Retrieval Augmented LLM Framework for Robust Automated Generation of HVAC Diagnostic Models
Chujie Lu (TU Delft - Architecture and the Built Environment, Katholieke Universiteit Leuven)
Laure Itard (TU Delft - Architecture and the Built Environment)
Dirk Saelens (Katholieke Universiteit Leuven, EnergyVille)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Diagnostic Bayesian networks (DBNs) offer effective and interpretable solutions for HVAC fault detection and diagnosis (FDD), yet their practical construction and deployment require substantial manual effort. The work proposes a retrieval-augmented LLM framework to achieve robust and automated DBN generation. Preliminary results demonstrate that the proposed framework effectively mitigates hallucinations and significantly improves the structural quality of the generated DBNs.
help_outline