Prompt engineering to inform large language model in automated building energy modeling

Gang Jiang, Zhihao Ma, Liang Zhang and Jianli Chen

Energy, 2025, vol. 316, issue C

Abstract: Application of large language models (LLMs) to facilitate auto-building energy modeling (ABEM) is complex and resource-intensive. This paper presents practical guidelines for ABEM using prompt engineering with LLMs. Unlike training LLMs for ABEM with fine-tuning (involving model weight adjustment), prompt engineering is to provide specific prompts along with modeling descriptions and demonstrations, informing LLMs to automatically generate building energy models (BEMs) through natural language expression, and enabling users to perform ABEM without specialized building knowledge or software proficiency. In this study, the capabilities of prompt engineering for ABEM using LLMs are investigated. To achieve this, six types of prompts are designed, encompassing two exploratory tasks and one real-world task with a total of 648 case studies. The results from the case studies suggest that prompt engineering is feasible to automatically obtain the desired BEMs using one-shot learning (one demonstration), few-shot learning (few-demonstration), and chain-of-thought strategies (i.e., task explanation and division). Finally, the modeling comparison between the tested lightweighted LLMs with GPT-4o suggests that compact LLMs with an appropriate context window are suitable to be deployed for various building applications. The source codes for prompt engineering are available on GitHub (https://github.com/Gangjiang1/Prompting-for-Auto-building-Modeling.git).

Keywords: Prompt engineering; Large language model; Building energy models; Automated building energy modeling; In-context learning; Generative artificial intelligence (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544225001902
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:316:y:2025:i:c:s0360544225001902

DOI: 10.1016/j.energy.2025.134548

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().