Abstract Child passenger safety is a critical public health concern, particularly in low- and middle-income countries where road traffic injuries are a leading cause of child mortality. Evidence suggests that child restraint systems (CRS) reduce the likelihood of children being injured or killed in road traffic crashes, yet CRS uptake remains low in many countries, including Pakistan. This study examines socio-demographic, experiential, and contextual factors associated with CRS use at the individual level in Lahore, Pakistan. Using survey data, a machine learning framework with nested cross-validation and hyperparameter tuning using Optuna was applied to evaluate six models: Logistic Regression, Decision Tree, Random Forest, XGBoost, LightGBM, and CatBoost. Among these, CatBoost achieved the highest mean outer-fold ROC-AUC and was further analyzed using SHAP for interpretability, although its performance was statistically indistinguishable from Random Forest. For the CRS-Yes class, CatBoost achieved a precision of 0.841 and a recall of 0.632. SHAP analysis indicated that education level, attending driving school, driving experience, minor past crash involvement, and household members contributed most to the model’s CRS-use predictions. These associations may help identify variables for further investigation and support the development of more targeted hypotheses for CRS awareness and policy research.
Abstract This study investigates the aerodynamic responses of the high-speed maglev train traveling at 600 km/h under strong crosswinds, particularly during tunnel entry and exit. While prior studies have addressed aerodynamic stability in conventional rail systems, few studies have examined such effects on maglev trains, which are highly sensitive to lateral forces due to their electromagnetic suspension and narrow levitation gap. Using computational fluid dynamics (CFD) based on the Reynolds-Averaged Navier–Stokes (RANS) equations and the k−ω Shear Stress Transport (SST) turbulence model, unsteady flow behaviours at a crosswind speed of 33.7 m/s, a threshold outlined in Chinese high-speed train standards, are simulated in this study. The results reveal substantial transient aerodynamic loads and asymmetric surface pressure fluctuations during tunnel portal interactions, leading to significant side-force variations and potential stability risks. This study provides new insights into the safety-critical aerodynamic environment of future ultra-high-speed maglev operations and emphasizes the need for aerodynamic design optimization under extreme weather scenarios.
Abstract As the operating speed of high-speed trains increases, the proportion of aerodynamic drag in total drag gradually rises. Currently, employing micro-structured units on the train surface represents a novel approach for drag reduction in high-speed trains. The cowcatcher at the bottom of the head car is significantly affected by airflow impact. The highly turbulent vortex structures generated there influence downstream components. However, applying micro-structured units across the train surface consumes substantial resources. Therefore, a roughness function is adopted to represent these micro-structured units. Numerical simulations are conducted to study the influence of cowcatcher surface roughness heights (0.5, 0.7, and 1.0 mm) on the aerodynamic characteristics of the train underbody. The results indicate that variations in cowcatcher surface roughness height significantly affect the underbody flow field. The rough surface reduces the underbody flow velocity, consequently decreasing the pressure distribution on downstream components, particularly the pressure intensity on the windward plates of the bogie region. Regarding aerodynamic drag, the drag reduction rates in the Bogie 1 region are 15.13%, 19.29%, and 18.69% for the different roughness heights, respectively. The optimal drag reduction effect in the Bogie 1 region occurs at a roughness height of 0.7 mm. This is attributed to the roughness height in the cowcatcher region reducing the velocity of airflow entering from the cowcatcher bottom, thereby mitigating the impact of high-speed airflow on the downstream bogie region and consequently reducing its aerodynamic drag. The findings of this study provide guidance for future research on higher-speed trains.
Abstract Remote-sensing-based slope stability monitoring is vital for railway safety. This study is grounded in the demands for high timeliness and multi-element monitoring of three-dimensional deformation of modern railway slopes. Aiming at the issue that existing technologies struggle to balance spatial and temporal resolution in the process of fusing and resolving three-dimensional deformation using the BeiDou Navigation Satellite System (BDS) and Interferometric Synthetic Aperture Radar (InSAR), a point-surface hierarchical fusion strategy is proposed. After constructing a unified fused observation equation, spatial regularization constraints and local joint adjustment are introduced to balance spatiotemporal resolution from two levels. This ensures the consistency and reliability of 3D accuracy under the hierarchical strategy, while achieves constrained representation of the large-scale deformation trend and high-precision temporal evolution of the 3D deformation field. Building on the hierarchical fusion results, this paper proposes a 3D deformation risk classification method for railway slopes, enabling refined slope monitoring and early disaster warning with a feasible technical scheme. The effectiveness of the method is validated on typical slopes along the Shijiazhuang-Taiyuan Railway. The results indicated that (1) the average 3D deformation accuracy obtained by the point–surface hierarchical fusion strategy is superior to 10 mm, ensuring both accuracy balance and reliability under the hierarchical strategy while taking into account both spatial and temporal resolutions. (2) The proposed three-dimensional deformation risk classification method for slopes preliminarily verifies the feasibility of the hierarchical fusion strategy for 3D deformation monitoring of railway slopes considering both spatial and temporal resolutions.
Abstract Electric vehicle battery packs consist of numerous cells, and the heat-generation rate varies with cell positions. Conventional liquid-cooling plates are generally unable to provide targeted cooling, resulting in poor temperature uniformity across the battery pack. Based on the influence of cell temperature rise and the characteristic parameters of liquid-cooling plates on cell heat-transfer performance, this study proposes a flow-channel calculation method for achieving targeted heat-dissipation performance. The core innovation is a regionalized design method that accounts for the different cooling needs of individual battery cells and uses a bionic vein structure to adaptively match the flow-channel width. Inspired by the hierarchical venation of heat-resistant plants, a novel vein-like liquid-cooling plate for battery packs is developed, and a heat-transfer experimental platform is built to calibrate the proposed numerical model. The heat-dissipation performance of the proposed liquid-cooling plate is then compared with that of a traditional parallel-channel liquid-cooling plate through numerical simulations, and the mechanism by which the flow-channel angle affects the overall performance of the plate is investigated. The results show that a smaller flow-channel angle reduces vortices at the junction between the main and branch channels, mitigates flow separation and decreases the pressure drop in the channels. The proposed bionic liquid-cooling plate exhibits low flow resistance and good temperature uniformity, maintaining the temperature difference of the battery pack within 3 K. This study offers theoretical support for low-energy thermal management in new-energy battery packs.
