Scholarlink

This study proposes a prediction model for elevator energy consumption parameters by integrating LSTM with a multi-head attention mechanism. Instead of directly forecasting total energy usage, the model targets key operational indicators such as daily running time and travel distance, which are closely related to energy consumption. The LSTM captures long-term dependencies in time series data, while the attention mechanism assigns adaptive weights to different features, improving information extraction and fusion. Experimental results show that the proposed method surpasses a standard LSTM model in both accuracy and stability, providing reliable predictions to support energy-efficient elevator operation and management.

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

Energy Consumption Parameter Prediction Method Based on LSTM and Multi-head Attention Mechanism

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mu Yuan,
  • Gang Xiao,
  • Maoyu Wang,
  • Yanfang Huang,
  • Xia Zhang,
  • Yue Lang,
  • Jun Chen,
  • Zhenbo Cheng

摘要

This study proposes a prediction model for elevator energy consumption parameters by integrating LSTM with a multi-head attention mechanism. Instead of directly forecasting total energy usage, the model targets key operational indicators such as daily running time and travel distance, which are closely related to energy consumption. The LSTM captures long-term dependencies in time series data, while the attention mechanism assigns adaptive weights to different features, improving information extraction and fusion. Experimental results show that the proposed method surpasses a standard LSTM model in both accuracy and stability, providing reliable predictions to support energy-efficient elevator operation and management.