Scholarlink

<p>The randomness and uncertainty of electric vehicle (EV) users’ charging behavior pose significant challenges to the prediction of EV charging load and the stable operation of the power grid. There is a problem of the influence of multiple factors, and so a prediction method of charging load of electric vehicle charging stations based on an improved dung beetle algorithm-support vector machine (IDBO-SVM) model is proposed. Firstly, the dung beetle algorithm is optimized by introducing Chebyshev chaotic mapping, the golden sine algorithm, and dynamically updated position weight coefficients. The advantages of the IDBO algorithm are validated using six test functions. Next, a feature input matrix is constructed using influencing factors such as holidays, weather conditions, and electricity prices, along with historical load data, to build the IDBO-SVM model. Finally, based on the actual charging load data from multiple EV charging stations in Nanchong city, Sichuan province, China, simulation experiments are conducted to compare and validate the accuracy and effectiveness of the model.</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

Research on Electric Vehicle Charging Load Forecasting Based on IDBO-SVM

  • 期刊论文

  • 发表日期：

    2025年9月19日

查看全文
  • Bowen Tang,
  • Ruihong Zhu,
  • Mingyang Wang,
  • Xinyu Fan,
  • Xinyu Zhang

摘要

The randomness and uncertainty of electric vehicle (EV) users’ charging behavior pose significant challenges to the prediction of EV charging load and the stable operation of the power grid. There is a problem of the influence of multiple factors, and so a prediction method of charging load of electric vehicle charging stations based on an improved dung beetle algorithm-support vector machine (IDBO-SVM) model is proposed. Firstly, the dung beetle algorithm is optimized by introducing Chebyshev chaotic mapping, the golden sine algorithm, and dynamically updated position weight coefficients. The advantages of the IDBO algorithm are validated using six test functions. Next, a feature input matrix is constructed using influencing factors such as holidays, weather conditions, and electricity prices, along with historical load data, to build the IDBO-SVM model. Finally, based on the actual charging load data from multiple EV charging stations in Nanchong city, Sichuan province, China, simulation experiments are conducted to compare and validate the accuracy and effectiveness of the model.