Scholarlink

As an important component of electric vehicles (EVs), electric taxis (ETs) will have adverse effects on the power grid when they are charged randomly and in a large scale and disorderly manner. The importance of large language models (LLMs) lies in their ability to simulate and predict complex large-scale systems. Therefore, this paper proposes an ET charging guidance strategy based on LLMs, which comprehensively considers multiple influencing factors such as time, passenger travel rate and traffic flow, and combines the real-time renewable energy supply, power load and voltage over-limit of the regional distribution network to formulate dynamic electricity prices for guidance. Finally, multiple LLM agents are defined, which are responsible for the real-time charging decision judgment of ETs, the selection of charging stations and the decision of charging time. The simulation results show that compared with the traditional guidance strategy, the method proposed in this paper can effectively reduce the impact of charging load on the distribution network, and can increase the daily income of ET by about 5.82%.

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

Electric Taxi Charging Guidance Optimization Strategy Based on Large Language Model Under Road-Network Coupling

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Zekai Jin,
  • Guangfei Geng,
  • Wei Tang,
  • Chengwei Lou,
  • Lu Zhang

摘要

As an important component of electric vehicles (EVs), electric taxis (ETs) will have adverse effects on the power grid when they are charged randomly and in a large scale and disorderly manner. The importance of large language models (LLMs) lies in their ability to simulate and predict complex large-scale systems. Therefore, this paper proposes an ET charging guidance strategy based on LLMs, which comprehensively considers multiple influencing factors such as time, passenger travel rate and traffic flow, and combines the real-time renewable energy supply, power load and voltage over-limit of the regional distribution network to formulate dynamic electricity prices for guidance. Finally, multiple LLM agents are defined, which are responsible for the real-time charging decision judgment of ETs, the selection of charging stations and the decision of charging time. The simulation results show that compared with the traditional guidance strategy, the method proposed in this paper can effectively reduce the impact of charging load on the distribution network, and can increase the daily income of ET by about 5.82%.