Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Physics-Informed Large Language Models for HVAC Anomaly Detection with Autonomous Rule Generation

Subin Lin, Chuanbo Hua

TL;DR

HVAC systems drive substantial building energy use, making reliable anomaly detection essential. The authors introduce PILLM, a physics-informed Large Language Model that evolves anomaly-detection rules through an iterative, physics-guided loop using reflection and crossover operators. By grounding rule generation in thermodynamic and control-theoretic principles, PILLM produces interpretable, executable Python rules and achieves state-of-the-art precision and $F_1$ on the public LBNL AFDD dataset, with ablations confirming the value of the physics-informed components. This work demonstrates that combining LLM reasoning with explicit physical constraints can yield trustworthy, deployable AFDD for smart buildings, bridging explainability, adaptability, and robustness.

Abstract

Heating, Ventilation, and Air-Conditioning (HVAC) systems account for a substantial share of global building energy use, making reliable anomaly detection essential for improving efficiency and reducing emissions. Classical rule-based approaches offer explainability but lack adaptability, while deep learning methods provide predictive power at the cost of transparency, efficiency, and physical plausibility. Recent attempts to use Large Language Models (LLMs) for anomaly detection improve interpretability but largely ignore the physical principles that govern HVAC operations. We present PILLM, a Physics-Informed LLM framework that operates within an evolutionary loop to automatically generate, evaluate, and refine anomaly detection rules. Our approach introduces physics-informed reflection and crossover operators that embed thermodynamic and control-theoretic constraints, enabling rules that are both adaptive and physically grounded. Experiments on the public Building Fault Detection dataset show that PILLM achieves state-of-the-art performance while producing diagnostic rules that are interpretable and actionable, advancing trustworthy and deployable AI for smart building systems.

Physics-Informed Large Language Models for HVAC Anomaly Detection with Autonomous Rule Generation

TL;DR

HVAC systems drive substantial building energy use, making reliable anomaly detection essential. The authors introduce PILLM, a physics-informed Large Language Model that evolves anomaly-detection rules through an iterative, physics-guided loop using reflection and crossover operators. By grounding rule generation in thermodynamic and control-theoretic principles, PILLM produces interpretable, executable Python rules and achieves state-of-the-art precision and on the public LBNL AFDD dataset, with ablations confirming the value of the physics-informed components. This work demonstrates that combining LLM reasoning with explicit physical constraints can yield trustworthy, deployable AFDD for smart buildings, bridging explainability, adaptability, and robustness.

Abstract

Heating, Ventilation, and Air-Conditioning (HVAC) systems account for a substantial share of global building energy use, making reliable anomaly detection essential for improving efficiency and reducing emissions. Classical rule-based approaches offer explainability but lack adaptability, while deep learning methods provide predictive power at the cost of transparency, efficiency, and physical plausibility. Recent attempts to use Large Language Models (LLMs) for anomaly detection improve interpretability but largely ignore the physical principles that govern HVAC operations. We present PILLM, a Physics-Informed LLM framework that operates within an evolutionary loop to automatically generate, evaluate, and refine anomaly detection rules. Our approach introduces physics-informed reflection and crossover operators that embed thermodynamic and control-theoretic constraints, enabling rules that are both adaptive and physically grounded. Experiments on the public Building Fault Detection dataset show that PILLM achieves state-of-the-art performance while producing diagnostic rules that are interpretable and actionable, advancing trustworthy and deployable AI for smart building systems.
Paper Structure (24 sections, 1 equation, 1 figure, 1 table)

This paper contains 24 sections, 1 equation, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. PILLM
  5. Initial Population.
  6. Physics-Informed Reflection.
  7. Physics-Informed Crossover.
  8. Elitist Rule Mutation.
  9. Experiment
  10. Main Results.
  11. Ablation Study.
  12. Explainability.
  13. Conclusion
  14. Task
  15. Metrics.
  16. ...and 9 more sections

Figures (1)

  • Figure 1: Overview of PILLM. The framework follows an evolutionary generate-and-reflect pipeline for anomaly detection rules. In each iteration, the current rule population undergoes Physical-Informed Reflection, where physical context is incorporated into candidate rules. These reflections are then used in Physical-Informed Crossover to produce the next generation of rules. Finally, elite rules are refined through mutation, resulting in evolved rules that are adaptive.