Scholarlink

With the advancement of smart grids, the acquisition of power loads has gradually diversified, and the accuracy of traditional power load forecasting no longer meets the demand. DBO algorithm has a faster convergence speed and higher stability. Based on LSTM-Attention prediction model, this paper uses DBO algorithm to optimize the parameters of the model, and proposes a model based on DBO-LSTM-Attention to more accurately predict future load. Taking the power load data of a certain region in China in 2018 as the data set, and taking humidity, temperature, wind speed, air pressure, precipitation, visibility, water pressure, perceived temperature and historical load values as input features, the prediction results of DBO-LSTM-Attention model are compared with those of other models. Among them, DBO-LSTM-Attention prediction model has the best effect, which proves that the introduction of DBO algorithm to optimize LSTM model parameters greatly improves the prediction accuracy of the power load forecasting model.

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

Power Load Forecasting Based on DBO-LSTM-Attention Model

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Shan Wang,
  • Ruiying Li,
  • Jingyi Yan,
  • Haoruo Sun,
  • Qiyuan Cui,
  • Zhibin Tan,
  • Dong Hu

摘要

With the advancement of smart grids, the acquisition of power loads has gradually diversified, and the accuracy of traditional power load forecasting no longer meets the demand. DBO algorithm has a faster convergence speed and higher stability. Based on LSTM-Attention prediction model, this paper uses DBO algorithm to optimize the parameters of the model, and proposes a model based on DBO-LSTM-Attention to more accurately predict future load. Taking the power load data of a certain region in China in 2018 as the data set, and taking humidity, temperature, wind speed, air pressure, precipitation, visibility, water pressure, perceived temperature and historical load values as input features, the prediction results of DBO-LSTM-Attention model are compared with those of other models. Among them, DBO-LSTM-Attention prediction model has the best effect, which proves that the introduction of DBO algorithm to optimize LSTM model parameters greatly improves the prediction accuracy of the power load forecasting model.