文献来源：
Liang Zhang
a
,
b
,
*
, Vitaly Ford
c
, Zhelun Chen
d
, Jianli Chen
e
,
f
a
University of Arizona, Tucson, AZ, United States
b
National Renewable Energy Laboratory, Golden, CO, United States
c
Arcadia University, Glenside, PA, United States
d
Drexel University, Philadelphia, PA, United States
e
Tongji University, Shanghai, China
f
University of Utah, Salt Lake City, UT, United States
文献链接：/2025.3.13.pdf
文章摘要：Building energy modeling (BEM) is a complex process that demands signiffcant time and expertise, limiting its broader application in building design and operations. While Large Language Models (LLMs) agentic workffow have facilitated complex engineering processes, their application in BEM has not been speciffcally explored. This paper investigates the feasibility of automating BEM using LLM agentic workffow. We developed a generic LLMplanning-based workffow that takes a building description as input and generates an error-free EnergyPlus building energy model. Our robust workffow includes four core agents: 1) Building Description Pre-Processing, 2) IDF Object Information Extraction, 3) Single IDF Object Generator Suite, and 4) IDF Debugging Agent. These agents divide the complex tasks into manageable sub-steps, enabling LLMs to generate accurate and reliable results at each stage. The case study demonstrates the successful translation of a building description into an error-free EnergyPlus model for the iUnit modular building at the National Renewable Energy Laboratory. The effectiveness of our workffow surpasses: 1) naive prompt engineering, 2) other LLM-based workffows, and 3) manual modeling, in terms of accuracy, reliability, and time efffciency. The paper concludes with a discussion on the interplay between foundational models and LLM agent planning design, advocating for the use of ffne-tuned, specialized models to advance this ffeld.
声明：本网站内容版权归许鹏教授课题组所有。未经本课题组授权不得转载、摘编或利用其它方式使用上述作品。
视频链接：全雅妮_利用大语言模型代理工作流自动开发和调试建筑节能模型_哔哩哔哩_bilibili