Gemma Nicholson

A Multiple Train Trajectory Optimization to Minimize Energy Consumption and Delay

In railway operations, if the journey of a preceding train is disturbed, the service interval between it and the following trains may fall below the minimum line headway distance. If this occurs, train interactions will happen, which will result in extra energy usage, knock-on delays, and penalties for the operators. This paper describes a train trajectory (driving speed curve) optimization study to consider the tradeoff between reductions in train energy usage against increases in delay penalty in a delay situation with a fixed block signaling system. The interactions between trains are considered by recalculating the behavior of the second and subsequent trains based on the performance of all trains in the network, apart from the leading train. A multitrain simulator was developed specifically for the study. Three searching methods, namely, enhanced brute force, ant colony optimization, and genetic algorithm, are implemented in order to find the optimal results quickly and efficiently. The result shows that, by using optimal train trajectories and driving styles, interactions between trains can be reduced, thereby improving performance and reducing the energy required. This also has the effect of improving safety and passenger comfort.

Benchmarking and evaluation of railway operations performance

Abstract A prototype evaluation framework has been used in railway traffic simulator benchmarking and in the quantitative evaluation and comparison of timetables and of real time traffic management decision taking for railway systems in the presence of both small and large-scale service disruption in the EU FP7 project ON-TIME. Quantified key measures allow an assessment of performance and can be used to compare timetables, control methods or delaying incidents. The resilience measures additionally provide a visualisation and information that can be analysed as an aid to the understanding of delay propagation, based on both real and simulated data. This benchmarking and evaluation method is based on collecting data at selected observation points on train and service ID, position, and time from either simulation or from data feeds of real operations. With the addition of data on train and network characteristics, this collected data can be processed in such a way as to evaluate the key measures outlined in a previously defined Quality of Service framework. The key measures quantify the following KPIs of the model: transport volume, journey time, connectivity, punctuality, resilience, energy consumption and resource usage.

Measurement and modelling of ACFM response to multiple RCF cracks in rail and wheels

Abstract This paper presents the experimental and model results of the response of an alternating current field measurement (ACFM) sensor to clusters of rolling contact fatigue (RCF) cracks typical of those found in rails and rail wheels. Both artificial and real cracks occurring in rails taken from service are considered. Currently, commercially available ACFM software is capable of producing an estimate of crack pocket length for isolated cracks, assuming they are regularly shaped. The results presented are part of continuing work to link the ACFM signal to the whole range of complex shaped RCF cracks that appear in rail and rail wheels, including those appearing in clusters. The challenges in accurately sizing clustered RCF cracks using the ACFM technique are discussed.

Impact of Automation on the Capacity of a Mainline Railway: A Preliminary Hypothesis and Methodology

The increase in demand for capacity on railways has presented railway organisations with the impetus to implement automation systems on their networks. Automation has been seen as a key solution for improving the reliability of production in various industries. In the case of railways this would be improving the reliability of service, so that more passengers and goods can be moved, essentially improving the capacity. This paper provides a preliminary hypothesis and methodology for identifying the increase in capacity that comes with increasing the level of automation within a railway. This paper proposes a classification of the various railway technologies, such as ETCS, DAS and ATO, into distinct levels of automation for a mainline railway. Further, contributing factors to delays are identified and a preliminary relationship is defined. A methodology to test the hypothesis is proposed at the end.

Investigation into train positioning systems for saving energy with optimised train trajectories

One approach to reduce energy consumption in railway systems is to implement optimised train trajectories. These are speed profiles that reduce energy consumption without foregoing customer comfort or running times. This is achieved by avoiding unnecessary braking and running at reduced speed whilst maintaining planned arrival times. An optimised train trajectory can be realised using a driver advisory system (DAS). The optimal train trajectory approach needs a variety of input data, such as the trains position, speed, direction, gradient, maximum speed, dwell time, and station locations. Many studies assume the availability of a very accurate train position in real time. However, providing and using high precision positioning data is not always the most cost-effective solution. The aim of this research is to investigate the use of appropriate positioning systems, with regard to their performance and cost specifications, with optimised trajectories. This paper first presents a single train trajectory optimisation to minimise overall energy consumption. It then explores how errors in train position data affect the total consumed energy, with regard to the tractive force due to gradient when following the optimised trajectory. A genetic algorithm is used to optimise the train speed profile. The results from simulation indicate that a basic GPS system for specifying train position is sufficient to save energy via an optimised train trajectory. The authors investigate the effect of error in positioning data, to guarantee the reliability of employing the optimised solution for saving energy whilst maintaining an acceptable journey time.

Determining the propagation angle for non-vertical surface-breaking cracks and its effect on crack sizing using an ACFM sensor

Alternating current field measurement (ACFM) probes are used to detect and size cracks in a range of engineering components. Crack sizing for this, and other electromagnetic (EM) based NDT systems, relies on relating the signal obtained to the actual crack length. For cracks that do not propagate vertically, such as rolling contact fatigue cracks in rails, predicting the crack depth, which determines the rail depth to be removed by grinding, requires an assumed propagation angle into the material as no method to determine crack vertical angle from the EM signals has been reported. This paper discusses the relationship between ACFM signals and propagation angles for surface-breaking cracks using a COMSOL model. The signal accurately predicts the crack pocket length when the vertical angle is 30–90° but underestimates pocket length for shallower angles, e.g. a 50% underestimate is seen for a 3.2 mm pocket length crack propagating at a vertical angle of 10°. A new measure, the trough–peak ratio, is proposed to determine the crack vertical angle. These are verified by experimental measurements using a commercial ACFM pencil probe for cracks with a range of vertical angles between 10° and 90°.

A microscopic evaluation of railway timetable robustness and critical points

One method to increase the quality of railway traffic flow is to construct a more robust timetable in which trains are able both to recover from delays and the delays are prevented from propagating ...

Dynamic train rescheduling using alternating algorithms

The highly uncertain and complex railway operations result in unpredictable operational dynamics to the railway system, such as uncertain delays, changing traffic flows, changes in traffic pattern, etc. In recent years, various algorithms have been developed to help dispatchers in making decisions during railway operation. However, these algorithms are short of the ability of providing solutions to the complicated traffic with operational dynamics. In order to deal with operational dynamics, a methodology using performance-based supervisory control is proposed to provide multiple rescheduling decisions over a wider area through the application of alternating rescheduling algorithms in appropriate sequences. The process of dynamic rescheduling is illustrated and implemented with the BRaVE simulator, and the rescheduling results have been evaluated with propagation of delays with respect to time. A case study on ECML is introduced to demonstrate how this process is realised, and it shows the improvement of delay recovery by applying alternating approaches.

Towards integrated simulation and formal verification of rail yard designs - an experience report based on the UK East Coast Main Line

The development of railway systems is often supported by a range of tools, each addressing individual, but overlapping concerns such as, e.g., performance or safety analysis. However, it is a challenge for users to organise work-flows; results are often in different, non-aligning data formats; furthermore, tools work on different levels of abstraction from macro to microscopic. Thus, tool integration would be beneficial, and also allow for more playful, experimental prototyping and design. This paper reports on lessons learned from the integration of BRaVE - the Birmingham Railway Virtual Environment - and OnTrack from Swansea University. BRaVE is an easy-to-use railway simulation software for development, modelling and flow analysis. OnTrack allows for the automatic verification of scheme plans against a number of safety properties via different formal methods. We present an approach that bridges the gap that occurs from varying details in data sources through automated transformations. This integration provides a first step towards a seamless environment for prototyping, concept development, and safety analysis under ”one roof”. We demonstrate the usefulness of our approach by giving integrated simulation and verification results for the UK East Coast Main Line. This work is part of the wider RSSBs Future Traffic Regulation Optimisation research programme.

Development of a high-level feasibility analysis model for the selection of railway upgrade options

Currently, many railway systems need to be upgraded to meet the demand for rapidly increasing railway capability, while there is a lack of the right models and tools required to support the early decision making stage of railway system upgrade projects. A rail selection upgrade process, combining macroscopic and microscopic simulation, has been developed to find appropriate and systematically selected solutions for railway system upgrade projects. As part of the approach, a high-level feasibility analysis model is developed in this research for determining serious candidate solutions from a large number of possible upgrade options in a quick and effective way. This paper outlines the development process and verification of this models feasibility.