Scholarlink

Ensuring the safe operation of the power supply system is crucial in urban rail transit system. In a direct current (DC) subway system, the running rail is the return path for the traction current. Therefore, the running rail must be insulated from the ground to prevent current leakage and safety hazards. However, due to the long subway lines, high humidity and metal dust in the tunnels, ground faults in the runway are a risk. A method is proposed in this paper for diagnosing running rail grounding faults equipped with One Dimension Efficient Channel Attention based on ShuffleNet (1DECA-SN). Firstly, a dynamic simulation platform is established to generate rail potential (RP) data for different ground fault locations. Secondly, ground fault diagnosis using ShuffleNet network is proposed, which can effectively identify the locations of ground faults existing in the subway system. Employing the suggested diagnostic method, the dataset was tested underneath multi-train and two-train operating conditions. In this way, the proposed method's effectiveness has been proven. The implementation of this technology will facilitate the secure and reliable operation of subway systems, thereby enhancing the transportation network for passengers and operators alike.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Running Rail Ground Fault Diagnosis Based on 1DECA-SN in Direct Current Subway System

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Chen Zidan,
  • Li Qiaoyue,
  • Xia Yaoda,
  • Kang Ziyuan,
  • Du Guifu,
  • Ge Yuchen

摘要

Ensuring the safe operation of the power supply system is crucial in urban rail transit system. In a direct current (DC) subway system, the running rail is the return path for the traction current. Therefore, the running rail must be insulated from the ground to prevent current leakage and safety hazards. However, due to the long subway lines, high humidity and metal dust in the tunnels, ground faults in the runway are a risk. A method is proposed in this paper for diagnosing running rail grounding faults equipped with One Dimension Efficient Channel Attention based on ShuffleNet (1DECA-SN). Firstly, a dynamic simulation platform is established to generate rail potential (RP) data for different ground fault locations. Secondly, ground fault diagnosis using ShuffleNet network is proposed, which can effectively identify the locations of ground faults existing in the subway system. Employing the suggested diagnostic method, the dataset was tested underneath multi-train and two-train operating conditions. In this way, the proposed method's effectiveness has been proven. The implementation of this technology will facilitate the secure and reliable operation of subway systems, thereby enhancing the transportation network for passengers and operators alike.