Scholarlink

Load forecasting plays a pivotal role in industrial demand response, enabling businesses to plan their electricity needs ahead of time through day-ahead scheduling. However, data is often limited or outdated due to frequent infrastructure modifications. To this end, load forecasting using limited data has recently attracted research interest. This paper explores the application of learning-based algorithms to day-ahead load forecasting in data-constrained environments. Moreover, we introduce Diff-Ensemble, an ensemble incorporating the long short-term memory (LSTM) and diffusion model, to enhance load forecasting capabilities when data is limited, and evaluate it using real-world data from an industrial site as part of the InStaFlex project. Results show that Diff-Ensemble reduces the normalized MAE (NMAE) by 8.1% and 18.5% compared to the LSTM and diffusion model, respectively. Furthermore, with just seven days of training, Diff-Ensemble achieves an NMAE of 2.77%, outperforming all other methods in this study. This work demonstrates the viability of ensemble learning for load forecasting with limited data.

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

Diff-Ensemble: An Ensemble of LSTMs and Diffusion Models for Day-Ahead Load Forecasting Using Limited Data

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Stijn Van Raemdonck,
  • Joris Van den Bergh,
  • Brecht Zwaenepoel,
  • Tomas Van Oyen,
  • Dieter Van den Bleeken,
  • Hossein Tabari,
  • Peter Hellinckx

摘要

Load forecasting plays a pivotal role in industrial demand response, enabling businesses to plan their electricity needs ahead of time through day-ahead scheduling. However, data is often limited or outdated due to frequent infrastructure modifications. To this end, load forecasting using limited data has recently attracted research interest. This paper explores the application of learning-based algorithms to day-ahead load forecasting in data-constrained environments. Moreover, we introduce Diff-Ensemble, an ensemble incorporating the long short-term memory (LSTM) and diffusion model, to enhance load forecasting capabilities when data is limited, and evaluate it using real-world data from an industrial site as part of the InStaFlex project. Results show that Diff-Ensemble reduces the normalized MAE (NMAE) by 8.1% and 18.5% compared to the LSTM and diffusion model, respectively. Furthermore, with just seven days of training, Diff-Ensemble achieves an NMAE of 2.77%, outperforming all other methods in this study. This work demonstrates the viability of ensemble learning for load forecasting with limited data.