ABSTRACT. Disruption management applying disruption programs helps to amend disruption operations. Pre-planned train dispatching instructions for a specific disruption situation facilitate the work of the dispatchers. Those instructions are mostly manually designed and often focus solely on the train runs. The proposed approach aims to improve the quality of disruption programs concerning operations and especially concerning the reduced passenger mobility. For this purpose, the algorithms to be presented evaluate the operating concept on its functionality and reachability in a solely train operations focused way. A stable and fast transitioning disruption program is already enhancing the passenger mobility in a disruption, but it is not enough to call it passenger-friendly. For this purpose, the algorithms design a transportation concept including passenger guidance measures and comprise a final evaluation of the disruption program in a passenger-oriented way.

ABSTRACT. When a train disruption happens, hundreds of train dispatchers need to quickly make a rescheduling timetable plan for train service recovery by taking account of various constraints. Complexity of rescheduled timetable for a large-scale train service disruption changes significantly depending on rescheduling policy like turn around operation because it requires additional train arrangements. Then, dispatchers should decide rescheduling policy first of all to make their timetable rescheduling plan. These dicision making are quite important and rescheduling operations depends on dispatcher’s individual skills currently. One of the most important issues is rescheduling techniques are not inherited among dispatchers effectively. Therefore we propose a new data-driven approach to support dispatcher’s dicision based on historical train timetable data. We analyzed the past timetable data of railway service line in the Tokyo Metropolitan area and automatically extracted primary rescheduling components such as train cancellation. We have verified accuracy of our method regarding turn around operation by comparing with dispatcher’s manual report. The result indicates estimation accuracy was more than 90%. In addition, we have developed the prototype system which suggests rescheduling policy to dispatchers.

ABSTRACT. This paper focuses on the real-time train platforming and routing problem at a busy complex high-speed railway station in a disrupted situation caused by malfunctioning railway infrastructure or train primary delay. When the disruption occurs, dispatchers need to reassign trains to the conflict-free platform and route, even reschedule the arrival and departure times. To this end, we develop a mixed-integer linear programming formulation that determines platforming and routing decision simultaneously, while allowing trains to be rescheduled when the initial schedule imposes irreconcilable conflicts. The objective of our model is to minimize the overall deviation from the planned schedule and the original platforming plans. To improve the solving efficiency, an iterative algorithm is proposed to compute near-optimal solutions in a short computation time, which is based on the decomposition of the overall problem into two sub-problems: (i) platform and route assignment with fixed arrival and departure times, (ii) partial conflict trains rescheduling. The connecting information between two sub-problems concerns the index of conflict trains and the new train timetable. To solve sub-problem (i) efficiently, we develop a branch and bound algorithm which includes implicational rules enabling to speed up the computation and still can acquire optimal solutions. Since the model of sub-problem (ii) is the same as the model of original problem but has a relative small scale, it can be efficiently solved by CPLEX solver. A real-world instance with operation data of Zhengzhou East high-speed railway station, is implemented to demonstrate the performance and effectiveness of the proposed algorithm.

ABSTRACT. Train timetables are built such that trains can drive without any delay. However, in real-time, unexpected events such as overcrowded platforms or small mechanical defects can cause conflicts, i.e., two trains requiring the same part of the infrastructure at the same time. Currently, such conflicts are typically resolved by experienced dispatchers. However, it is impossible for them to fully anticipate the impact of their actions on the entire network. Conflict detection and prevention tools embedded in a Traffic Management System can help them in making informed decisions. Though some advanced train movement prediction and conflict detection has been developed in the last years, there still exists a need for conflict prevention strategies capable of delivering conflict resolutions on large and complex networks based on retiming, reordering and rerouting some of the trains in real-time. Our previous work introduced such a conflict prevention strategy that, based on offline calculations, determined which part of the network should be regarded when deciding on a conflict resolution. This work is significantly extended here by considering several new parameters for the Dynamic Impact Zone heuristic. This paper compares results on different sizes of networks, and tackles the challenges for applying the strategy on even larger networks.

ABSTRACT. The onboard collection of data related to train operation enables a better calibration of the current train motion models, which are fundamental for the elaboration of optimized train control solutions. Here, the possibility to implement an online calibration of train motion models, i.e. to set the model’s parameters for the single train on the go, is explored. For this purpose, a comparison of different calibration models is proposed. Then, the performances of the models are evaluated according to the requirements for online elaborations. In the end, possible further requirements and limitations on the use are discussed.

ABSTRACT. Since years and decades, IT systems are used to plan, to monitor and to control train operations and railway traffic on network regions. Especially in technically advanced railway networks, the usage of computer based systems for dispatching and controlling traffic started quite early, e.g. large innovation programs in the 90s like Bahn2000 in Switzerland or BZ2000 in Germany and more. One component of these systems is a database component managing infrastructure data and topology information. Infrastructure databases usually are dedicated to regions of responsibilities, e.g. “Niederlassung” or “Regionalbereich”, managing infrastructure data for the corresponding region and bordering areas, e.g. lines and tracks from the region leading to surrounding regions. In this way, several, similar databases exists with partially overlapping data. With new IT systems, larger systems covering more regions are possible. Following an evolving (instead of an revolution) approach processes to merge and consolidate existing data and systems are required. This talk presents an approach for migrating legacy databases and data structures from different, overlapping regions and for performing consistency checks with respect to topology, connectivity and infrastructure content and semantical matching of regional data. The talk will present algorithms, heuristics and processes to merge and analyse consistency of consolidated infrastructure databases as developed and evaluated within real-world projects.

ABSTRACT. Simulation is one of the powerful means within the toolset of railway operations research. In contrast to pure timetabling and to queuing theory, it supports a precise representation of interdependencies. There is long-lasting series of research and only some exemplary publication can be listed. Some microscopic simulation tools provide an explicit conflict detection and solution a synchronous and/or asynchronous manner. Recently optimisation components are applied within conflict solution, too. While either the simulation algorithm or the simulation evaluation is addressed within a variety of publications, the execution of studies relies on an important interim step: aggregation, validation and interpretation of simulation (raw) data. Barely no literature can be found. This fundamental task is subject of this paper. A -- mostly -- standardised principle how to eliminate modelling errors and ensure comparability of results is introduced. The principle is based on experience gathered in various consultancy activities.

ABSTRACT. You might be aware of the following gap: There are by far more publications on promising projects on how mathematical optimization could improve the performance of railway companies, than true success stories in the sense that operations research methods really entered the practice of railways. In this paper, we shed a bit of light on those projects, which finally did not enter the practice of railways. We do so by conducting a survey in which we ask both, railway practitioners who served as ordering party, and optimization experts who served as R&D solution provider. By summarizing the most frequent replies to our question about the key factors why in the past mathematical optimization methods did not enter the practice of railways, we offer this knowledge base to future project managers. Acting accordingly with respect to definition of project goals, project design, and project management, hopefully lets them come up with even more true success stories of operations research methods in the practice of railways

ABSTRACT. In order to enable widespread use of data driven analysis and machine learning methods for rail operations problems, large volumes of operational data are needed. This data has the potential to contain erroneous or missing values, especially given its size and dimensionality. In this work a data reconciliation problem for rail dispatch data is proposed to identify and correct errors, as well as to impute missing data. The data reconciliation problem finds the least-perturbed modification of the historical data that satisfies operational constraints, such as feasibility of meet and overtake events, safety headway, siding allocation, and running time. It also imputes missing values with estimates that satisfy all operational constraints. The data reconciliation method is applied to a large historical dataset from freight rail territory in Tennessee, United States, containing over 3,000 train records over six months. The method identifies and corrects errors in the historical data, and is able to impute data on a synthetically decimated version of the historical data. The quality of the imputed data from data reconciliation is compared to imputed data using naive interpolation. The results show that data reconciliation reduces timing error of imputed points by up to 15% and increases the number of meet and overtake events estimated at the correct historical location from less than 40% to approximately 95%. These findings indicate that the data reconciliation method is a useful preprocessing step for analysis and modeling of railroad operations that are based on real-world physical dispatching data.

ABSTRACT. Delay prediction is an important issue associated with train timetabling and dispatching. Based on real-world operation records, accurate estimation of delays is of immense significance in train operation and decisions of dispatchers. In the study, we established a model that illustrates the interaction between train actual delays and the factors affecting the same via train operation records from a Dutch railway system. Based on the main factors that affect train delay and the time series trend, we identified the independent and dependent variables. A long short-term memory (LSTM) prediction model in which the actual delay time corresponded to the dependent variable was established via Python. Finally, the prediction accuracy of the random forest model and artificial neural network model was compared. The results indicated that the LSTM model outperformed other models.

ABSTRACT. This paper presents a combined method of fuzzy theory and rough sets theory for the early warning of high-speed railway (HSR) under adverse weather conditions. Based on the monitoring data of meteorological indicators along the railway, a fuzzy c-means (FCM) clustering is first applied in order to figure out the fuzzy distribution of sample data and to fit the corresponding membership function of every indicator. According to the clustering results, every original sample is transformed into its cluster level as string data for the subsequent application of rough sets theory. Then a series of effective rough rules between conditional indicators and the decision indicator is extracted after attribute reduction by the Rosetta toolkit, where the decision indicator is represented by the train deceleration rate. Since the value of an indicator may correspond to several fuzzy levels, the multiple combinations of different conditional indicators will activate multiple rough rules. In order to forecast a clear value of the decision indicator, a centroid-based Max-Min compound arithmetic is applied to clarify relevant rules and determine the warning level. Using the designed algorithm, a case analysis of an HSR line section is conducted to verify the feasibility of the combined method, all meteorological data and operation records are provided by the Shanghai Railway Bureau in China. The results prove that the hybrid algorithm can be applied in the real-time forewarning of HSR train operation, with a global accuracy over 86%.

ABSTRACT. To reduce passenger congestion during morning rush hour, railway companies in the Tokyo metropolitan area have increased the number of trains. On the other hand, once a train exceeds a dwell time due to sudden events such as passengers rushing onto a train, passengers agglomerating in specific cars and doors, objects getting caught in doors etc., delays propagate to subsequent trains quickly. To evaluate daily train transport stability and countermeasures against train delays, a train travel time simulation model is needed. However, it has been difficult so far to replicate the occurrence of sudden events and the fluctuations in passenger demand. In this paper, we use detailed data based on dwell time structure and on-site inspections to construct a train travel time simulator. In addition, we evaluate several case-studies of timetable adjustments and passenger demand variations.

ABSTRACT. Recently, railway-infrastructure managers and suppliers have launched major programs with the aim to revise – or even reinvent – the overall interlocking architectures plus adjacent systems and operational principles. Representatives of such initiatives are “SmartRail 4.0” and “Digitale Schiene Deutschland”. Those programs are backed-up by a set of motivations, namely: Existing command and control technology is overaged or becomes outdated, skills to maintain technology get lost due to demographic aging, applied technology is expensive and does not allow any further capacity gain.

Since capacity improvement is a core target of all programs, there is a severe need to express the capacity impact of the related system architecture. Such quantifications serve the broad portfolio from political decision processes to detailed requirement specifications.

This article describes necessary extensions of the blocking-time model to meet the requirements of future interlocking architectures in detail. It contributes to extent the standardised blocking-time model in such a manner, that its applicability is ensured in different setups. Besides providing and extending the theoretic background, the article provides examples of practical computations and applications. They cover sensitivity analysis to elaborate the decisive impact parameters on headway times as well as timetabling-studies based on a future technology setup.

ABSTRACT. To accommodate increased demand for safe, efficient, and reliable freight and passenger rail transportation, North American railways are investigating implementing advanced communications-based train control (CBTC) systems with moving blocks to increase capacity of existing track infrastructure and leverage previous investments in Positive Train Control (PTC) technology. Under moving blocks, trains can operate at short headways based on individual train braking characteristics and track occupancy. Therefore, railway traffic flow, traditionally governed by fixed block signals, may begin to resemble highway traffic flow consisting of a continuous stream of vehicles interacting directly with each other. This research investigates if operations of freight trains around capacity bottlenecks caused by vertical grades or speed restrictions and controlled by moving blocks or fixed block wayside signal systems create shockwaves similar to those found in highway traffic streams. A macroscopic model is developed to produce density-flow curves describing operations of saturated, unidirectional, homogenous rail traffic around various combinations of grades and track speeds. Industry-standard railway traffic simulation software is used to validate the methodology. Model inputs and simulation headways are based on braking distances calculated for a typical North American bulk unit freight train. Results indicate that shockwave theory can predict the behaviour of rail traffic under fixed block and moving block control systems around stationary capacity bottlenecks caused by geography or civil speed restrictions. This knowledge can inform railway practitioner decisions on investment in advanced CBTC with moving blocks, future railway capital plans, and operating plans for new and existing lines with stationary bottlenecks.

ABSTRACT. This research focus on developing simulation system which can estimate train traffic or passenger flow under new train control systems. The goal of our research is to evaluate effects of installing intelligent signalling system or train control system, such as moving block or prediction control. In our previous work, we had developed “Train operation/passenger behaviour simulator ,” which estimates passengers’ train paths, train congestion, and train delay under a certain timetable. In that simulator, train delay is predicted by estimating train operation curves of each train between stations. However, train operation curves could be calculated under fixed block signalling systems only. So, we cannot estimate effects of new train control systems, like moving block or prediction control. In this research, we devised the fast estimation method for train operation curves under moving block. By that method, train operation curves can be calculated almost the same calculation time as that under fixed block. We also devised train control algorithm based on prediction control theory under moving block to minimize headway between successive trains. We then implemented those methods to the simulator, and evaluated effects of installing moving block signalling system and prediction control in an actual commuter line in Japan.

ABSTRACT. The increasing need for capacity has led the railway industry to explore next generation signalling concepts such as Virtual Coupling which takes moving-block operations further by separating trains by a relative braking distance, like cars on the road. By means of a Vehicle-to-Vehicle (V2V) communication architecture trains can move in a virtually coupled platoon which can be treated as a single convoy at junctions, to improve capacity. This concept however introduces the need for additional safety constraints, especially at diverging junctions, which could make actual capacity improvements insufficient to justify investments. Hence, there is a need to understand capacity performances of Virtual Coupling and potential gains over state-of-practice signalling systems. This paper addresses this need by developing an innovative train-following model that captures operational states and corresponding transitions of trains running under Virtual Coupling. A comparative capacity analysis has been conducted for a portion of the South West Main Line in the UK. Promising results have been obtained, showing that the biggest capacity gains returned by Virtual Coupling relate to operational scenarios normally found in practice with trains having service stops and using different routes.

ABSTRACT. This paper investigates the real-time train rescheduling problem in a high-speed railway line under a complete segment blockage by exploring the effectiveness of incorporating train coupling strategy on the train timetable rescheduling. The problem lies on determining the actual arrival and departure time as well as the platform track assignment of trains at stations after a complete segment blockage caused by disruptions, where reasonable trains satisfying strict rules could be coupled with others to avoid being cancelled. A novel mixed integer linear programming model is formulated to minimize the total deviation of trains' arrival and departure time to that in the planned timetable. In the model, both the acceleration and deceleration time of trains when departing from and arriving at stations are explicitly considered, while the platform track of trains at passed stations is jointly optimized. A rolling horizon algorithm is designed to effectively solve large-scale problem instances since the rescheduling of timetables is usually determined in stages in practice. Test instances constructed based on the Wuhan-Guangzhou High-Speed Railway in China are utilized to test the effectiveness and efficiency of the proposed approaches. Computational results demonstrate that the train coupling strategy is likely to reduce the total deviation and to relief the propagation of delays. Meanwhile, the rolling horizon algorithm can provide practically acceptable rescheduled timetables quickly. Thus, the train coupling strategy is promising in the field of train timetable rescheduling to cope with large-scale disruptions.

ABSTRACT. Railway networks are often operated close to their full capacity due to limited infrastructure expansion and increasing traffic demand. Hence, basic timetables are fairly vulnerable to random operational disturbances. In consequence of this, the service level for passengers decreases through a combination of delay propagation and delay accumulation. To solve this problem, a possibility widely used in research is to add extensive recovery and buffer times. Nevertheless, the resulting robust basic timetables would lead to a deterioration of the operating capacity, especially in congested areas. Another approach to reduce the impact of operational disturbances on railway operation is to use conventional dispatching algorithms. Unfortunately, most of them ignore further potential disturbances during the dispatching process, which is why the generated dispatching solution might even worsen train’s punctuality.

In this context, at the Institute of Railway and Transportation Engineering (IEV) at the University of Stuttgart a proactive dispatching algorithm has been developed, that generates dispatching solutions under the consideration of random disturbances in dynamic circumstances. The algorithm is divided into two main processes. First, the block sections are classified depending on their operational risk index by simulating numerous timetables with random disturbances generated in a Monte Carlo scheme and the related negative impacts in the studied railway network are calculated. Second, near-optimal dispatching solutions are automatically generated based on Tabu Search algorithm. This is achieved within a rolling time horizon framework, taking risk-oriented random disturbances in each block section into account.

ABSTRACT. The train rescheduling problem is quite a popular topic in the railway research community. Many approaches are available to reschedule traffic in a network partition but very few works address the coordination of these partitions. In railway systems with very dense traffic, e.g. the Swiss one, it is not always possible to partition the network such that the rescheduling algorithms can work on completely independent regions. This paper proposes a coordination approach for adjacent local rescheduling algorithms. These algorithms are based on the Resource Conflict Graph model, which enables the representation of the interlocking system at a very fine granularity. Simulations on data from the Swiss Federal Railways show the validity of this approach in improving the consistency of decisions at the common boundaries of adjacent local rescheduling algorithms.

ABSTRACT. Train platforming is critical for ensuring safety and efficiency of train operations within the stations, especially when train delays occur. This paper studies the problem of re-optimization of train platforming, where the train station is modeled using discretization of the platform track time-space resources. To solve the re-optimization problem, we propose a binary integer programming model which minimizes the weighted sum of total train delays as well as platform track utilization costs, subject to constraints defined by operational requirements. Moreover, we design an efficient heuristic algorithm to solve the model with a good precision. A real-world case is taken as an example to show the effectiveness of the proposed model and algorithm. The results show that the model established in this paper can describe re-optimization of train platforming accurately and can be solved quickly by the proposed heuristic algorithm. In addition, the model and algorithm developed in this paper can provide an effective computer-aided decision-making tool for the train dispatchers in case of train delays.

ABSTRACT. In this paper, an analytical model for strategic decision-making is developed, which makes it possible to compare and select the best solution of several railway junction designs according to theoretical infrastructure capacity, completely independent of timetable. The model achieves triple effects because, besides the above, it enables the selection of the most favorable sequence of train traffic through the junction, as well as determining the theoretical junction capacity. Model uses well-known combinatorics problems to determine the minimum infrastructure occupancy time. Testing has shown the simplicity of model implementation, as well as a satisfactory level of accuracy.

ABSTRACT. We propose a method to describe capacity utilization for railway infrastructure that applies blocking time theory to managing train runs. Different from traditional capacity evaluation, infrastructure capacity utilization description shows detailed information on infrastructure utilization hidden in timetabling data instead of sheer number of trains that can be operated, or capacity consumed. Using a function system defined upon necessary operational inputs for timetabling in blocking time theory, we can obtain the feasibility condition for operating consecutive trains. Thus, the method to identify critical block section can be deduced from the feasibility condition. Structural indication determines the capacity utilization of consecutive train paths, which can be further integrated into a bi-directional graph to model infrastructure capacity utilization description followed by infrastructure time allocation. Consumed capacity of railway infrastructure by operating train runs can be formulated. Besides, a general procedure is proposed to analyse the sensitivity of consumed capacity to operational inputs. An experimental case study is conducted to demonstrate the application of this method in analysing the impact of speed and recovery time.

ABSTRACT. Traffic variability is well known to have a substantial effect on railway capacity. Varying train running and stopping times entail larger train separations and yield non-usable time slots in train timetables and operations. In the present paper we aim to assess the effects of variability on the capacity of railway corridors in long term planning of rail traffic. Our main focus are metro transit systems in urban areas, where the effects of fluctuations of running and stopping times are particularly pronounced due to dense operations. The investigation comprises uncertainty of train stopping and running times as well as of the traffic concept itself, such as variations of headways and line frequencies.

On the system level, inhomogeneity tends to propagate in the network, yielding non-usable capacity, which is why we propose a stochastic modeling of rail corridors. Our model is motivated by finite capacity transfer lines, which have been studied in the context of manufacturing systems. The corridor is viewed as a sequence of heavily correlated service stations representing line segments and stations, for which effective throughput, distributions of train running times and service quality are calculated. The performance of the model is tested in a case study for the central link of the mass transit system in Cologne. In addition, an outlook on how the model can be extended to general heavy-rail corridors with different types of train services is provided.

ABSTRACT. Different types of trains may have substantially dissimilar characteristics, resulting in various capacity impacts. The concept of base train equivalent (BTE) was proposed to standardize different train types into a universal unit, namely, base train unit (BTU). However, the previously developed delay-based model suffers from consistency issue, and its application is limited to only two train types. Thus, this study proposes a new concept of delay-based BTE computation and corresponding BTE models. The dynamic BTE model considers volume and heterogeneity and aims to reflect fully the actual capacity impact of non-base trains. The fixed BTE model identifies the most appropriate BTE value at a particular traffic heterogeneity. Results from the case studies demonstrate that the proposed method can address scenarios with all types of traffic mixes and multiple train types. The unit of delay-based rail capacity can be converted into a standard unit using the proposed models. The effect of an additional train can be easily assessed, and the capacity measurements from different lines or systems can be compared and evaluated.

ABSTRACT. Railway capacity analysis usually aims to allocate the existing tracks to the desired traffic. Usually, track occupancy or indicators as delays are used to estimate how the system performs. In this article, we analyse the interaction of commuter and express train services. With help of microscopic simulation with RailSys, characteristics as delays are estimated for the regarded timetable slots and converted into a performance index. Assuming that the number of travellers is known, these individual indices can be converted into an index for the timetable alternative. A case study is performed on the Swedish Western Main line. Simulation is procured for all services on the line. Based on the results, an index is calculated for the respective timetabling alternatives. In that way, the effect of in-/decreased primary delays and supplements as well as major disturbances is analysed. The model is shown to be relevant for capacity allocation and can be helpful for the timetabling process when the passenger demand can be estimated. Estimated delays can have larger impact than differences in travel times. The analysis also showed that with the commonly used valuations, socio-economic effects on commuter trains are much lower than for long-distance trains. For further development of the model, an adjustment of these parameters should be considered.

ABSTRACT. Train timetabling plays an important part in train management, not only for passengers, but also for train operators. In a highly dynamic transportation market, train timetabling is an essential bridge connecting the service supplier with transportation demand. However, in present operations, train scheduling without considering passenger demand can reduce competitive advantages of railway in the multimodal transportation market and will further lead to passenger dissatisfaction. Therefore, it’s important to schedule trains responding to passenger demand in the train planning process. In this paper, we focus on the problem of train timetabling with passenger demand, specifically deciding train stop plan based on different origin-destination passenger demand pairs. Taking the stop indicators as important decision variables, a mixed integer linear programming model is proposed to address this train timetabling and stopping plan integration issue, with minimizing total train travel time and maximizing the number of transported passengers. The weighted-sum method is used to find the Pareto optimal solutions for the proposed bi-objective mathematical model. A set of numerical tests is presented based on Beijing-Jinan high-speed railway line (part of Beijing-Shanghai high-speed railway line) by Cplex optimization solver to validate the model.

ABSTRACT. We introduce a concurrent solver for the periodic event scheduling problem (PESP). It combines mixed integer programming techniques, the modulo network simplex method, satisfiability approaches, and a new heuristic based on maximum cuts. Running these components in parallel speeds up the overall solution process. This enables us to significantly improve the current upper and lower bounds for all benchmark instances of the library PESPlib.

ABSTRACT. See extended abstract.

ABSTRACT. Description/outline of the development trends beyond current standalone Driver Advisory Systems; C-DAS, Intelligent Cruise Control and use of the technology for ATO.

ABSTRACT. Braking curves from on-train data recorders from 2002 and 2018 are compared to identify changes in braking styles in the UK over time. The changes, typically of four seconds for each braking curve from 60mph to 0mph, are then assessed to understand how these small changes affect the operation of fixed block signalling systems.

The contribution of these small delays are then simulated in the "trenissimo" micro simulation tool to better understand how small changes in driving styles can contribute wider system performance through the spread of sub-threshold delay, and thus increase the delay per incident.

ABSTRACT. A user-centered, agile approach was used to develop a high-standard train simulator for applications in train driver education. A user group of train operators and train driver educators was formed to share experience and cut development cost. Joint prioritisation by the user group was used in combination with agile development to iteratively develop new variants of the train simulator, with features tailored to the user group’s needs. The user group has grown from 2 organisations in 2015 to 10 organisations in 2019, each of which now use the train simulator in education and training. They find that this has been beneficial in terms of quality, cost or time. Regarding research, this scheme has resulted in three PhD projects and several other projects on driver behaviour, railway signalling systems and capacity.

ABSTRACT. The purpose of this study is to develop a method for calculating the train trajectory in an uncertain environment in which the values of system parameters are difficult to determine. The method takes into account the human operator needs of avoiding possible hazards and extra workloads. A novel Approximate Dynamic Programming (ADP) method is proposed for optimization the train trajectory in real-time. To demonstrate the effectiveness of the proposed algorithm, we conduct a numerical simulation with real-life infrastructure, train and timetable data. We conduct deterministic optimization, which regards the system as having no uncertainty in its parameters, and then perform stochastic optimization (ADP), which takes the uncertainty into account in the optimization process, and compare the outcomes. The results successfully verify the effectiveness of our proposed algorithm.

ABSTRACT. Railways are essential for the sustainable accessibility of the big cities and the national and international mobility. Many urban and national railway networks face increasing capacity saturation issues, while the transport demand keeps growing for both passengers and freight. Likewise, high-speed networks in Europe and China are expanding rapidly and attract more and more passengers as sustainable and fast alternative for flying. And similar to the demand for international freight train corridors that often share tracks with national passenger railway networks. To enable this growth the railways are in a transition to modern digital railway traffic systems with continuous wireless communication for efficient signalling and traffic management. The timetable becomes increasingly more accurate to allow optimized infrastructure occupation, while train operation is supported by driver advisory systems connected to intelligent traffic management systems. Automatic train operation is already the standard in modern metro systems and will also be applied to other railway systems to realize the highest capacity consumption. The transfer to this modern digital technology is a challenge to the railway sector and requires an innovative integrated approach with the aim to achieve a safe, efficient and reliable transport system. This presentation gives an overview of the challenges ahead and the role of railway operation research in these innovations.

ABSTRACT. Planning rolling stock movements in industrial passenger railway applications is a long-term process based on timetables which are also often valid for long periods of time. For these timetables and rotation plans, i.e., plans of railway vehicle movements are constructed as templates for these periods. During operation the rotation plans are affected by all kinds of unplanned events. An unusal example for that is the collapse of a tunnel ceiling near Rastatt in southern Germany due to construction works related to the renewal of the central station in Stuttgart. As a result the main railway connection between Stuttgart and Frankfurt am Main, located on top of the tunnel, had to be closed from August 12th to October 2nd 2017. This had a major impact on the railway network in southern Germany. Hence, all rotation plans and train schedules for both passenger and cargo traffic had to be revised. In this paper we focus on a case study for this situation and compute new rotation plans via mixed integer programming for the ICE high speed fleet of DB Fernverkehr AG one of the largest passenger railway companies in Europe. In our approach we take care of some side constraints to ensure a smooth continuation of the rotation plans after the disruption has ended.

ABSTRACT. When large maintenance work is done at a double track line, it is often possible to have one of the two tracks open for traffic. The traffic then run with single track operation which heavily affects the capacity and need to be planned in an early stage, before the yearly timetable is finalized. Today, in Sweden, there are some difficulties when planning for maintenance works and how to adapt the reduced capacity in the timetable. Due to an increased demand for capacity and for better punctuality from train operators, there is a need for more well thought-out strategies for how to handle the capacity restriction and for how much robustness is needed in the timetable to preserve a certain quality. In this paper, we present a study which assess strategies for double track maintenance work leading to single track operations. A simulation study is performed in which three different timetable strategies are tested and evaluated. The aim is to find strategies and timetable rules to better handle capacity reductions at double track lines so that trains can run with high quality even though there are maintenance works at the same time. In the paper we discuss the advantages and disadvantages with the three strategies and how they affect train slots, runtimes and punctuality. We propose a strategy for how to best consider single track operations, depending on the length of the single track section, speed restriction and traffic pattern, including hands-on timetable rules.

ABSTRACT. Train passengers expect a high level of service under all circumstances, while disruptions occur on a daily basis. In some countries, dispatchers have predefined emergency plans at hand. These plans' flexibility and applicability are questionable as disruptions may be unique in timing, occurrence, or traffic state. In this paper, we present a novel approach for timetable rescheduling by introducing the concept of \textit{archetypical infrastructure pieces} (AIPs): frequently occurring parts of the infrastructure. We consider three AIPs outside station areas, ranging from (i) double-track corridors between two double switches without, and (ii) including stops, to (iii) multi-track corridors with stops. For each AIP, a machine-scheduling based mathematical model is presented. Additional service constraints can be included to retrieve more practicable results, or to improve passenger experience. Extensive experiments show that the models outperform heuristics representative for dispatchers' decision-making by effectively balancing between re-timing, re-ordering, stop-skipping, and cancelling trains. We estimate that disruptions on about 50\% of the Belgian railway network outside of station areas, can be considered using our three AIP models. Dispatchers could employ the models by characterizing the disruption and the infrastructure around it with a limited set of parameters. This approach allows to combine the models’ merit of efficiently balancing between measures, with the dispatcher’s knowledge on the surroundings of the corridor, and passenger expectations.

ABSTRACT. Global warming and climate change are indisputable theories. Since the Industrial Revolution, the mean temperature of the planet has increased by 1°C. Now, temperatures are approaching a higher stage of +1.5°C and the attention is on both CO2 emissions and energy consumption. Transportation is a major component of the environmental impact, accounting for approximately 30% of air pollution and energy consumption. Due to the rapid urbanization in the EU, with an estimated 74.3% of the population living in cities, forecasted to rise to 80% by 2050, urban mobility is dramatically increasing its relevance. Therefore, a reduction in energy consumption and pollutant emissions is a crucial factor to consider in developing urban transportation and particularly rail-based systems, able to provide energy saving transport services by improving urban environment. Several methods and techniques are under development to improve the energy performance of Light Rail Transport (LRT), which spread from different typologies of power supply to improving energy efficiency. The purpose of this paper is to start from the last developments and innovative energy sources for LRT systems. The focus is on two parts: a) trams running on Hydrogen in parallel with on board batteries with energy saving control techniques, b) potential renewable energy sources to meet power demand. The comparison is with traditional power sources and equipment (e.g. Catenary-based). The methods, based on selected indicators, are under development and test by calculations and simulations with reference to the case study of the new tramlines in the city of Brescia (Italy).

ABSTRACT. With the increase of the operating mileage, a large amount of energy consumption generated by metro systems needs to be taken seriously. One of the effective ways to reduce the energy consumption is to collaboratively optimize the driving strategy and train timetable by considering the regenerative energy (RE). However, the dimensionality and computational time will increase accordingly in optimization as the number of operating trains rises. With the intention of tackling this problem by efficiently reducing dimensionality, the energy-efficient metro train operation problem is optimized in this paper by applying the discrete differential dynamic programming (DDDP) approach. Firstly, the model calculating the net energy consumption that takes into account the RE is formulated. Then, the optimization model will be transformed to a discrete decision problem by using Space-Time-Speed (STS) network methodology, and the corresponding solution will be obtained through the DDDP based algorithm. Finally, two case studies will be conducted in a metro network to illustrate the effectiveness of the proposed approach.

ABSTRACT. Energy-efficient driving strategies are often disrupted by train separation constraints, particularly when there are short time headways between trains and some trains are delayed. When a train encounters a restrictive signal it will usually have to slow significantly, which disrupts efficient driving and introduces delays that can propagate back through the network.

Driver Advice Systems (DAS) can help trains follow a schedule precisely, and at the same time save energy. Connected Driver Advice Systems (C-DAS) extend this capability by adding communication with a central control system, which can provide real-time updates to individual train schedules in response to disruptions on the network.

In this paper, we use examples from a long-haul freight line and from an intercity passenger line to show how small adjustments to train schedules can be used to ensure safe separation of trains while minimising energy use, for trains travelling along a line between junctions.

We consider a pair of trains travelling along a line between junctions. First, we calculate independent energy-optimal driving strategies for each of the two trains. Next, we identify time intervals where the separation is less than the required separation distance. We then search for timing constraints that satisfy the separation constraints and for which the combined energy use of the two trains is minimised. This will result in smoother, more efficient train flows by avoiding encounters with restrictive signals.

ABSTRACT. With the rapid development of high-speed railway in China, train line planning encounters more challenges for holiday operations. Compared with weekdays, the number of high-speed railway passengers increases sharply during holidays and the spatial distribution of the passenger trips is diverse in different days. Thus, how to optimize holiday train line plan (HTLP) for high-speed railway to meet the fluctuation of passenger demand is becoming more and more important. Meanwhile, the objectives of reducing the railway operation cost and ensuring the similarity between the holiday and the weekday line plan should be considered. Nowadays, HTLP is usually made by manually adjusting train formation and adding extra trains based on weekday line plans, which have low flexibility and adaptability to satisfy holiday’s passenger demand. This study mainly focuses on the optimization of HTLP based on weekday line plans using a theoretical method to improve the adaptability and benefit of line plans. Compared with the holiday line plans in real-world, our experimental results show that our optimized HTLP has less train operation cost and more similarity with the weekday line plans. And the service level of high-speed railway is also improved. Based on the experimental results, the effectiveness of our proposed model and solution method has been proved. The proposed mathematical optimization model and method can be used to improve the holiday line plan of high-speed railway.

ABSTRACT. For long distance high-speed rails, designing trains using different stopping strategies that satisfy both long and short distance travel need is a difficult task. Limiting the number of train stops to reduce travel time conflicts with the goal of increasing passengers’ opportunity to gain direct connections. In addition, long trains with frequent stops occupy too much route capacity, while disconnected short trains make passengers to transfer. To handle the trade-offs, this paper presents a mixed integer linear programming model that determines optimal combinations of train stopping patterns, and we particularly consider the practice that many long trains use skip-stopping strategy. The objective of the model minimizes time loss and transfer time of passengers. Specifically, we intend to run fast trains between major stations, such that travel time of many passengers with long distance trips can be shortened; meanwhile, to operate a certain number of short trains with dense-stopping or all-stopping patterns so as to ensure connectivity level between non-major stations and passengers can have frequent direct or transfer train options. We design a genetic algorithm procedure to manage real-size instances. We have tested our model and algorithm for the Beijing-Guangzhou high-speed rail line in China, which is a 2,298 km line with 36 stations. Numerical results indicate that our approach improves the travel efficiency of passengers compared with an existing train stopping schedule. We also explore the trade-offs among different train stopping strategies and their impact on travel efficiency of passengers.

ABSTRACT. Infrastructure management and railway undertakings were organisationally separated in Great Britain as a result of the 1993 Railways Act. However, this vertical separation has proved problematic, not least in terms of assessing the trade-offs between capacity utilisation and performance in terms of reliability and punctuality. A possible solution is virtual integration where an alliance is formed between the infrastructure manager and the dominant railway undertaking. The first such alliance in Great Britain was between Network Rail and South West Trains (SWT) and ran from 2012 to 2015. This paper undertakes a statistical analysis of public performance measures and cancelled and significantly late services for SWT and two comparator railway undertakings, Govia Thameslink Railway (GTR) and Southeastern. There are no indications that virtual integration led to an improvement in the performance of SWT services but some indication that deteriorations were less than those for GTR, but this could be explained by other external factors. Overall, there is no clear indications that there were changes in punctuality and reliability trends for SWT as a result of virtual integration. Thus performance did not present a reason to end the virtual alliance between NR and SWT but nor did it provide reasons for continuing.

ABSTRACT. The main purpose of this study is to develop an understanding of the effect of trackwork on train punctuality. The study is also a pilot on combining railway traffic data and maintenance data for analytical purposes. This type of analysis can show how different types of trackwork influence punctuality. Previous studies on this subject have mainly utilised causes of delay data. In this study we have found methods to strengthen such analyses also by taking into account the high-resolution data available from registered records of train movements. The current study illustrates how such time data can be utilised in order to enhance the decision support system for maintenance. A data processing method is used to identify the affected lines. A search engine is designed in order to correlate the two datasets (including train traffic data and trackwork data) and to extract the required data for the analysis. Through the overlap between the duration of trackwork and the delay in departure time, it is demonstrated that it is possible to combine railway infrastructure data with train traffic data. As a result of this analysis, it is discovered that for railway traffic, in general, trackwork contribute, in fact, to a minor part of all causes of delays. However, a relatively large share of the trackwork was related to train delays. Another important finding from this study is that, when trackwork are conducted, it is a high probability that they will cause delays in train schedules and these delays are relatively large too.

ABSTRACT. A railway track near Rastatt, Germany, lowered on 12 August 2017 and caused a complete blockage of a sector of a major rail corridor, which lasted until 1 October 2017. This track closure had severe effects on the railway freight and passenger transport. This work investigates the effects on the Swiss railroad network, using openly available realized operation data. The behavior of the delays before, during and after the disruption is investigated on three different levels. First, the delay of arriving trains to Basel SBB, as it can be seen as the input delay into the Swiss railway system. Secondly, it is investigated how the delay evolves on the Swiss intercity and interregional lines in short distance (i.e. first stop) and thirdly how this delay evolves over the course of the lines. The results display a consistent improvement of punctuality during the disruption period, which however decreases when considering stations farther away from Basel SBB. This can be explained by the fact that during the disruption period, trains arriving from Germany at Basel SBB exhibit, due to the shorter running distance, significantly lower delays than during other periods. The improved punctuality is therefore a result of a reduced delay propagation of the trains arriving from Germany. The effects of this severe and long lasting disruption can be quantified even in some spatial and temporal distance. It can be used as an example to test theoretical models, which forecast delays, or examine train network complexity and interconnectivity.

ABSTRACT. This study aims to investigate delays, delay increases, and delay recovery characteristics, by using statistical methods to clarify delay propagation patterns according to historical records of the Wuhan-Guangzhou high-speed railway (HSR) in China in 2014 and 2015. Specifically, we examined arrival and departure delay duration distributions and used heatmaps to demonstrate the spatiotemporal frequency distribution of delays, delay increases, and delay recovery, and the heatmaps clearly show hot spots (coordinates with high frequencies) in a timetable. Then, we separated delays as discrete intervals according to their severity, and analyzed the delay increasing frequency and the delay increasing severity within each interval, so as to clarify the relationships of delay increasing probability and delay increasing severity with delay extents. Next, we investigated the observed delay recoveries and prescheduled buffer times at (in) station (section), which demonstrate the recovery ability of each station and section. Finally, to understand the key influencing factor of delay propagation, we analyzed the relationship between capacity utilization and delays, delay increases, and delay recoveries, by examining their Pearson correlation coefficients. These indicate that delay frequencies and delay increasing frequencies with Pearson correlation coefficients as high as 0.9 are highly dependent on capacity utilization. The uncovered delay propagation patterns can enrich dispatchers’ experience, and improve their decision-making ability during real-time dispatching in HSR.

ABSTRACT. On deregulated railway markets, efficient capacity allocation is important. We study the case where commercial trains and publicly controlled traffic (“commuter trains”) use the same railway infrastructure and hence compete for capacity. We develop a method that can be used by an infrastructure manager trying to allocate capacity in a socially efficient way. The method calculates the loss of social benefits incurred by changing the commuter train timetable to accommodate a commercial train path request and based on this calculates a reservation price for the train path request. If the commercial operator’s willingness-to-pay for the train path exceeds the loss of social benefits, its request is approved. The calculation of social benefits takes into account changes in commuter train passengers’ travel times, waiting times, transfers and crowding, and changes in operating costs for the commuter train operator(s). The method is implemented in a microscopic simulation program, which makes it possible to test the robustness and feasibility of timetable alternatives.
We show that the method is possible to apply in practice by demonstrating it in a case study from Stockholm, illustrating the magnitudes of the resulting commercial train path prices. We conclude that marginal societal costs of railway capacity in Stockholm are considerably higher than the current track access charges.

ABSTRACT. As railway markets are increasingly deregulated, coordinating and prioritising between capacity requests becomes more complex. Conventional administrative procedures are often based, explicitly or implicitly, on prioritising criteria, which make them insufficient to differentiate between competing operators with similar services. An allocation method that improves social welfare would be desirable, but calculating welfare requires information that is typically commercially sensitive, such as number of passengers, fares and cost structures, making such an approach difficult to implement.

This paper describes the respective virtues and challenges of different types of allocation methods in the context of a deregulated market with mixed traffic and government-owned tracks. Three broad groups of allocation methods are described: commercial traffic with long planning horizons, traffic organised by public traffic agencies, and short-term traffic.

We then outline an allocation method that better meets the requirements of a deregulated market. It is a mixed method, which uses an auction-like mechanism to allocate pre-defined paths to commercial operators on specified, capacity-constrained lines. The value of capacity for non-commercial public transport is assessed through social cost-benefit analysis (social CBA) of timetables. These social CBA results then decide the access charges for commercial operators.

ABSTRACT. This paper seeks a concept to include fixed-interval paths with manageable train slots to satisfy the flexible needs of freight traffic in a strict fixed-interval passenger timetable. The primary method is constituted by literature review and theoretical slot construction. The terms timetable and railway operation are specified and illustrated. Different levels of timetables will be discussed and further developed. Two concepts, slots and pulses, are described together with precondition and modelling to accomplish a mixed timetable level for flexible freight traffic and fixed passenger traffic. Finally, a comparison of the timetable levels with the rail freight corridor Rhine-Alpine is presented. In conclusion, three points for further research are made, and an experiment is suggested to validate the result in the future.

ABSTRACT. Railway schedules are often planned after timetabling principles derived from practical experience and based on the macroscopic limitations of a system, rather than the microscopic conflicts inherent in its signaling system. This inaccuracy in planning principles can lead to infeasible timetables, which induce delays and thus reduce the service reliability of railway transport. The objective of this paper is to support the design of fact-based timetables by introducing a systematic analysis of historical data from operations on the Danish railway, which includes testing the principles used in the timetable design, specifically regarding the allocation of headway time, and identifying possible improvements to the scheduling of trains according to their specific characteristics. The data records used in this analysis are generated by the signaling system and the automatic train detection system. The records state the scheduled and realized times of each train at every timing point on the network, as well as, categorical information about the trains and the measuring points. The timestamps are rearranged by an automatic algorithm to calculate the scheduled and realized running times, the scheduled and realized headways at the timing points, and the change in deviation between consecutive timing points for each train run. The relationships between these factors are then analyzed statistically to identify the adjustments that can be made to the timetabling principles to reduce the generation of secondary delays in operation and to improve the possibility of delay recovery under disturbed operation.

ABSTRACT. Trains are inevitably subject to interference from the external environment and internal systems during operation, leading to delays and conflicts. In this regard, there are usually buffer times allocated at (in) the station (section) in the train timetable, to recover delays. Most of the existing methods that deal with the buffer time allocation mainly consider the length of the section and the traffic density. These methods usually fail to consider the impact of the actual delay of trains, and the buffer time allocation (BTA) is unreasonable. The integration of the actual delay effects into the BTA needs to be resolved. Based on this, in this work, a delay time distribution model was established, and the models were compared according to the standard error of each parameter in the model. Subsequently, based on the delay distribution, a BTA model with weighted average delay expectation time as the objective function was constructed in which the weight coefficients were determined based on the delay strength, and the model was solved by a mathematical analysis method. Different allocation models were designed for different ranges of the total buffer time values. Finally, taking the Dutch railway network trunk section Maarssen–Utrecht Centraal (Mas–Ut) as an example, the results show that the buffer time after redistribution of the BTA model reduces the expected delay time in the segment by 5.25% compared with the original buffer time of the station, indicating that the BTA is reasonable.

ABSTRACT. In the context of the urbanization and the development of Urban Rail Transit (URT). The reliability of URT network is getting attention. The connectivity reliability is the basis of it. Three indicators are constructed to measure the connectivity reliability of URT network from passengers’ tolerable travel paths, passengers’ travel efficiency and passengers’ travel realization on URT network, respectively. The tolerable coefficient which is the ratio of passengers’ tolerable travel time and shortest travel time of all connected paths is proposed and added to evaluation indicators to reflect passengers’ choice behaviour of travel paths. Based on the maximum impact of passengers, ratio of affected passenger volume (RPV) is proposed to identify the important stations. The AFC data, train running time data are used to calculate the passenger volume and passengers’ tolerable travel paths in Wuhan subway. Finally, the connectivity reliability of Wuhan subway network is analysed through simulate attack stations. The results show that the identification methods of Degree Centrality (DC), Betweenness Centrality (BC) and ratio of affected passenger volume (RPV) can effectively identify the important stations on connectivity reliability of Wuhan subway. In particular, the method of RPV can identify the important stations that can influence the passengers’ travel realization in URT network most. In addition, attacking stations has greater impact on the passengers’ travel paths choice than passengers’ travel efficiency and travel realization.