Quick Entry

  • Author Login Author Login
  • Expert Center Expert Center
  • Editor Center Editor Center
  • Legacy system entry Legacy system entry

Quick Icon

WeChat
  • Current Issue
  • Articles in Press
  • Back Issue Browse
更多
Volume  期 5，2025 2025年第卷第5期

  • Special Issue on Autonomous Safety Assurance and Intelligent Operation and Maintenance of Rail Transit

    TANG Jun, HUANG Wenjing, ZOU Zhirong, LI Siyuan, TAO Jiyu

    DOI：10.13890/j.issn.1000-128X.2025.05.001
    摘要：With the deep integration of new-generation information technologies such as artificial intelligence, big data, digital twins, and the internet of things, the rail transit industry is accelerating its evolution toward networked and intelligent operations to enhance transport efficiency and service quality. However, open network interconnections and multi-system collaboration have eroded the security boundaries of traditional closed systems, exposing rail transit systems to increasingly severe cybersecurity risks. From the perspective of technology evolution in rail transit cybersecurity, this paper presents a systematic review of the current status of relevant industry standards and technologies, and details cybersecurity activities and risk identification processes across the system lifecycle. It identifies and classifies key protection technologies and summarizes key mainstream techniques currently implemented, including firewalls, security audit, intrusion detection, identity authentication, access control, and blockchain technology, along with a concise analysis of their applications.  
    关键词：rail transit;cybersecurity;IEC 63452;risk identification;security protection;firewall;artificial intelligence;high-speed train   
    620
    |
    93
    |
    0
    <HTML>
    <L-PDF><WORD><Meta-XML>
    <引用本文> <批量引用> 149947754 false
    更新时间：2026-02-24

    SHEN Zelin, WANG Biao, QIN Yong, CHENG Xiaoqing, WANG Zhipeng, WANG Li

    DOI：10.13890/j.issn.1000-128X.2025.05.002
    摘要：Diffusion models enable the generation of virtual fault samples, alleviating issues such as scarcity and imbalance in measured data and showing promising applications in intelligent fault diagnosis of train gearboxes. However, existing methods have two main limitations. First, they ignore correlations between a priori physical information and sample features, hindering use of condition differences to guide the generation process, and resulting in insufficient diversity of generated samples. Second, they neglect the differential contributions of time steps to the diffusion model optimization process, wasting computing resources on low-contribution time steps, and causing low training efficiency and slow convergence rate of the model. To address these limitations, this paper proposes a virtual sample augmentation method for fault diagnosis of train gearboxes based on a dual-condition diffusion model. Within the diffusion-model framework, this method used U-Net as the main architecture for reverse denoising, and integrated an attention mechanism with a dual-condition encoder. This design enables the model to dynamically guide the generation process based on conditions during the denoising stage and to suppress the generation of non-ideal samples by introducing template conditions, thereby enhancing the diversity and distributional completeness of generated data. In addition, a time-step optimization sampling module was introduced to dynamically evaluate and screen time steps in the diffusion process based on their importance, focusing learning on those with greater contributions to denoising to improve training efficiency and convergence rate of the model. By combining a priori information of conditions with adaptive optimization of time steps, the model efficiently generates diverse virtual fault samples. Experimental results demonstrate that the proposed method generates a richer set of fault samples, significantly improves accuracy in fault diagnosis tasks, and achieves superior efficiency of model training to existing methods.  
    关键词：train gearbox;fault diagnosis;virtual sample augmentation;dual-condition diffusion model;time step optimization;metro train;artificial intelligence   
    483
    |
    46
    |
    0
    <HTML>
    <L-PDF><WORD><Meta-XML>
    <引用本文> <批量引用> 149947404 false
    更新时间：2026-02-24

    SUN Bo, LIU Bin

    DOI：10.13890/j.issn.1000-128X.2025.05.003
    摘要：The Lanzhou-Xinjiang passenger dedicated line traverses four major wind zones in Xinjiang, namely Yandun, Baili, Sanshili, and Dabancheng. Sections exposed to strong-wind impact along this line total 462.41 km, accounting for 65% of its total length within Xinjiang. The average annual number of strong-wind days in these zones exceeds 75, and the maximum recorded wind speed reaches 46.7 m/s, posing a serious threat to the safe operation of this railway. In response to the lack of strong-wind trend forecast capabilities in the existing disaster prevention system, this paper proposes a refined strong-wind trend forecast system. Based on 2018—2024 data from multiple monitoring stations, a probabilistic strong-wind forecast system was developed by integrating a multi-model forecast method with bilinear interpolation and the least squares method for parameter estimation. This system provides 24-hour strong-wind trend forecast at 1-hour resolution and supports data release based on fusion of forecasts and real-time monitoring. Experimental results revealed the highest average monthly forecast accuracy in September, reaching 83.0%, and the highest station-level average forecast accuracy at K3021+217, reaching 86.2%. Furthermore, the relationship between extreme wind speeds and alarm levels was statistically analyzed, and a strong-wind forecast data application strategy based on alarm rules was proposed. This strategy shortens unnecessary shutdown time of trains, improves transport efficiency, and provides important support for the safe operation of railways in strong-wind zones.  
    关键词：Lanzhou-Xinjiang passenger dedicated line;trend forecast;refined;strong wind alarming;application strategy;high-speed railway   
    410
    |
    52
    |
    0
    <HTML>
    <L-PDF><WORD><Meta-XML>
    <引用本文> <批量引用> 149947660 false
    更新时间：2026-02-24

    CHANG Zhaorui, LIANG Shulin, CHI Maoru, LI Zhenqian, HUO Wenbiao, PENG Lei

    DOI：10.13890/j.issn.1000-128X.2025.05.004
    摘要：To enhance the operational quality and safety of electric multiple units (EMUs), this study investigates parameter optimization for their underfloor suspension dampers based on vertical dynamic performance and addresses service state management. A rigid-flexible coupled vehicle dynamic model was established, which accounts for the elastic vibration characteristics of car bodies to accurately simulate relevant behaviors. On this basis, the influence patterns of design parameters on the vehicles' operational quality were systematically analyzed, from aspects such as center-of-mass offsets resulting from the structural layout of suspension dampers, damping ratio matching in resonance models, and damping characteristics. From the analysis results, an optimized matching scheme for key suspension parameters was derived. To reveal the service state of these dampers, an in-depth analysis was conducted to examine the effects of performance degradation and typical faults of key components on the vehicles' vertical dynamic performance, which verified the optimized scheme provides sufficient safety margins to accommodate performance degradation. Research results show that parameter optimization significantly improves the vehicles' vertical dynamic performance and underscore the need for effective performance degradation monitoring during service. This study provides a theoretical basis for refined parameter design and performance optimization of underfloor suspension dampers for high-speed trains, with important implications for ensuring the long-term safe and stable operation of EMUs.  
    关键词：high-speed trains;underfloor suspension dampers;vertical vibration;rigid-flexible coupling model;modal analysis;high-speed railway;high-speed EMU;finite element   
    452
    |
    55
    |
    0
    <HTML>
    <L-PDF><WORD><Meta-XML>
    <引用本文> <批量引用> 149947500 false
    更新时间：2026-02-24
查看更多
  • Column Album
更多
Hydrogen energy rail transit system column

Hydrogen energy rail transit system column

Journal of low carbon transport

Journal of low carbon transport

Power semiconductor technology

Power semiconductor technology

Journal of world transport convention

Journal of world transport convention

Journal of serialized Chinese standard subway trains

Journal of serialized Chinese standard subway trains

Journal of power semiconductors

Journal of power semiconductors

60th anniversary column

60th anniversary column

high-speed railway column

high-speed railway column

查看更多
0