Train Delays Research Articles

A Case-Control Analysis ofBone Stress Injury onAdvancement andHealth Care Utilization inUS Air Force Basic Military Trainees.

Every year, approximately 35,000 recruits enter the United States Air Force Basic Military Training (BMT). Musculoskeletal problems are particularly vexing for BMT because a broad array of differentials render definitive diagnosis burdensome and while most sufferers will fully recover, healing often requires protracted training delays resulting in increased unrecoverable training costs to the program. The purpose of this study is to conduct a case-control study of basic military trainees entering service from 2012 to 2015 to assess detectable differences in demographics, retention, and health care utilization. This is a retrospective case-control study of existing data and records obtained from existing and active information record systems for individuals who attended BMT between the years 2012 and 2015. Cohort assignment was based on incidence of first stress fracture diagnosis (i.e., having a select diagnosis code in one's medical record) relative to arrival at BMT. Health care utilization is operationalized as medical encounter frequencies and associated international classification of disease (ICD) codes occurring for all patients from BMT recruit date of arrival to 6 months post-graduation (i.e., 180 days). The primary dataset includes 132,359 distinct individuals, of which 129,637 (98%) had no history of stress fracture diagnosis and 1,487 (1%) of which have a diagnosis of stress fracture more than 60 days after arrival at BMT; these are assigned to the CONTROL group. There are 1,235 (0.9%) assigned to the CASE group. CASE members presented with 30.1 unique ICD-10 codes post-graduation compared to 6.3 in the CONTROL group. Six of the top thirteen ICD-10 codes were musculoskeletal in nature and all presented significantly higher rates for the CASE group. Trainees who suffered a bone stress injury (BSI) during the United States Air Force BMT advanced in rank on par with trainees who did not, but BSI sufferers exhibited greater health care utilization than those who did not suffer BSI. The cost to replace a trainee unable to complete BMT is considerable; however, it is also important to consider the reduction in resiliency and readiness to the Department of Defense and the financial burdens from increased health care utilization. The 2-fold increase in BSI presentation in the graduated control group reflects that more needs to be done for all populations to identify better proactive efforts to foundationally support wellness and prevention rather than accept injury as the status quo.

An Automatic Train Operation Based Real-Time Rescheduling Model for High-Speed Railway

With the continuous development of the Automatic Train Operation (ATO) system in high-speed railways, automatic driving is progressively supplanting manual operations, ushering in a new era of predictability and reliability for high-speed railway transport. Concurrently, the advent of the ATO system provides a notable impact on real-time rescheduling during disruptions, as it equips dispatchers with precise insights into train operation statuses. This paper is dedicated to a thorough analysis of how the transition to automatic driving in train operations influences the real-time rescheduling model. Based on the distinctive impact of the ATO system on real-time rescheduling, we have proposed a mixed-integer linear programming model that combines train re-timing, reordering, and the minimization of passenger delays. To validate the effectiveness of our model, we present several experiments conducted using data from the Beijing–Shanghai high-speed railway line. The results unequivocally demonstrate that our ATO-based model significantly mitigates train delay time, demonstrating its practical value in optimizing high-speed railway operations.

Train rescheduling and platforming in large high-speed railway stations

To deal with train delays in large high-speed railway stations, a multi-objective mixed-integer nonlinear programming (MO-MINLP) optimization model was proposed. The model used the arrival time, departure time, track occupation, and route selection as the decision variables and fully considered the station infrastructure layout, train operational requirements, and time standards as limiting factors. The optimization objective was to minimize train delays and reduce track and route adjustments. To realize the large-scale and rapid solution of the MO-MINLP model, this study proposed a rolling horizon optimization algorithm that used “half an hour” as a time interval and solved the rescheduling and platforming problem of each time interval step-by-step. In numerical experiments, 227 train movements under delay circumstances in Hangzhoudong station were optimized by using the proposed model and solution algorithm. The results show that the proposed MO-MINLP model could resolve route conflicts, compress unnecessary dwell times, and reduce train delays; the solution algorithm could efficiently increase the computational speed. The maximum solution time for optimizing the 227 train movements is 15 min 24 s.

Prediction of departure delays at original stations using deep learning approaches: A combination of route conflicts and rolling stock connections

Rolling stock connections and route conflicts of trains at terminal stations are critical to train delay propagation and prediction on the railway networks. Previous studies primarily focused on delay prediction/propagation from a macro perspective, without considering the two factors. To address this problem, a hybrid neural network architecture, called TLF-net, is proposed to predict departure delays at original stations (DDOSs), considering the rolling-stock connection and potential route conflicts in the railway network. TLF-net consists of a transformer, a long short-term memory (LSTM), and a fully-connected neural network (FCNN) block, to separately address variables with different characteristics. Based on the real-world data from two terminal stations in China, the experimental results show that the consideration of the potential route conflicts from the network can considerably improve TLF-net's prediction performance over the model that only considering train interactions on a single railway line. Also, it is proven that arrival/departure routes of consecutive trains are crucial for delay prediction, while using the transformer block can efficiently reveal the route conflict severity between arrival/departure routes. The sensitivity analysis to influence factors demonstrates the significance of considering the rolling stock connection and potential route conflicts. Finally, to support real-time railway traffic management, a train arrival delay prediction model from our previous study is integrated to predict the input of TLF-net (i.e., train arrival delays), enabling the proposed model to dynamically predict the DDOSs. This dynamic updating lengthens the prediction horizon (the prediction time ahead), making it better support real-time train traffic control and management.

A multi-agent simulation based train platforming research for facilitating passenger transfer in a high-speed railway station

The train platforming problem is one of the most investigated topics in the train operation field. However, a systematic understanding of how train platforming improves the convenience of passenger transfer procedures is still lacking. To face the rapidly growing transfer demands in the high-speed railway station, this paper proposes a transfer-facilitating-oriented train platforming approach based on a multi-agent train operation simulation model. By evaluating the transfer connection quality using an indicator named Transfer Connection Strength(TCS), which considers transfer distance and transfer time, the proposed transfer-facilitating-oriented train platforming strategies can establish more convenient transfer connections and so improve the transfer service. A track occupation horizon based on the modelling of accurate train running processes is utilized to detect route conflicts and guarantee conflict-free train operations in the simulation model, enabling the solutions’ feasibility of train platforming to be taken into account. A series of simulation experiments are conducted using real data of Zhengzhou East station in China. The experiment results show that, the proposed transfer-facilitating-oriented train platforming strategies are capable of obtaining a greater number of convenient transfer connections. The feasibility can be guaranteed by permitted train delays and route conflict detection for the succeeding trains in a forward-looking period. Additionally, the platform assignment plans can be generated with a specified priority of transfer distance or transfer time.

Delay prediction with spatial–temporal bi-directional LSTM in railway network

Train delay prediction is a vital part of railway system, but due to uncertain factors such as the complexity of the railway system and spatial–temporal features, it is often difficult to predict train delay in practice. In this paper, we propose a Spatial–Temporal and Bi-directional Long Short-Term Memory (ST-BiLSTM) model to deal with the train delay prediction problem. The model contains spatial–temporal blocks to capture spatial and temporal features and a bi-directional Long Short-Term Memory (LSTM) block to introduce bi-directional information through an attention mechanism. Experiments demonstrate that ST-BiLSTM outperforms the existing baselines in two evaluation metrics.

In-Network Computation for Large-Scale Federated Learning Over Wireless Edge Networks

Most conventional Federated Learning (FL) models are using a star network topology where all users aggregate their local models at a single server (e.g., a cloud server). That causes significant overhead in terms of both communications and computing at the server, delaying the training process, especially for large scale FL systems with straggling nodes. This paper proposes a novel edge network architecture that enables decentralizing the model aggregation process at the server, thereby significantly reducing the training delay for the whole FL network. Specifically, we design a highly-effective in-network computation framework (INC) consisting of a user scheduling mechanism, an in-network aggregation process (INA) which is designed for both primal- and primal-dual methods in distributed machine learning problems, and a network routing algorithm with theoretical performance bounds. The in-network aggregation process, which is implemented at edge nodes and cloud node, can adapt two typical methods to allow edge networks to effectively solve the distributed machine learning problems. Under the proposed INA, we then formulate a joint routing and resource optimization problem, aiming to minimize the aggregation latency. The problem turns out to be NP-hard, and thus we propose a polynomial time routing algorithm which can achieve near optimal performance with a theoretical bound. Simulation results showed that the proposed algorithm can achieve more than 99<inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> of the optimal solution and reduce the FL training latency, up to 5.6 times w.r.t other baselines. The proposed INC framework can not only help reduce the FL training latency but also significantly decrease cloud&#x2019;s traffic and computing overhead. By embedding the computing/aggregation tasks at the edge nodes and leveraging the multi-layer edge-network architecture, the INC framework can liberate FL from the star topology to enable large-scale FL.

Optimization of system resilience in urban rail systems: Train rescheduling considering congestions of stations

In urban rail systems, unexpected events, due to equipment breakdown, extreme weather, etc., frequently cause the delay of trains and overcrowding of platforms, significantly affecting the service quality to passengers. Therefore, improving system resilience to recover to normal conditions more quickly has become a critical problem in the management of urban rail systems. In contrast to most previous studies that focused only on train rescheduling, our paper proposes optimizing the resilience of urban rail systems by considering the influence of delayed trains and overcrowded passenger flows. Specifically, we define a resilience-oriented quantitative evaluation index. Using the evaluation index as the objective function, we construct a mixed-integer nonlinear programming model. In our formulation, we consider a practical case in which the boarding and alighting of passengers at overcrowded stations may further delay trains. To solve the model, we develop a two-phase iterative optimization approach, in which phase 1 focuses on the adjustment of trains to enhance system resilience, and phase 2 evaluates the impact of passengers on the adjusted train schedule. Through the comparison of three different rescheduling methods, we prove the validity of the derived inequalities and our approach balances the computational time and system performance. Case studies are conducted on the Beijing Metro Line 1 to verify the effectiveness of the proposed approach. The results demonstrate that system resilience can be improved by as much as 39.7% using our optimization approach.

Joint Optimal Train Rescheduling and Passenger Flow Control for Speed Limit and High-Demand Scenarios of Urban Rail Transits

In the operation of urban rail transits, train delays occur frequently, and emergency management is one of the key factors to ensure the service quality. For speed limit scenarios with high demand, this paper takes the train running in the restricted manual (RM) mode in the speed limit zone as an example and discusses a coping method by jointing train rescheduling and passenger flow control. With the goal of maximizing the number of passengers served and minimizing total train delay, a nonlinear optimization model is constructed by taking the train operation-related requirements and passenger flow control-related indicators as constraints, and the model is reformulated to a mixed-integer programming (MIP) model with quadratic constraints, which can be solved by the Gurobi solver. In order to obtain effective solutions faster, a two-stage approach is discussed, which first obtains a rescheduling timetable and then dynamically adjusts the requirement of boarding equalization to obtain the passenger flow control scheme. The validity of the model and the solution approach are discussed with the help of numerical experiments. The results suggest that the model and solution approach are feasible. When the number of trains is fixed, the reasonable implementation of passenger flow control will help to increase the number of passengers served, and the pursuit of a higher equalization of boarding is not conducive to the number of passengers served and the full utilization of transport capacity. The two-stage approach has certain advantages over the direct computing method. The methods in this paper have a guiding value for emergency decision-making in similar delay scenarios.

Capacity-constrained mean-excess equilibrium assignment method for railway networks

The reliability and unreliability of travel time variability caused by stochastic train delays considerably influence passengers’ travel path choices. This paper proposes an equilibrium assignment method that can capture the passengers’ path choice behaviors considering both the reliability and unreliability of travel time variability in the railway network. The passengers’ travel path choices are modeled using the mean-excess travel cost (METC), which is the combination of the conditional expectation of travel cost exceeding the effective travel cost, and the endogenous advanced ticket booking cost associated with the rigid train capacity. In general, heterogeneous passengers with different risk-aversion attitudes toward stochastic train delays evaluate their travel options differently and have different requirements for successful transfer. Therefore, first, the passengers’ transfer options are modeled with the concept of METC based on the delay distribution of the two connecting trains. Subsequently, the METC of the passenger trips is calculated. A mean-excess passenger flow assignment model is formulated based on the user equilibrium condition. A path-based capacity-constrained passenger assignment algorithm is designed to solve the model. Moreover, three illustrative examples are presented to demonstrate the effectiveness of the model and solution algorithm.

Improved Genetic Algorithm for Train Platform Rescheduling Under Train Arrival Delays

In this study, the train platform rescheduling problem (TPRP) at a high-speed railway station is analyzed. The adjustments of the train track assignment and train arrival/departure times under train arrival delays are addressed in the TPRP. The problem is formulated as a mixed-integer nonlinear programming model that minimizes the weighted sum of total train delays and rescheduling costs. An improved genetic algorithm (GA) is proposed, and the individual is represented as a platform track assignment and train departure priority, which is a mixed encoding scheme with integers and permutations. The individual is decoded into a feasible schedule comprising the platform track assignment and arrival/departure times of trains using a rule-based method for conflict resolution in the platform tracks and arrival/departure routes. The proposed GA is compared with state-of-the-art evolutionary algorithms. The experimental results confirm the superiority of the GA, which uses the mixed encoding and rule-based decoding, in terms of constraint handling and solution quality.

A train timetable rescheduling approach based on multi-train tracking optimization of high-speed railways

PurposeThis paper aims to propose a train timetable rescheduling (TTR) approach from the perspective of multi-train tracking optimization based on the mutual spatiotemporal information in the high-speed railway signaling system.Design/methodology/approachFirstly, a single-train trajectory optimization (STTO) model is constructed based on train dynamics and operating conditions. The train kinematics parameters, including acceleration, speed and time at each position, are calculated to predict the arrival times in the train timetable. A STTO algorithm is developed to optimize a single-train time-efficient driving strategy. Then, a TTR approach based on multi-train tracking optimization (TTR-MTTO) is proposed with mutual information. The constraints of temporary speed restriction (TSR) and end of authority are decoupled to calculate the tracking trajectory of the backward tracking train. The multi-train trajectories at each position are optimized to generate a time-efficient train timetable.FindingsThe numerical experiment is performed on the Beijing-Tianjin high-speed railway line and CR400AF. The STTO algorithm predicts the train’s planned arrival time to calculate the total train delay (TTD). As for the TSR scenario, the proposed TTR-MTTO can reduce TTD by 60.60% compared with the traditional TTR approach with dispatchers’ experience. Moreover, TTR-MTTO can optimize a time-efficient train timetable to help dispatchers reschedule trains more reasonably.Originality/valueWith the cooperative relationship and mutual information between train rescheduling and control, the proposed TTR-MTTO approach can automatically generate a time-efficient train timetable to reduce the total train delay and the work intensity of dispatchers.

Modeling the Duration of the Impact of Unplanned Disruptions on Passenger Trips Using Smartcard Data in Urban Rail Systems

Many urban rail systems operate near capacity given the rapid increase in passenger demand, and unplanned disruptions are unavoidable. From a passenger perspective, the duration of trip delays is a major concern, and passenger trip delays may be longer than the train delays. Several studies have focused on predicting train delays, but the research on the duration of the disruption impacts on passenger trips is limited given that the duration is not observed directly. This paper proposes a probabilistic method to estimate the disruption impact duration using smartcard data, explores statistical and machine learning models to predict the duration of impacts on passengers, and identifies influencing factors including incident characteristics, operating conditions, infrastructure, external factors, and demand. The results highlight that prediction accuracies are acceptable for multiple linear regression, accelerated failure time, and random forest models. Disruptions caused by power failures have longer impact durations than other causes, followed by platform screen doors. The fixed block signaling system leads to a larger disruption duration than the moving block system. The study provides, for the first time, a data-driven approach to understanding the duration of the impact of disruptions on passenger trips using smartcard data which can facilitate timely and informed decision-making under unplanned disruptions.

Deep Q-Network Approach for Train Timetable Rescheduling Based on Alternative Graph

The disturbance of local areas with complex railway networks and high traffic density not only impedes the efficient use of rail networks in those areas, but also propagates delays to the entire railway network. This has motivated research on train rescheduling problems in high-density local areas to minimize train delays by modifying their planned arrival and departure times. In this paper, we present a train rescheduling method based on Q-learning in reinforcement learning. More specifically, we used deep neural networks to approximate the action-value function, and the underlying Markov decision process (MDP) is based on the alternative graph formulation for the train rescheduling problem. In the proposed MDP formulation, the status of the alternative graph corresponding to the current schedule is defined as the state, and the alternative arc corresponds to the action the agent can take. The MDP is approximately solved via deep Q-learning in which deep neural networks are used to approximate the action-value function in Q-learning. Although the size of the alternative graph depends on the number of trains, our MDP formulation is independent of the number of trains, which makes the proposed method more scalable. The evaluation of the method was performed on a simple railway network and a real-world example in Seoul, South Korea, with randomly generated initial train schedules and train delays. The experimental result showed that the proposed method is comparable to the mixed-integer-linear-programming (MILP)-based exact approach with respect to the quality of the solution.

Privacy‐preserving model splitting and quality‐aware device association for federated edge learning

AbstractFederated edge learning (FEEL) provides a promising device‐edge collaborative learning paradigm, which enables edge devices to parallel participate in model co‐creation while preserving user privacy, opening countless opportunities to enable edge intelligence. With the growing demand for intelligent services, extensive FEEL deployment is inevitable. Nevertheless, existing FL schemes neglect two unique features (i.e., resource heterogeneity and data heterogeneity) in real‐world edge learning and thus may negatively affect the training efficiency and accuracy. Specifically, (1) heterogeneous and limited device resources cause massive laggards, which bring intolerable training delay; (2) heterogeneous data distribution causes device quality divergence, bringing severe training accuracy degradation. This article proposes a split‐based FEEL framework and an adaptive model splitting and quality‐aware device association scheme (MSDA) to tackle the aforementioned challenges. MSDA contains two levels: at the model splitting level, according to device capability and model structure, an adaptive splitting mechanism is proposed to provide a low‐latency and privacy‐preserving model splitting strategy for each device and guide subsequent device association. At the device association level, each device is simulated as a player with a quality weight in the potential game. Then a quality‐aware decentralized device association mechanism is designed to ensure that more high‐quality devices upload local updates before the deadline with the help of the edge server. Finally, experimental results demonstrate that MSDA yields significant improvements, achieving up to 3.1 training speedup and 39% accuracy improvement compared to state‐of‐the‐art methods.

The Influence of Service Quality and Timeliness on Passenger Loyalty, Yogyakarta International Airport Train

Yogyakarta International Airport (YIA) is a new airport located in Kulon Progo Regency, Yogyakarta. One of the transportation modes that can be used to reach the airport is the YIA Airport Train. With the increasing number and routes of airlines operating at YIA, the number of airport train passengers has also increased. However, this increase in passenger numbers has not been accompanied by an improvement in the quality of service provided and train delays are still occurring. This study aims to determine the influence of service quality on passenger loyalty, on-time performance on passenger loyalty, and the combined influence of service quality and on-time performance on passenger loyalty. The methods used in this study are Structural Equation Modeling (SEM) and multiple linear regression analysis. Data processing was done using SmartPLS 3.2.9 software for the SEM method and IBM SPSS 20 for the multiple linear regression analysis method. The results of this study indicate that both methods used support each other's findings. The results show that service quality has an impact on passenger loyalty. On-time performance also has an impact on passenger loyalty. Additionally, service quality and on-time performance have a positive and significant influence on passenger loyalty. In conclusion, it can be inferred that the better the service quality and on-time performance provided by the service provider, the higher the passenger loyalty will be.

Extensive Model and Matheuristic Algorithm for the Train Platforming Problem with Two-Train-Capacity Tracks: A Case Study of Prague Central Station

This paper provides a deeper insight into the train platforming problem (TPP). Many studies have focused on different versions of train scheduling and routing problems, and most of them assume that the platform track’s capacity is one train. However, especially in busy and complex railway stations, most platform tracks are divided into two parts, allowing two trains to simultaneously share the same platform track for passenger boarding/alighting. This results in more efficient train assignment to the platform tracks. In addition, consideration of the track capacity makes the problem more difficult because directions of trains are problematic. Motivated by this challenge, we consider the TPP with two-train-capacity tracks. We first describe the problem in detail and then propose a mixed-integer programming model. The objective of the considered problem is to minimize the total weighted train delays, which are defined as the difference between the departure times calculated by the mathematical model (M1) and the scheduled departure times of the trains in the timetable. Because of the NP-hard nature of the problem, the proposed M1 may not find feasible solutions for large-size problems. Thus, a matheuristic algorithm (MA) is developed to solve large-size problems. We used randomly generated test problems to demonstrate the performance of the proposed M1 and MA. Experimental results showed that MA outperforms M1 in both solution quality and solution time. Additionally, a case study was conducted at the central station of Prague, Czechia.

Aerodynamic intensification effect and dynamic response of cracks on high-speed railway tunnel linings

Falling concrete blocks are serious problems in high-speed railway tunnels, and they cause delays of high-speed trains (HSTs) and even compromise driving safety. When HST-induced aerodynamic shockwaves propagate into cracks, the intensified pressure makes the cracks grow and connect, resulting in falling concrete blocks. In this study, the longitudinal, oblique and circumferential cracks at the tunnel vault are selected as representatives. The temporal, spatial and spectral characteristics of the aerodynamic pressure in the cracks are studied using the unsteady viscous k-ε turbulence model. The aerodynamic intensification effect in the lining crack is also revealed. Results show that the maximum pressure of the circumferential crack is 1.51 and 1.37 times that of the longitudinal and oblique cracks, respectively. On the basis of the results of soil–tunnel–crack models established in ABAQUS, two response surface models (RSMs) are established to quantify the effects of important factors, including crack length, crack depth, crack width, incident angle of the aerodynamic shockwave and train velocity, on the maximum tensile strain of the crack tip of circumferential cracks. Train velocity exerts the greatest influence on the maximum tensile strain in the circumferential crack, followed by crack width. Moreover, a field measurement is performed to investigate the dynamic strain of crack tips under train-induced aerodynamic loads and verify the proposed RSM. The maximum tensile strain caused by aerodynamic pressure is approximately 0.31–0.86 times of the maximum tensile strain of concrete. The proposed RSM can realise a reasonable prediction of the dynamic strain of crack tips. This research may be valuable for analysing the crack tip stability of tunnel lining cracks and guaranteeing the structural health of high-speed railway tunnels.

An efficient evolutionary algorithm for high-speed train rescheduling under a partial station blockage

This paper investigates the high-speed train rescheduling (HSTR) problem under a partial station blockage and proposes an efficient problem-specific strengthen elitist genetic algorithm (PS-SEGA) for HSTR. Firstly, a HSTR model subject to train operation constraints is established to minimize the total train delay. A permutation-based encoding method is developed to define an efficient search space based on the train departure sequence. A heuristic decoding method is employed to eliminate all train operation constraints and output the rescheduled timetable. Moreover, a hybrid initialization method involving an efficient heuristic strategy (EHS) is put forward to accelerate the convergence speed of PS-SEGA. Using problem-specific knowledge, EHS generates an efficient and feasible solution for the initial population. Finally, a restart strategy is presented to maintain genetic diversity. Compared with other advanced evolutionary algorithms and their improved variants also using the improvements of PS-SEGA, experimental results demonstrate the effectiveness of the proposed PS-SEGA for addressing HSTR scenarios under the partial station blockage. As for the scenarios that CPLEX cannot obtain optimal solutions within 10 min, PS-SEGA can provide quasi-optimal solutions in real time. Furthermore, compared with the other two heuristics algorithms (i.e., First-Scheduled-First-Served and EHS), PS-SEGA can give the train departure sequence with a smaller total train delay.

Understanding passenger travel choice behaviours under train delays in urban rail transits: a data-driven approach

The analysis of passenger travel choice behaviours under train delays has become a crucial topic in research on urban rail transit operation management. In this paper, we focus on analysing travel choices of affected regular passengers under train delays by utilizing the data collected through an automatic fare collection (AFC) system along with train delay log records. Along this line, we propose a data-driven four-stage framework for studying regular passengers’ responses under delays, consisting of data profiling, regular passenger screening and travel patterns extraction, affected regular passenger identification, and affected passenger behaviour prediction modelling. Using a real-world case of the Shenzhen Metro in China, we conduct extensive experiments for method validation and feature insights analysis. The proposed framework could provide a microscopic view of passenger travel behaviours under train delays for fine prediction and exhibit a possibility for multi-source heterogeneous data mining in passenger behaviour analysis and train delay-related tasks.