Scholarlink

<p>The global metro system is experiencing rapid development due to the benefits of subway transportation as a clean mode of travel and its expansion in many cities worldwide. Therefore, reducing energy consumption and minimizing voltage fluctuations in the contact network have become urgent issues. Aiming to address these challenges, a control strategy for the output voltage of metro line substations, based on fully controlled rectifiers, is proposed. This strategy uses an intelligent agent model for the substation voltage source converter, incorporating a reward function designed to optimize energy conservation and voltage stabilization in the catenary system. A traction power system simulator is used to emulate the natural environment, and the agent's actions are refined through each training cycle using the twin-delayed deep deterministic policy gradient (TD3) algorithm. The proposed strategy was validated through simulations based on the Guangzhou Subway Line 6, demonstrating its capability to dynamically adjust substation rectifier voltage for highly efficient distribution of regenerative braking energy over time. Results indicate considerable improvements in energy savings and voltage stabilization when compared to multi-pulse rectifiers (diode rectifier unit) and constant voltage control strategies.</p>

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

Intelligent control strategy for subway overhead line substation outputs based on controllable rectifiers

  • 期刊论文

  • 发表日期：

    2025年4月18日

查看全文
  • Hongyi Li,
  • Jiyuan Li,
  • Yujie Hu,
  • Tongshuo Yang,
  • Qinyue Zhu

摘要

The global metro system is experiencing rapid development due to the benefits of subway transportation as a clean mode of travel and its expansion in many cities worldwide. Therefore, reducing energy consumption and minimizing voltage fluctuations in the contact network have become urgent issues. Aiming to address these challenges, a control strategy for the output voltage of metro line substations, based on fully controlled rectifiers, is proposed. This strategy uses an intelligent agent model for the substation voltage source converter, incorporating a reward function designed to optimize energy conservation and voltage stabilization in the catenary system. A traction power system simulator is used to emulate the natural environment, and the agent's actions are refined through each training cycle using the twin-delayed deep deterministic policy gradient (TD3) algorithm. The proposed strategy was validated through simulations based on the Guangzhou Subway Line 6, demonstrating its capability to dynamically adjust substation rectifier voltage for highly efficient distribution of regenerative braking energy over time. Results indicate considerable improvements in energy savings and voltage stabilization when compared to multi-pulse rectifiers (diode rectifier unit) and constant voltage control strategies.