Scholarlink

By implementing a comprehensive energy management system, the integrated energy system (IES) can effectively improve and optimize the allocation of multiple renewable energy sources, playing an important role in the process of energy transformation. However, IES load forecasting is a difficult task because of the strong randomness and fluctuation of multi-energy loads in IES. This paper proposes a GWO-SVR-LSTM model which can predict different kinds of loads in IES with high accuracy. Firstly, we use GWO to optimize the hyper-parameters of SVR. Then, the IES load sequences are input into the LSTM network and GWO-SVR for prediction respectively. Finally, the prediction values of the 2 models are combined by GWO-SVR to obtain the final load-forecasting results. Through a lot of experiments, our proposed model is proved to have higher accuracy than other comparison models.

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

Short-Term Load-Forecasting of Integrated Energy System Based on GWO-SVR-LSTM

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Zhenyu Wang,
  • Qingyong Zhang

摘要

By implementing a comprehensive energy management system, the integrated energy system (IES) can effectively improve and optimize the allocation of multiple renewable energy sources, playing an important role in the process of energy transformation. However, IES load forecasting is a difficult task because of the strong randomness and fluctuation of multi-energy loads in IES. This paper proposes a GWO-SVR-LSTM model which can predict different kinds of loads in IES with high accuracy. Firstly, we use GWO to optimize the hyper-parameters of SVR. Then, the IES load sequences are input into the LSTM network and GWO-SVR for prediction respectively. Finally, the prediction values of the 2 models are combined by GWO-SVR to obtain the final load-forecasting results. Through a lot of experiments, our proposed model is proved to have higher accuracy than other comparison models.