Footnote:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Description:
Global Navigation Satellite Systems (GNSS) are potentially applicable for various railway applications, especially the safety-related applications such as train localisation for the purpose of train control. In order to integrate GNSS for train localisation, a trustable stand-alone GNSS-based localisation unit should be developed. Then to comply with EN 50126 (reliability, availability, maintainability, and safety; RAMS) standards, the demonstration of GNSS quality of service (QoS) should be evaluated in consistent with RAMS. However there are currently no appropriate performance evaluation methods on GNSS for railway safety-related applications. This dissertation identifies the required performance for train localisation in consideration of GNSS QoS and railway RAMS. The common and different properties of the performance are analysed in detail using consistent attribute hierarchy structures based on UML class diagram. Then formalised performance requirements are proposed quantitatively on four properties (accuracy, reliability, availability, and safety integrity). After that, the evaluation and verification methodologies are introduced. The evaluation methodology is using a reference measurement system for GNSS receiver measured train location accuracy identification, and a stochastic Petri net (SPN) model for GNSS receiver measured train location accuracy categorisation. The SPN model illustrates the GNSS receiver measured train locations into three states (up state, degraded state, and faulty state). Then the four proposed properties are allocated and estimated formally using the three states in the SPN model. The verification methodology is used to verify the GNSS receiver measured train location in real time based on a localisation unit. The GNSS receiver measured train locations are verified using hypothesis testing methods based on the accurate digital track map provided beforehand. Then train location estimation from the localisation unit is verified according to the mileage of the train. With the ...