Journal Article (Refereed)
A fault tolerance approach for railway scheduling and train control
Ursani, Z & Mei, T X & Whiteing, A 2013, 'A fault tolerance approach for railway scheduling and train control', Transportation Research Part B: Methodological, 56, pp.161-173.
This paper makes two major contributions. It firstly proposes the use of a fault tolerance approach for railway operations and secondly it develops a minimum time gap matrix model for capacity computation and the study of perturbation effects through the generation of a compressed timetable. A fault tolerance approach is proposed to improve the operational efficiency of the railway network in terms of the network capacity and the robustness of train timetables. The term fault tolerance is used in a broad sense, to represent any abnor- malities or unexpected events in operations or equipment. Enhanced fault tolerance capa- bility provides safety assurance so that, in normal operating conditions, trains can adopt much faster speed profiles when approaching a ‘to-be-cleared’ signal block at stations and junctions than those currently permitted, effectively turning the status of ‘be ready to stop’ to that of ‘proceed with caution’. In the rare event of a ‘fault’ in the system, e.g. if a conflicting train fails to move out of a signalling block as expected or a switch fails to operate as required, the train would be re-routed to take an alternative path. In this study, the new approach is developed on three scenarios i.e., a standard classic right turn junction, a terminus station, and a small network combining both of these elements to demonstrate the performance gains, but the concept can be readily extended for other types of junctions/stations. Results so far show great potential in the proposed fault toler- ance approach to increase the capacity and enhance operational robustness to perturba- tions at such locations. A novel method for capacity computation called minimum time gap matrix model is also introduced that has capability to produce compressed timetables directly from a given train sequence.
This paper makes a major contribution to the field of rail transport, as it has established the theoretical basis and also set the theme for the development of a radically novel approach for the train scheduling and optimisation that can provide a significant improvement on network capacility and operation resilience. The reserach project has been invited to present the work at a number of events organised by rail industry (e.g. RSSB and IRSE seminars) and part of the work has been invited for funding application to carry out further research.
Transportation Research Part B: Methodological