Abstract Ride-hailing services, characterized by convenience, flexibility, and on-demand availability, have substantially reshaped urban mobility patterns, posing uncertain impacts on subway ridership. To clarify the competitive and cooperative dynamics between two modes, a latent class choice model (LCCM) is employed using revealed preference survey data (N = 2061) collected from Beijing, Shanghai, and Shenzhen in China. Three distinct commuter segments are identified through latent class analysis, and results indicate that ride-hailing and subway predominantly exhibit a competitive relationship (46.34%). Moreover, user heterogeneity is evident across user groups. Competitive users ride more frequently and commute shorter distances, and carless users tend toward competitive relation, while cooperative behaviors are linked to income and metro accessibility. Interestingly, commonly used spatial proximity assumption in previous studies exerts few influences on empirical modal interactions. Designing user-targeted takeaways for different transportation participants, this work offers valuable insights for improving multimodal travel efficiency and sustainable urban mobility.
Through large-scale new construction, expansion and comprehensive development of suburban railways, the Tokyo metropolitan area has effectively driven the development of the peripheral areas of the metropolitan circle, which has achieved good socioeconomic benefits and is a model for the construction of human settlements. This study adopts an exploratory sequential mixed-methods research design, taking the JR Chuo Express Line, Tokyu Den-en-toshi Line, and TX Express Line as typical cases. It first conducts a historical policy discussion based on textual analysis of the transformation and renovation of Tokyo’s suburban railways and their relationship with urban development, sorting out the characteristics and successful experiences of integrated development along suburban railways in Tokyo. Then, it carries out a data-based discussion on the current situation, summarizing the effects of the lines and their stations on driving population growth and land prices, the TOD (transit-oriented development) model along the lines, and the collaborative strategies between railways, stations, and surrounding cities, so as to analyze the synergistic relationship between the transformation of suburban railways and urban development in Tokyo as well as the effects after the transformation. Research results indicate that during the early construction and development stages of the outer areas of the Tokyo metropolitan area, suburban railways have had a positive effect on the development of population and land prices along the line, as well as the integrated development of stations and cities. Their construction experience can provide reference for the formulation of suburban railway development strategies and integrated development along the lines in similar international metropolitan circles.
We are presenting a crew planning optimization project for the train drivers of an urban rail transportation company. The core is about crew rostering, i.e., defining over a certain period of time (e.g., one year) for each train driver sequences of working days and rest days, and also specifying whether working days shall contain either some early, late, or night shift. But the project was not only about solving just one classical crew rostering problem. Rather, during the last three years, the entire process of crew rostering had been investigated. Apart from having modeled, solved, and implemented two different variants of crew rostering—a cyclic and an acyclic one—there have been designed and used three further optimization models for less typical surrounding sub-processes, as well as a simple crew assignment model. During the design and implementation process of these six optimization models, the focus had been put on their annual and daily usability, respectively. In particular, many practical requirements had been collected and implemented in rather straightforward ways. The result is a family of mathematical optimization models, whose results cover the valid annual crew rosters from the year 2024 on, as well as the daily assignment of specific duties to train drivers from May 2025 on at S-Bahn Berlin GmbH.
Abstract Subway Station Construction Site Layout Planning (SSCSLP) in dense urban cores is characterized by extreme spatial constraints. Conventional Constraint-Preserving Search (CPS) paradigms often exhibit significant limitations in such environments. Specifically, the strict rejection of infeasible solutions fragments the search space, frequently causing stagnation in local optima. To address these challenges, a novel Graph-based Dynamic Constraint-Relaxation Multi-Objective Optimization Framework is proposed. An Edge-Attributed Weighted Graph is utilized to capture complex spatial dependencies. Uniquely, the Graph-based Dynamic Constraint-Relaxation NSGA-II (GDCR-NSGA-II) is developed to overcome optimization bottlenecks. A Dynamic Constraint-Relaxation Strategy (DCRS) transforms hard constraints into a continuous penalty landscape. This mechanism establishes an infeasibility-driven search trajectory, guiding the population from the infeasible region toward the global optimum at the feasible boundary. The proposed framework was validated using a case study of Chongqing Rail Transit Line 27. Comparative analysis demonstrated that, when the single best feasible solution identified by the conventional method was strictly used as the benchmark, the proposed framework reduced the average construction cost by approximately 49.4% and improved average safety performance by 63.7%. Consequently, this study provides robust theoretical support for intelligent decision-making in ultra-constrained engineering scenarios.
Flexible (un)coupling operation within a Y-shaped network allows for the dynamic adjustment of metro train formations at the diverging junction based on fluctuating passenger demand. Train timetable and rolling stock schedule serve as essential elements in organising efficient flexible (un)coupling train operation. However, reducing overall waiting time during the optimisation process may lead to disparities, where some passengers experience much longer waiting periods compared to others. To address this issue, we formulate a multi-objective optimisation model that jointly determines train formations, timetable, and rolling stock schedule while balancing operating cost, total waiting time, and fairness. Moreover, we design an operationalisation method for balancing the objectives and solve the problem using the Gurobi solver. Numerical experiments are conducted under different scenarios on a Y-shaped metro network in Guangzhou, China. The results reveal that the model can effectively improve fairness compared to cases without fairness considerations. By allowing a moderate increase in total waiting time or operating costs, passengers with the longest waiting times experience reduced waiting durations. The standard deviation of waiting times also decreases. For instance, a 25.6 % rise in operating costs yields a 27.4 % reduction in passenger waiting time cost and a 36.5 % reduction in the standard deviation of waiting times. Ultimately, a more efficient balance between fairness, operating costs, and overall passenger waiting time is achieved.
Efficient weekly scheduling, including task assignments, maintenance, washing, and parking, for urban rail transit vehicles is essential for reducing operational costs and ensuring smooth depot operations. However, optimizing these components separately may cause inefficiencies, conflicts, and suboptimal vehicle utilization, ultimately compromising system performance and reliability. This study presents an integrated optimization model to minimize mileage variance, driving distances, and shunting operations. To address the multiobjective nature of the problem while satisfying operational constraints, a novel utility-balancing simulated annealing (U-SA) algorithm is developed. In this algorithm, the utility-balancing mechanism is embedded in the SA framework and guided by fuzzy programming to coordinate multiple objectives effectively. The algorithm utilizes three distinct operators—mileage-balancing operator, distance-minimization operator, and shunting-optimization operator—to steer the optimization toward an efficient solution. The model and algorithm are validated through a real-world case study of Beijing Subway’s Line 12. Computational results demonstrate that the approach effectively balances objectives, reduces operational costs, and enhances resource utilization. Sensitivity analysis reveals that adding parking tracks reduces shunting operations, while earlier washing windows helps shorten driving distances. This method offers a practical decision-support tool to improve the efficiency of urban rail transit operations.
Urban rail transit (URT) service disruptions, though infrequent, involve high risk and uncertainty, often triggering passenger behaviors that deviate from normal patterns. How stranded passengers make travel decisions under such conditions is essential for improving emergency response and system resilience. However, traditional models based on expected utility theory (EUT) often fail to capture the psychological complexity and bounded rationality exhibited during disruptions. This study develops a decision-making framework that integrates cumulative prospect theory (CPT) with multi-attribute decision-making to model travel mode choices of stranded passengers under non-mandatory evacuation URT service disruption scenarios. Stated-preference (SP) surveys were designed to simulate varied disruption scenarios, capturing passenger evaluations of alternative modes such as detour URT routes, buses, ride-hailing, and cycling. A baseline EUT model and three CPT-based models with different parameter constraints were calibrated using empirical data. Results show that CPT-based models, especially those applying nonlinear value and probability functions to travel time, significantly outperform EUT models in fitting behavioral patterns. Passengers tend to overestimate low-probability events and underestimate high-probability ones, with heightened sensitivity near the extremes of probability. The value function calibration reveals increasing sensitivity to time gains and diminishing sensitivity to losses, reflecting anxiety-driven urgency and “loss numbness”—a sharp reaction to initial delays that flattens as delays grow. The proposed model effectively captures key psychological mechanisms in disruption scenarios, offering a more behaviorally realistic basis for predicting passenger responses. These insights support more targeted emergency resource deployment and behavior-sensitive management strategies, contributing to enhanced resilience in urban rail systems.
The continuous expansion of the scale of global urban rail transit (URT) network has led to the aggravation of traction energy consumption and the utilization of regenerative braking energy (RBE). Applying energy storage technology in URT can enhance the stability of traction power supply system (TPSS), absorb RBE, reduce traction energy consumption, and bring an opportunity to access new energy. This paper focuses on the flywheel energy storage system (FESS). First, the basic working principle and basic structure combined with the application summary of FESS in URT are analyzed in detail. Second, the technical breakthroughs of flywheel devices that need further improvement are illustrated in terms of rotor, bearing, and converter. Then common energy management and control methods of single flywheel device and flywheel array are summarized and compared. Finally, from the aspects of high-performance development of flywheel unit as well as intelligent energy management of FESS, the future development of FESS is put forward, which provides a reference for the standardization, high efficiency, and simplicity of FESS in URT.
Abstract Insufficient wheel–rail adhesion during braking poses a significant threat to operational safety and accelerates component wear. Although anti-slide valves modulate braking force to mitigate sliding, the resulting speed fluctuations can inadvertently trigger emergency braking commands from the signaling system, potentially exacerbating the slide. Existing detection methods are largely reactive, relying on post-event identification rather than predictive foresight. To bridge this gap, an onboard module for predictive risk warning is proposed. Utilizing real-world vehicle and signaling data, a cross-attention CNN–GRU model is proposed for the accurate prediction of whole-train wheel slide probability. In parallel, a train dynamics model projects the braking trajectory to assess the displacement deviation relative to target positions and safety margins. By integrating the data-driven slide probability with the model-based displacement deviation, a dynamic risk matrix is constructed to assess different levels of train slide risk. With an optimal prediction horizon of 2.5 s and established alarm thresholds, the proposed module provides actionable inputs for the train operation control system. Experimental results demonstrate that the proposed method effectively predicts both wheel sliding and the associated overrun risk. This enables the signaling system to proactively intervene in sliding control, preemptively mitigating sliding occurrences and enhancing overall operational safety.
Abstract Virtual coupling (VC) has attracted increasing research attention as a train control concept for rail systems, enabling trains to operate at very short separations using cooperative, relative distance-based control. This paper reviews how VC affects railway operations across three interconnected layers: operation control, transport organization and system-level performance evaluation. At the control layer, we synthesize research on separation and collision avoidance models, trajectory planning, formation and cooperative control, and uncertainty-aware control methods, covering both model predictive control-based designs and alternative control strategies that address parametric uncertainty, heterogeneous braking, delays and cyber-attacks. At the planning layer, we survey line planning, timetabling, rolling stock circulation, train constitution and rescheduling models that embed VC-enabled headway relaxation, dynamic (de)coupling and platoon composition as decision variables, linking VC to passenger-oriented capacity allocation and disruption management. At the evaluation layer, we review signaling system safety assessment, capacity and headway analysis, and energy and environmental studies that quantify VC impacts under realistic signaling, infrastructure, and disturbance conditions. A thematic–methodological mapping highlights the prevalence of optimization and model predictive control, the limited integration between control and passenger-oriented planning, and the currently limited body of evidence for high-speed and heavy-haul applications. The review identifies consistent modeling patterns, critical gaps, and implementation challenges, and proposes a realistic research roadmap for treating VC as an operational and service innovation rather than a purely signaling upgrade.
Stray current in urban railway systems presents a significant challenge, including localized corrosion of rails, fastening systems, and adjacent metal structures, which compromises infrastructure safety. This study focuses on enhancing the resistance of rail fastening systems to mitigate stray current. Laboratory tests were conducted on samples where rails were fixed using two different fastening systems. During testing, the samples were subjected to 26 VDC, and electrical resistance and potential distribution were measured under controlled conditions. The results indicated that both systems were inadequately insulated, resulting in stray current, although in one system the anchor bolts were insulated primarily to prevent current leakage into the track substructure. Material analysis showed that the elastomeric elements in this system were not specified for high electrical resistance, so stray current was not prevented. Modifications for both systems were proposed to improve stray current prevention and were analyzed using numerical models in COMSOL Multiphysics software. The study highlights the importance of establishing precise requirements for the electrical resistance of fastening systems and developing a standardized methodology for measuring resistance.
Abstract Accurate 3D reconstruction of railway tunnels is crucial for infrastructure maintenance, safety assessment, and digital twin development. However, existing methods often fail in real-world scenarios due to sensor noise and spurious geometry generation in non-structural ground regions—particularly caused by rails, sleepers, and ballast. To address these domain-specific challenges, we propose TunnelSDF-FilterNet, a domain-aware neural signed distance function (Neural SDF) framework explicitly designed for high-fidelity railway tunnel reconstruction. Our approach introduces three key innovations: (1) a robust preprocessing pipeline that integrates DBSCAN–RANSAC clustering with normal and height-based geometric filtering to automatically remove ground artifacts without manual intervention; (2) a tunnel-tailored Neural SDF training objective featuring a novel ground suppression loss to constrain reconstruction within the tunnel envelope and an edge-aware regularization term to preserve fine structural details such as segment joints; and (3) a NeuralPull-based surface refinement strategy applied during inference to achieve sub-voxel precision in surface extraction. Extensive experiments on real-world tunnel datasets demonstrate that TunnelSDF-FilterNet significantly outperforms state-of-the-art methods in both qualitative fidelity and quantitative metrics—including lower Chamfer Distance (CD), higher Normal Consistency (NC), reduced Ground Inclusion Rate (GIR), and improved F-Scores. The proposed framework delivers operationally deployable, high-precision 3D tunnel models, offering a practical AI-driven solution for intelligent railway infrastructure management and digital twin construction in modern railway systems.
Urban air mobility (UAM) leverages advanced air transportation systems to facilitate efficient and reliable air travel within and around urban areas. It involves the use of electric aircraft capable of vertical take-off and landing (VTOL), such as air taxis, which navigate urban airspace to transport passengers quickly between locations. High-speed rail (HSR) service, on the other hand, is a rail transport system that utilizes trains operating at higher speeds than traditional rail services, providing rapid connectivity between major metropolitan areas. In this study, we aim to develop a hub-and-spoke network design (HSND) problem that integrates UAM and HSR (UAMHSR–HSND) to improve the efficiency of multimodal transportation systems. The hub-and-spoke model captures the interfacility interactions between passenger flows and vertiport–station assignments via a semi-quadratic assignment model, which poses significant computational challenges. To address this challenge, we compare different linearization models that leverage the special network configuration of UAMHSR–HSND and allow the problem to be solved using off-the-shelf solvers. We further propose a hybrid solving strategy that uses NET-QAP to generate warm-start solutions and then refine them with stronger linearization models, thereby enhancing scalability for large instances. Numerical results show that the NET-QAP model consistently solves large-scale cases with reasonable optimality gaps, and that integrating it with the multicommodity-flow linearization can obtain optimality (GAP = 0
Smart card data (SCD) from Automatic Fare Collection (AFC) systems provide fine-grained insights into urban mobility rhythms. Yet most existing studies focus on static ridership, with limited attention to the relationship between station-level travel rhythms and the surrounding built environment. This study develops a spatiotemporal and data-driven framework to classify urban rail transit (URT) stations and examine their land-use determinants. Using two weeks of AFC data from 128 stations in Nanjing, China, a two-stage approach is implemented. First, a Gaussian Mixture Model (GMM) is applied to cluster stations based on weekday ridership profiles, yielding six distinct categories: residential oriented, employment oriented, hub comprehensive, spatial mismatched, predominantly residential mixed use, and predominantly employment mixed use. These categories reveal a clear transition from mixed and hub functions in the city center to residential-dominated stations in suburban areas. Second, a Random Parameter Logit (RPL) model is employed to assess the influence of socio-demographic, land-use, and amenity variables, capturing heterogeneity more effectively than the conventional Multinomial Logit (MNL) model. Results highlight the decisive roles of population density, housing prices, and employment land in shaping employment- and hub-oriented stations, while community-oriented facilities, such as healthcare and daily services, exert stronger effects on residential-oriented stations. These findings enrich theoretical understanding of station heterogeneity and provide empirical evidence for transit-oriented development (TOD), land-use coordination, and multimodal integration. The proposed framework is transferable to other rapidly urbanizing cities, offering practical guidance for building efficient and sustainable URT systems.
Real-time prediction of dynamic origin–destination (OD) passenger flows is essential for efficient passenger flow management in urban rail transit (URT) systems. Existing studies have primarily focused on commuting OD flows, which exhibit strong regularity and are supported by abundant data samples. In contrast, non-commuting OD flows—especially those generated by irregular passengers with limited historical data—are characterized by high stochasticity and data sparsity and have received relatively little attention, with existing studies often reporting unsatisfactory predictive performance. To address these challenges, this study proposes a novel real-time OD flow prediction framework for irregular non-commuting passengers through multi-source data fusion and feature extraction. Specifically, individual-level spatiotemporal behavioral features are extracted from metro AFC data using a density-based clustering algorithm. Land-use and geo-economic data are then integrated to characterize individual travel preferences and construct a multidimensional behavioral indicator system. Building upon these features, hierarchical clustering and machine learning models are employed to perform personalized destination prediction. Empirical experiments conducted on Nanjing Metro data demonstrate that the proposed framework substantially improves prediction accuracy for non-commuting passengers and provides new insights into dynamic OD modeling. The results highlight the strong applicability and potential of the method for real-time passenger flow prediction in complex urban rail systems.
Whether urban rail accessibility is distributed equitably has significant implications for social equity, especially in rapidly urbanizing cities. This research analyses how metro accessibility was distributed across different neighborhoods in Hangzhou during 1998-2021. The results indicate that public housing projects tended to be located in areas with lower metro accessibility compared to market housing. The unequitable distribution of metro accessibility between public and market housing appeared to worsen as urban expansion continued. These findings imply that institutional interventions are required to enhance inclusive transit-oriented development.
Analyzing route choice behavior and understanding the heterogeneous impacts on passenger decisions are key to improving transportation service quality in integrated suburban railway and metro networks. While existing studies focus on metro systems using Mixed Logit (MXL) models or clustering methods with Multinomial Logit models, these approaches struggle to capture the diverse factors in travel scenarios and the complex nonlinear relationships in decision-making. This study integrates Latent Dirichlet Allocation with Machine-Learning models like eXtreme Gradient Boosting (XGBoost), Support Vector Machine, and Random Forest to analyze passenger behavior within Shanghai’s composite rail network. It identifies key influencing factors such as socio-economic demographic characteristics, travel purposes, and the express-to-local departure ratio. XGBoost outperforms other models, including the traditional MXL, in predictive performance. The study highlights significant heterogeneity in route choices across different passenger groups, underscoring the need for personalized transportation solutions. Based on these findings, this study proposes the following actionable suggestions for suburban railway in the composite network: In terms of operational optimization, it is proposed to add express train overtaking stations in key commuting corridors and optimize the timetable to reduce transfer waiting times, thereby improving overall travel efficiency. Besides, time-of-day differentiated pricing and combined-ticket discounts are proposed to improve the rationality of the ticket-price structure. Finally, it is recommended to enhance the personalized route recommendation system.
This study investigates the decoupling relationship between meteorological comfort and urban rail transit ridership in China. Daily meteorological data and passenger volume data from 28 major cities were processed to construct a meteorological comfort index using the entropy weighting method, in which precipitation levels were converted into continuous values based on national standards. A decoupling model was then applied to examine the dynamic interaction between weather comfort and transit use. The analysis identifies three classes of decoupling states: Class A, where passenger travel remains stable despite unfavorable weather; Class B, where moderate sensitivity to meteorological variation is observed; and Class C, where travel is strongly influenced by weather conditions. Results show that most cities predominantly fall under Class B, but with notable fluctuations across seasons and regions. The findings highlight that meteorological comfort does not uniformly determine ridership, but instead reveals differentiated patterns of resilience and vulnerability across urban rail systems. This contributes to a deeper understanding of how external environmental factors interact with public transit demand and provides methodological guidance for improving the robustness of transport planning under climate variability.
During rush hours, the capacity of metro in megacities is insufficient to meet the travel demand, resulting in oversaturation and high risk on platform in stations, especially transfer stations. This paper addresses this problem through the joint optimization of some operational interventions, aiming to alleviate passenger overloads while maintaining travel efficiency. To make the model more realistic, the stochastic characteristics of passengers are considered, including the probability distribution of passenger arrival time, inbound and transfer walking times. To provide a high-quality solution for the complex constraint model, three cooperative agents-governing passenger inflow, transfer flows, and train skip-stopping mode-are architected within improved Double Deep Q learning Network (IDDQN) to form a multi-agent reinforcement learning solution. Empirical validation on Beijing Metro Line 13 and Changping Line demonstrates that the multi-agent framework proposed in this paper can eliminate 100% of passenger over-limit flow while reducing the average waiting time of passengers. It also has a significant improvement in reducing stochastic characteristic impact and accelerating convergence.
Abstract Urban rail transit systems contain numerous sharp radius curves where rail fatigue defect occurs frequently. In severe cases, this leads to extensive rail scaling, rail surface fatigue cracks, and intensified wheel-rail vibration and noise. Appropriate rail profiles can effectively optimise wheel-rail contact status and mitigate the deterioration development of wear and wheel/rail fatigue defect. Addressing this issue, this paper combines field testing evaluation and simulation analysis methods to establish a metro vehicle-track coupled dynamics model validated by measured data. Three profile schemes and three simulation scenarios were considered to comprehensively study and analyze the matching performance between sharp radius curve rail profiles and actual wheel profiles from perspectives of wheel-rail contact stress distribution, fatigue and wear indices, and vehicle dynamic performance. Results indicate that the 60N profile consistently exhibits optimal stress control effectiveness under all operating conditions. Compared to the traditional 60 profile condition, normal and tangential contact stress are reduced by up to 64% and 61%, respectively. Relative to the measured profile condition, stress reductions of up to 64% and 45% are achieved at the gauge corner. The 60N profile demonstrates the lowest fatigue index compared to the 60 profile and measured profile, while maintaining stable wear characteristics and improving vehicle lateral stability with reduced bogie acceleration peaks. Based on comprehensive analysis, the 60N profile is recommended as the optimal solution for both original rail profile selection and grinding target profile.
