Evaluation of railway system performance under changing levels of automation using a simulation framework

Guruvayur Venkateswaran, Krishnan (2019). Evaluation of railway system performance under changing levels of automation using a simulation framework. University of Birmingham. Ph.D.

Preview

GuruvayurVenkateswaran2019PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (4MB) | Preview

Abstract

Modern mainline railways are under constant pressure to meet the demands of higher capacity and to improve their punctuality. Railway system designers and operators are increasingly looking to use automation as tool to enable proactive strategies to optimise the timetable, improve the reliability of the infrastructure & rolling stock, to allow for a more dynamic command & control system which can respond to passenger demand and overall to linearize the response behaviour of the system under duress.

In the first part of this thesis, I, the author, will discuss the development of automation over the years and the techniques that have been developed to analyse automation changes in a system. Further to this, I outline the various changes to the railway technology over the last century in brief.

In the second part, I apply the techniques described earlier to design an automation framework to develop a grade of automation for the railway system to meet the demands of improved capacity and performance. Further to this, I develop parallel testable levels of automation using existing railway technology to demonstrate the effectiveness of a framework developed using the methodology discussed before. These levels are then tested on a network topology using micro-simulation to verify if they produce improved capacity and performance.

In the final part, A case study is developed for the network from Kings Cross station to Welwyn Garden on the East Coast Main Line with the traffic dense branch line from Hertford north joining this line. The network is simulated under similar conditions to that adopted for the theoretical network and the results are compared with the previous outcomes.

Results from the above studies have several significant outcomes. Firstly, the methodology developed over the course of this thesis can produce automation levels that are distinct from each other. Secondly, these simulation results show that there is a step change in the performance of the systems when organised into distinct levels of automation. Thirdly, and perhaps the most important conclusion from the studies, I show that automation of a single railway sub-system does not yield beneficial results unless there are complementary solutions produced for the surrounding sub-systems.

In the theoretical phase of the study, the journey time calculations were repeated for 5000 iterations using a Quasi Monte Carlo framework. The results indicate a clear separation between each of the level and stages of automation proposed within the framework. The results from the simulation show that the reduction in journey times between the various levels can be as much as 5%. In the case study, the results were not as distinct but the overall trendlines indicate a reduction in journey times for both intercity and suburban services.

Publications produced during the research period:

• Venkateswaran, K., Nicholson, G., Chen, L. & Pelligrini, P. 2017. D3.3.2 Analysis of European best practices and levels of automation for traffic management under large disruptions In: IFFSTAR (ed.) Capacity for Rail. UIC.

• Venkateswaran, K. G., Nicholson, G. L., Roberts, C. & Stone, R. Impact of Automation on the Capacity of a Mainline Railway: A Preliminary Hypothesis and Methodology. 2015 IEEE 18th International Conference on Intelligent Transportation Systems, pages 2097-2102.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):