Using information engineering to understand the impact of train positioning uncertainties on railway subsystems

Many studies propose new advanced railway subsystems, such as Driver Advisory System (DAS), Automatic Door Operation (ADO) and Traffic Management System (TMS), designed to improve the overall performance of current railway systems. Real time train positioning information is one of the key pieces of input data for most of these new subsystems. Many studies presenting and examining the effectiveness of such subsystems assume the availability of very accurate train positioning data in real time. However, providing and using high accuracy positioning data may not always be the most cost-effective solution, nor is it always available. The accuracy of train position information is varied, based on the technological complexity of the positioning systems and the methods that are used. In reality, different subsystems, henceforth referred to as ‘applications’, need different minimum resolutions of train positioning data to work effectively, and uncertainty or inaccuracy in this data may reduce the effectiveness of the new applications. However, the trade-off between the accuracy of the positioning data and the required effectiveness of the proposed applications is so far not clear. A framework for assessing the impact of uncertainties in train positions against application performance has been developed. The required performance of the application is assessed based on the characteristics of the railway system, consisting of the infrastructure, rolling stock and operational data. The uncertainty in the train positioning data is considered based on the characteristics of the positioning system. The framework is applied to determine the impact of the positioning uncertainty on the application’s outcome. So, in that way, the desired position resolution associated with acceptable application performance can be characterised. In this thesis, the framework described above is implemented for DAS and TMS applications to understand the influence of positioning uncertainty on their fundamental functions compared to base case with high accuracy (actual position). A DAS system is modelled and implemented with uncertainty characteristic of a Global Navigation Satellite System (GNSS). The train energy consumption and journey time are used as performance measures to evaluate the impact of these uncertainties compared to a base case. A TMS is modelled and implemented with the uncertainties of an on-board low-cost low-accuracy positioning system. The impact of positioning uncertainty on the modelled TMS is evaluated in terms of arrival punctuality for different levels of capacity consumption. The implementation of the framework for DAS and TMS applications determines the following: • which of the application functions are influenced by positioning uncertainty; • how positioning uncertainty influences the application output variables; • how the impact of positioning uncertainties can be identified, through the application output variables, whilst considering the impact of other railway uncertainties; • what is the impact of the underperforming application, due to positioning uncertainty, on the whole railway system in terms of energy, punctuality and capacity