Scholarlink

<p>The rapid adoption of electric vehicles (EVs) has intensified the demand for accessible and efficient charging infrastructure. However, disparities in the spatial and temporal distribution of EV chargers remain a critical obstacle, often leading to congestion during peak hours and under utilization during off-peak periods. Fixed charging stations (FCSs) serve as the backbone of public charging infrastructure, but their static nature limits their ability to adapt to dynamic demand. Mobile charging stations (MCSs) provide a flexible alternative, yet existing studies primarily focus on their standalone deployment, with limited consideration of real-world compatibility with FCS networks. To address these challenges, this paper introduces the <b>MAC-DP</b> framework, which leverages multi-agent reinforcement learning (MARL) to optimize the dynamic placement of MCSs in coordination with FCSs. MAC-DP incorporates novel action selection strategies and reward formulations to enhance charging station utilization and reduce EV users’ travel and waiting times. Experiments conducted on real-world datasets demonstrate that MAC-DP outperforms baseline approaches by minimizing travel and waiting times for EV users and balancing utilization variance among all charging stations. These results underscore the potential of MAC-DP as a practical solution for hybrid deployments of FCSs and MCSs, offering benefits for both EV users and charging station operators.</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

Mac-dp: multi-agent control for dynamic placement of electric vehicle charging stations

  • 期刊论文

  • 发表日期：

    2025年6月20日

查看全文
  • Lo Pang-Yun Ting,
  • You-Cheng Guo,
  • Chi-Chun Lin,
  • Shih-Hsun Lin,
  • Kun-Ta Chuang

摘要

The rapid adoption of electric vehicles (EVs) has intensified the demand for accessible and efficient charging infrastructure. However, disparities in the spatial and temporal distribution of EV chargers remain a critical obstacle, often leading to congestion during peak hours and under utilization during off-peak periods. Fixed charging stations (FCSs) serve as the backbone of public charging infrastructure, but their static nature limits their ability to adapt to dynamic demand. Mobile charging stations (MCSs) provide a flexible alternative, yet existing studies primarily focus on their standalone deployment, with limited consideration of real-world compatibility with FCS networks. To address these challenges, this paper introduces the MAC-DP framework, which leverages multi-agent reinforcement learning (MARL) to optimize the dynamic placement of MCSs in coordination with FCSs. MAC-DP incorporates novel action selection strategies and reward formulations to enhance charging station utilization and reduce EV users’ travel and waiting times. Experiments conducted on real-world datasets demonstrate that MAC-DP outperforms baseline approaches by minimizing travel and waiting times for EV users and balancing utilization variance among all charging stations. These results underscore the potential of MAC-DP as a practical solution for hybrid deployments of FCSs and MCSs, offering benefits for both EV users and charging station operators.