Scholarlink

<p>Energy consumption can be decreased by integrating phase change materials (PCM) into building envelopes. However, the correlation between climate conditions and energy savings potential of using PCM in buildings remain unclear and unquantified. This study investigates the energy savings and thermal performance of PCM-integrated buildings in 50 rural areas of China. The heating and cooling energy use intensity (EUI) reduction, PCM liquid fraction and wall inner surface temperature profiles were compared and analyzed in five representative rural areas. A machine learning-based climate classification approach tailored for PCM-integrated buildings was proposed by combining XGBoost and clustering methods. The XGBoost algorithm was employed to identify and rank the influential climate variables by comparing their importance scores. The K-means clustering method was then applied to the key climate variables to enable climate classification reflecting PCM performance. The most substantial heating and cooling EUI reductions occur in the severe cold region at 8.35 kWh/(m<sup>2</sup>·year) and hot summer and warm winter region at 4.31 kWh/(m<sup>2</sup>·year), respectively. Outdoor air temperature and solar irradiance were identified as the most significant factors. Clustering of key climate variables facilitated the evaluation of the effectiveness of using PCM in buildings. The daily average heating energy savings were 0.67, 1.31 and 1.60 kWh/day for different clusters, while the daily average cooling energy savings were −0.25, 0.29 and 1.02 kWh/day.</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

Climate adaptation study of phase change materials in rural buildings in China using a machine learning approach

  • 期刊论文

  • 发表日期：

    2025年11月1日

查看全文
  • Yaping Zhou,
  • Yaotian Hu,
  • Jiangtian Zhu,
  • Jing Luo,
  • Xiaofeng Zhang,
  • Saleh Nasser Al-Saadi,
  • Jie Li,
  • Xiaoqin Sun

摘要

Energy consumption can be decreased by integrating phase change materials (PCM) into building envelopes. However, the correlation between climate conditions and energy savings potential of using PCM in buildings remain unclear and unquantified. This study investigates the energy savings and thermal performance of PCM-integrated buildings in 50 rural areas of China. The heating and cooling energy use intensity (EUI) reduction, PCM liquid fraction and wall inner surface temperature profiles were compared and analyzed in five representative rural areas. A machine learning-based climate classification approach tailored for PCM-integrated buildings was proposed by combining XGBoost and clustering methods. The XGBoost algorithm was employed to identify and rank the influential climate variables by comparing their importance scores. The K-means clustering method was then applied to the key climate variables to enable climate classification reflecting PCM performance. The most substantial heating and cooling EUI reductions occur in the severe cold region at 8.35 kWh/(m2·year) and hot summer and warm winter region at 4.31 kWh/(m2·year), respectively. Outdoor air temperature and solar irradiance were identified as the most significant factors. Clustering of key climate variables facilitated the evaluation of the effectiveness of using PCM in buildings. The daily average heating energy savings were 0.67, 1.31 and 1.60 kWh/day for different clusters, while the daily average cooling energy savings were −0.25, 0.29 and 1.02 kWh/day.