truth data, leveraging on regular train trips in different operational scenarios and time.
This paper presents the main requirements and the selected architectural design to build a reliable, robust and accurate GT and a DM toolset exploiting specific data fusion algorithms. The paper also discusses the methodology, integration of processes, sensors and systems to ensure the validity, scalability of the GT and DM toolsets to contribute to the standardization process for the adoption of GNSS positioning in the ERTMS standard.
Anzeige
2 Ground Truth
The concept of “Ground Truth” is a well-established principle in cartography, where remotely collected data is validated by in-situ measurements. These measurements can be used, for instance, to calibrate remote sensing devices (e.g., satellite sensors), verify or correct experimental inferences, and update geographic databases. Cartographic methods have significantly improved thanks to GNSS-based positioning methods, interoperable data standards for the rapid exchange of highly interrelated and accurate information and devices and interfaces that visually deliver information on demand for different user needs.
2.1 Ground Truth Definition
In RAILGAP, the Ground Truth is defined as a set of georeferenced data with known accuracy, availability and reliability, built by means of a well-described process, to be considered (a) a stable and true reference, suitable for the purpose of comparison and validation of other data sources in the railway domain according to established requirements and (b) the basis for developing other railway components such a high accurate Digital Map [1]. In practice, this concept translates into a set of variables that, for each epoch, determines the position and dynamics of the train, like:
Timestamp (UTC): Instant of time (k epoch) in the time scale to which all other GT variables are referred – usually provided by the GNSS receiver
Absolute position (WGS84): Location of the vehicle, with respect to the Earth, at the time specified by Timestamp – determined by GNSS/INS measurements
Speed: derived from the GNSS/INS measurements
3D Acceleration (with respect to the Body-frame): instantaneous 3D acceleration – derived from GNSS/INS measurements
Attitude (with respect to the Body-frame) – derived from GNSS/INS measurements
2.2 High Level Requirements
In RAILGAP project, a specific attention has been devoted to the analysis and definition of the user requirements for the “Ground Truth as a “product” concept, at functional and non-functional levels. They are aligned to the following issues:
Handling and pre-processing of data from different sources (37% of the requirements)
Computational aspects related to the derivation of the Ground Truth (47% of the requirements)
Output data management and outcomes visualization (8% of the requirements)
User interfaces (8% of the requirements).
These requirements have been then further detailed up to a lower level corresponding to the toolset development, considering also the algorithmic recommendations and the outcomes from the experimentation performed within the project.
Anzeige
2.3 Functional Architecture
The functional architecture of the Ground Truth is presented in Fig. 1. Particularly, this diagram provides the relation between the GT logical architecture and the GT toolset.
Input data. A set of sensor devices is integrated into an On-board Unit (OBU), deployed on board of each train and properly configured as per the analysis performed during the project. This OBU will record all the observables (input data) required for the Ground Truth Computation.
Data Fusion and FDE (Faut Detection & Exclusion) Algorithm. The input data are then validated and processed in order to detect the presence of outliers and faulty data and eventually exclude them from further processing. To obtain a combined full solution of a potential Ground Truth, a data fusion is then performed.
Ground Truth Construction. The fused data set is then processed by the Ground Truth Computation Module. The GT computation may be supported by additional sensors or technologies such as cameras, or LIDAR for certain functionalities, such as track discrimination. After this computation, a set of checks are performed to validate the results.
Output management and Database construction. After this computation, a set of checks are performed on the obtained values for their validation and computed data are stored in a dedicated database, which will be accessible to the user by means of the corresponding interface.
User interface. The User Interface is in charge of handling input data and configuration settings defined by the user according to the different use cases, and presenting the output data in a kindly and understandable manner.
The “Digital Map” as a concept is related to the description of the railway network infrastructure in terms of the following information [2]:
Topology: the track network as described as a topological node edge model, graph based [3].
Geometry: it connects the objects of the topology to the physical word.
Railway infrastructure elements: a variety of railway signaling relevant assets that can be found on, under, over or next to the railway track, e.g., balises, platform edges.
Immaterial objects: encompass features that are closely linked with the railway infrastructure (e.g. speed limits, gradient, radio coverage area…).
3.1 Trackside Digital Map
In RAILGAP project, the Trackside Digital Map (TDM) tool is a comprehensive digital mapping tool designed for the needs of railway companies, operators and maintenance personnel. It provides an up-to-date and detailed representation of the entire railway network, including track layouts, infrastructure, and signaling systems. The map displays information such as track configurations, railway signals, switch points, and grade crossings, making it easy to understand the location and status of various components of the railway system. In addition, the TDM could provide information on rail yards, maintenance facilities, and other critical infrastructure components, helping to improve overall operational efficiency and reduce downtime and also supporting maintenance activities such as inspecting tracks, identifying potential problems, and planning track upgrades. Overall, the TDM provides a powerful and intuitive tool for optimizing railway operations and maintenance ensuring the efficient and reliable operation of the rail network.
3.2 High Level Requirements
User requirements for the TDM have been set in the RAILGAP project, at functional and non-functional levels. Particularly, TDM shall be able to:
Process time-synchronized data coming from different sensors based on GT, imaging and ranging sensors (i.e., cameras and LIDAR) acquired through commercial trains equipped with the RAILGAP measurement subsystem.
Detect, classify and georeference typical signaling object components of the TDM.
Build the TDM in standardized format and semiautomatic way.
Perform continuous monitoring, control and automatic update of the TDM when significant changes occur.
Scale the TDM structure allowing aggregation of additional information layers and the connection of different sections of the railways network (national or European).
3.3 Functional Architecture
The TDM functional architecture is presented in Fig. 2. Sensors block provides raw measurements from IMU, GNSS, stereo camera and LIDAR sensors, acquired by the On-Board Measurement block and made available to the TDM by the Trackside Data Collector block for off-line processing. Fault Detection and Exclusion block is responsible for detecting faults in the sensors data, and if necessary, correcting or excluding them by appropriately configuring the processing chain.
Analysis of confidence boundaries block provides the computed confidence boundaries associated with the Train absolute 3D positioning and Object absolute 3D positioning blocks. Train absolute 3D positioning block processes IMU and GNSS raw data to provide accurate 3D position of the train [4]. Railway object detection and classification block consists of a set of techniques for object detection and classification based on AI techniques for image, video, and point cloud analysis [5]. 3D Object relative positioning respect to body frame block computes the position of the detected objects with respect to the train itself as both resulting bounding boxes from camera and LIDAR are expressed in the local sensor coordinates. 3D Object absolute positioning respect to global frame block takes as inputs the detections provided by the Object relative 3D positioning block and the output of the Train absolute 3D positioning block to compute the 3D position of the detected objects in the global reference frame. Digital Map block is the database containing the DM, the toolsets and algorithms responsible for building it based on the coordinates (and any other data) provided by the Object absolute 3D positioning block. Topology and geometry information of the track network and relevant railway infrastructure elements data, are stored in the DM. The DM is also used by the Object absolute 3D positioning block to retrieve and update objects coordinates and associated information in a certain area.
4 Conclusion
The paper provided an overview of the main goals of the EU project RAILGAP addressing two gaps still hindering the adoption of satellite-based systems in railway applications: the lack of high-quality Ground Truth data and the need for a modernized process for mapping existing train tracks cost-effectively, by deriving mapping information directly from trains in commercial operation. GT and TDM toolsets were described in terms of architectural design, methodology and integration of processes, sensors and systems. The outcomes will enable obtaining high accuracy and reliable reference data and characterizing even the most challenging railway environments by exploiting the huge amount of data collected from GNSS receivers and other sensors such as IMU, LIDAR and camera, without operational overheads, at minimal cost in hardware while removing any need for trackside infrastructure. Digital Map should be part of the characteristics of any line. At near future, after having properly defined a procedure, the Infrastructure Managers (as RFI and Adif) should provide such information (and such a service) to any operator that will run on a line equipped with HL3 or moving block.
Acknowledgements
The RAILGAP project is funded by the H2020 EUSPA with GA N. 101004129.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
RAILGAP, deliverable D2.1. User Requirements Document related to Digital Map and Reference Measurement Values & Procedures. website https://railgap.eu/
UIC. IRS 30100 RailTopoModel – Railway infrastrucure topological model (2016)
4.
Brizzi, M., Neri, A.: Multi-ride fusion for rail digital map construction. In: 2023 IEEE/ION Position, Location and Navigation Symposium (PLANS), Monterey, CA, USA, pp. 740–747 (2023)
5.
Vilgertshofer, S., Stoitchkov, D., Borrmann, A., Menter, A., Genc, C.: Recognising railway infrastructure elements in videos and drawings using neural networks. In: 2020 Proceedings of the Institution of Civil Engineers - Smart Infrastructure and Construction, vol. 172, pp. 1–16
