Scholarlink

Fused magnesia is an important strategic material. During the fused magnesia smelting process (FMSP), if the electricity demand exceeds the specified limit, one or more of the electric furnaces may be shut down to manage the load. However, during the FMSP, there is a phenomenon of demand spikes, where the demand briefly exceeds the limit and then quickly drops back below the limit. In such cases, there is no need to shut down the electric furnaces. Therefore, accurately forecasting the electricity demand is crucial for improving the quality of fused magnesia products and reducing production costs. FMSP is crucial for reducing production costs and enhancing overall production efficiency. In this paper, we adopted a demand forecasting model for the FMSP based on the long short-term memory (LSTM) network and the attention mechanism (Attention). We propose an intelligent forecasting method for demand. The effectiveness of the proposed method was validated through real data from the FMSP.

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

Electricity Demand Forecasting for FMSP Using LSTM and Attention Mechanisms

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Jingwen Zhang,
  • Xiaoran Yang,
  • Qinwei Zhang

摘要

Fused magnesia is an important strategic material. During the fused magnesia smelting process (FMSP), if the electricity demand exceeds the specified limit, one or more of the electric furnaces may be shut down to manage the load. However, during the FMSP, there is a phenomenon of demand spikes, where the demand briefly exceeds the limit and then quickly drops back below the limit. In such cases, there is no need to shut down the electric furnaces. Therefore, accurately forecasting the electricity demand is crucial for improving the quality of fused magnesia products and reducing production costs. FMSP is crucial for reducing production costs and enhancing overall production efficiency. In this paper, we adopted a demand forecasting model for the FMSP based on the long short-term memory (LSTM) network and the attention mechanism (Attention). We propose an intelligent forecasting method for demand. The effectiveness of the proposed method was validated through real data from the FMSP.