Scholarlink

With the increase of electric vehicle penetration, more and more electric vehicles are traveling on highways, which brings great challenges to the construction of charging stations in highway service areas and the safe and stable operation of power distribution networks. Aiming at the difficult problem of electric vehicle charging demand prediction in highway service areas, this paper proposes a charging station load prediction method based on LSTM-Seq2Seq for highway service areas. In this paper, based on the operation data of charging piles in a service area in Jiangsu, China, a highway service area charging station load prediction model is established based on Long Short-Term Memory (LSTM) network and Seq2Seq (Sequence to Sequence, Seq2Seq). Finally, the algorithm proposed in this paper is compared with other prediction algorithms, and compared with the LSTM model, its mean absolute error (MAE) decreases by 4.48%, and the root mean square error (RMSE) decreases by 7.79%, which further validates the superiority of the algorithm.

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

Load Prediction of Charging Stations in Highway Service Areas Based on LSTM-Seq2Seq

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Zongwei Zhang,
  • Liang Dai,
  • Hongke Xu,
  • Hongsheng Xia,
  • Jiwei Li

摘要

With the increase of electric vehicle penetration, more and more electric vehicles are traveling on highways, which brings great challenges to the construction of charging stations in highway service areas and the safe and stable operation of power distribution networks. Aiming at the difficult problem of electric vehicle charging demand prediction in highway service areas, this paper proposes a charging station load prediction method based on LSTM-Seq2Seq for highway service areas. In this paper, based on the operation data of charging piles in a service area in Jiangsu, China, a highway service area charging station load prediction model is established based on Long Short-Term Memory (LSTM) network and Seq2Seq (Sequence to Sequence, Seq2Seq). Finally, the algorithm proposed in this paper is compared with other prediction algorithms, and compared with the LSTM model, its mean absolute error (MAE) decreases by 4.48%, and the root mean square error (RMSE) decreases by 7.79%, which further validates the superiority of the algorithm.