Scholarlink

<p>The next-generation heating systems, crucial for rational heat distribution and refined management, rely heavily on accurate zone-specific heat load predictions. This paper introduces a method for rapid zone-specific heat load prediction based on heat consumption allocation and data-driven techniques. The approach involves predicting the overall heat load of the building and then redistributing the total heat according to a heat consumption matrix. This eliminates the need for real-time data collection from each room, resulting in cost savings on hardware and improved computational efficiency. The overall building heat load data is obtained through a data-driven algorithm, while the heat consumption matrix is constructed through energy software simulation analysis. Using Building 2 in the Baotou Industrial Park, China, as a case study, the paper analyzes the differences between actual measurements and room estimates. Experimental results indicate an average error of 7.02% for the proposed estimation method. Although not achieving high precision (&gt;95%) in heat load prediction, this level of accuracy is deemed sufficient to meet the requirements of feedforward control.</p>

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

Heat Load Prediction of Building Rooms Using Only the Whole Building Data via Heat Allocation Approach

  • 期刊论文

  • 发表日期：

    2025年3月15日

查看全文
  • Xin Tan,
  • Yahui Wang,
  • Guoxin Sun,
  • Linfeng Wu,
  • Qihui Yu,
  • Yongheng Yu

摘要

The next-generation heating systems, crucial for rational heat distribution and refined management, rely heavily on accurate zone-specific heat load predictions. This paper introduces a method for rapid zone-specific heat load prediction based on heat consumption allocation and data-driven techniques. The approach involves predicting the overall heat load of the building and then redistributing the total heat according to a heat consumption matrix. This eliminates the need for real-time data collection from each room, resulting in cost savings on hardware and improved computational efficiency. The overall building heat load data is obtained through a data-driven algorithm, while the heat consumption matrix is constructed through energy software simulation analysis. Using Building 2 in the Baotou Industrial Park, China, as a case study, the paper analyzes the differences between actual measurements and room estimates. Experimental results indicate an average error of 7.02% for the proposed estimation method. Although not achieving high precision (>95%) in heat load prediction, this level of accuracy is deemed sufficient to meet the requirements of feedforward control.