Conference2026IEEE

Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control

Ulaş Filiz, Sajjad Baghaee, Mubeen Ahmed Fayyaz, Ilkay Ulusoy

2026 34th Signal Processing and Communications Applications Conference (SIU)

Official publicationDOI

Abstract

Deploying reinforcement learning (RL) agents for heating, ventilation, and air-conditioning (HVAC) control across climatically diverse buildings is challenging, as policies trained in one climate often fail to generalize to others. We investigate transfer and meta-reinforcement learning for building climate control using Sinergym with a custom EnergyPlus model of a real room, comparing Double Deep Q-Network (DDQN) and Model-Agnostic Meta Learning DDQN (MAML-DDQN) across three transfer scenarios on a target very hot climate. Fine-tuned DDQN reduces temperature violations by a factor of 35 over the from-scratch baseline and improves reward by 86% over zero-shot transfer, while MAML-DDQN converges in less than half the fine-tuning episodes and exhibits substantially lower variance across seeds. Results highlight a key trade-off: MAML-DDQN is preferable at adaptation-focused deployments, while fine-tuned DDQN is superior when maximum performance is the objective.

Selected figures

Figure 1 from “Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control”: Meta-reinforcement learning transfer procedure between source and target environments
Fig. 1 · Meta-reinforcement learning transfer procedure between source and target environments.
Reproduced from: U. Filiz et al., “Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control,” 2026 34th Signal Processing and Communications Applications Conference (SIU), 2026. View the original publication.
Figure 3 from “Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control”: Fine-tuning results on very hot. Left: Episodic cost over 100 episodes. Right: Total execution time in evaluation
Fig. 3 · Fine-tuning results on very hot. Left: Episodic cost over 100 episodes. Right: Total execution time in evaluation.
Reproduced from: U. Filiz et al., “Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control,” 2026 34th Signal Processing and Communications Applications Conference (SIU), 2026. View the original publication.

Cite this work

Plain text

U. Filiz, S. Baghaee, M. A. Fayyaz and I. Ulusoy, "Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control," in 2026 34th Signal Processing and Communications Applications Conference (SIU), pp. 1-4, 2026. doi: 10.1109/SIU71813.2026.11636860.

BibTeX 
@inproceedings{filiz2026analysis,
  author    = {Filiz, Ulaş and Baghaee, Sajjad and Fayyaz, Mubeen Ahmed and Ulusoy, Ilkay},
  title     = {Analysis of Meta-Reinforcement Learning on Transfer Learning for HVAC Control},
  booktitle = {2026 34th Signal Processing and Communications Applications Conference (SIU)},
  pages     = {1-4},
  publisher = {IEEE},
  year      = {2026},
  doi       = {10.1109/SIU71813.2026.11636860},
  url       = {https://ieeexplore.ieee.org/abstract/document/11636860}
}

All publications