Optimizing air conditioner system efficiency through reinforcement learning

Date

2025-8

Author

Hekimoğlu, Mehmet Burak

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

359
views

0
downloads

Cite This

Heating, ventilation, and air conditioning (HVAC) systems play a crucial role in maintaining thermal comfort and indoor air quality, but they also account for a substantial portion of a building’s energy consumption. This thesis presents a reinforcement learning (RL) approach for optimizing the operation of an HVAC system in a single-zone environment. The proposed framework comprises multi-agent Deep Q-Network (DQN) controllers operating in both competitive and cooperative settings, aiming to minimize energy consumption while ensuring thermal comfort and maintaining indoor air quality. Results show that the best multi-agent competitive configuration achieves 24.8% lower energy consumption than the rule-based controller and 18.9% lower than the single-agent DQN. These findings show that multi-agent RL can outperform both single-agent RL and traditional control strategies.

Subject Keywords

Reinforcement learning, Multi-agent systems, HVAC, Energy optimization, Indoor air quality

URI

https://hdl.handle.net/11511/115653

Collections

Graduate School of Natural and Applied Sciences, Thesis