Scholarlink

<p>This study evaluates the potential of ChatGPT for hotspot detection by comparing its performance with the Geographical Information System (GIS) software to detect temperature outliers in a photovoltaic (PV) cell. ChatGPT demonstrated similar performance to ArcGIS on z-score statistics and spatial distribution of hot and cold spots, but it exhibited a more sensitive detection capability. This study underscores the promise of ChatGPT in hotspot detection, with a high correlation of z-score statistics (R<sup>2</sup>: above 0.99) and high similarity of spatial distribution (70–97%) to ArcGIS. The preliminary results from our comparative study indicate that ChatGPT holds significant promise for hotspot detection of defective solar panels. It can be considered a complementary tool with potential utility, particularly for users lacking specialized GIS expertise. As multimodal artificial intelligence spreads to the solar energy field, ChatGPT is anticipated to be utilized in exploring temperature outliers generated in solar cells by individuals without specialized GIS expertise.</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

Using ChatGPT for detecting temperature anomalies in solar cells: potentials and constraints

  • 期刊论文

  • 发表日期：

    2025年7月28日

查看全文
  • Young-Seok Hwang,
  • Jung-Sup Um,
  • Biswajeet Pradhan,
  • Stephan Schlueter

摘要

This study evaluates the potential of ChatGPT for hotspot detection by comparing its performance with the Geographical Information System (GIS) software to detect temperature outliers in a photovoltaic (PV) cell. ChatGPT demonstrated similar performance to ArcGIS on z-score statistics and spatial distribution of hot and cold spots, but it exhibited a more sensitive detection capability. This study underscores the promise of ChatGPT in hotspot detection, with a high correlation of z-score statistics (R2: above 0.99) and high similarity of spatial distribution (70–97%) to ArcGIS. The preliminary results from our comparative study indicate that ChatGPT holds significant promise for hotspot detection of defective solar panels. It can be considered a complementary tool with potential utility, particularly for users lacking specialized GIS expertise. As multimodal artificial intelligence spreads to the solar energy field, ChatGPT is anticipated to be utilized in exploring temperature outliers generated in solar cells by individuals without specialized GIS expertise.