Enhancing thermal comfort while reducing cooling demand is a pressing challenge for Mediterranean school buildings, especially as climate change accelerates the frequency and intensity of heat-wave events. Hybrid ventilation (HV), switching between natural ventilation (NV) and air conditioning (AC), can meet this challenge, but conventional passive or rule-based controllers cannot adapt to the shrinking potential of comfortable NV hours. This study proposes a reinforcement learning (RL) framework that continuously learns when to deploy NV and when to shift to AC set-points, thereby preserving comfort and efficiency under both present and projected future climates. A single-zone EnergyPlus model of a Beirut classroom is coupled to an RL environment with a 20 min control interval. The RL agent observes indoor temperature, predicted mean vote (PMV), and weather conditions. Its actions include (i) window opening and (ii) AC activation at different temperature setpoints. The reward function penalizes PMV deviation and HVAC energy consumption, but awards bonuses when comfort is achieved via NV. The agent is trained on a typical meteorological year (TMY) and evaluated on 2050 and 2080 hot-scenario weather files to test climate resilience. Ongoing simulations will benchmark the learned policy against a calibrated rule-based baseline, focusing on three metrics: percentage of time within the comfort band (−0.5 ≤ PMV ≤ +0.5), cooling energy demand, and indoor overheating. The study expects the RL approach to reveal adaptive control patterns, such as dynamically adjusting AC set-points in response to outdoor heat severity, and to demonstrate superior performance across current and future climates. By delivering a data-driven, climate-adaptive HV controller and an EnergyPlus testbed, this work aims to supply building operators and researchers with a practical pathway toward the resilient, low-energy operation of educational spaces in warming Mediterranean regions.
Author Contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by H.K., J.Y. and N.G. The first draft of the manuscript was written by H.K. and all authors commented on previous versions of the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Institutional Review Board Statement
The study did not require ethical approval.
Informed Consent Statement
Not applicable.
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors on request.
Conflicts of Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).