BROWSE
Detail
Conference Paper
Reinforcement Learning-Based HVAC System Operation Under Limited Data Acquisition
Cited 0 time in 
Share 
Share
- Authors
- Ye-Eun Jang, Wan-Ki Park
- Issue Date
- 2024-11
- Citation
- International Conference on Renewable Energy Research and Applications (ICRERA) 2024, pp.1-5
- Language
- English
- Type
- Conference Paper
- Abstract
- The optimal operation of heating, ventilation, and air-conditioning (HVAC) systems has been extensively studied to improve energy efficiency in building operation. With the increasing integration of renewable energy sources into power systems, this research has gained additional importance, particularly in addressing the uncertainties in power generation and utilizing HVAC systems as demand response resources. Traditional optimization approaches rely on precise parameters from physics-based models or full datasets that include all environmental factors influencing indoor temperature variations. However, it remains a significant challenge to acquire such detailed datasets and accurate model parameters. In this paper, a reinforcement learning-based strategy is proposed to operate the HVAC system under limited data acquisition. The proposed strategy does not require exact modeling parameters or a full dataset. Case studies demonstrate the effectiveness of the proposed strategy by comparing it to conventional strategies in terms of operating cost, thermal comfort, and practical data acquisition.
- KSP Keywords
- Accurate model, Air conditioning, Case studies, Data Acquisition(DAQ), Demand response resources, Energy efficiency, Environmental Factors, HVAC system, Learning-based, Limited data, Model parameter