Scholarlink

<p>Oil demand forecasting is an essential national energy security and economic development issue. With the acceleration of global energy transformation, countries gradually replace traditional energy with new energy to ensure a safe and stable national energy supply. This paper intends to explore the impact of the increase in the supply of new energy, such as wind energy, solar energy, hydropower, and nuclear energy, on the demand for oil imports in the context of energy transformation. Through the analysis of the relationship between the alternation of new and old energy, and the collection of China’s statistical yearbook data, the combination of key indicators is determined using the grey correlation method and machine learning method. A prediction method for oil import demand with the highest prediction accuracy and stability is proposed through the comparative experiments of different models.</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

Oil Import Demand Forecasting Based on Machine Learning Technology Under Energy Transformation Background

  • 期刊论文

  • 发表日期：

    2026年1月1日

查看全文
  • Yi Qin,
  • Yanrong Huang,
  • Yingying Wang,
  • Zhijiang Zhao,
  • Chuanhui Zhu,
  • Jihong Wang,
  • Bo Wei

摘要

Oil demand forecasting is an essential national energy security and economic development issue. With the acceleration of global energy transformation, countries gradually replace traditional energy with new energy to ensure a safe and stable national energy supply. This paper intends to explore the impact of the increase in the supply of new energy, such as wind energy, solar energy, hydropower, and nuclear energy, on the demand for oil imports in the context of energy transformation. Through the analysis of the relationship between the alternation of new and old energy, and the collection of China’s statistical yearbook data, the combination of key indicators is determined using the grey correlation method and machine learning method. A prediction method for oil import demand with the highest prediction accuracy and stability is proposed through the comparative experiments of different models.