Artificial intelligence-based optimization of HVAC systems: Advances in thermal comfort, energy efficiency, energy consumption, challenges, and research gaps

초록

As buildings consume roughly a third of all the energy used in the world, improving their energy efficiency and environment-friendliness is essential. Artificial intelligence (AI) methods, especially machine learning and deep learning, have been gradually adopted over the last decade to handle the complexity of built-environment systems. This review summarizes the recent research on AI-driven integrated and holistic frameworks for building management. Functions such as occupancy detection and behavior modeling, energy consumption forecasting and optimization, indoor air quality and thermal comfort assessment, and fault detection and diagnosis must be examined to streamline the management of buildings and address interrelated heating, ventilation, and air-conditioning issues. While studies have evaluated the performance, modeling approaches, and methodological structures of AI-based techniques, identifying viable AI models remains challenging due to the rapid proliferation of algorithms across spatial scales, including system components, urban-level studies, temporal resolutions, and assessment metrics. Because of this methodological diversity, drawing direct comparisons and selecting optimal techniques for specific applications is difficult. Although AI techniques can improve the energy efficiency of buildings, most studies have focused on controlled laboratory settings or simulations, with few post-occupancy evaluations and real-world deployments. Standardized evaluation frameworks and practical guidelines are required to implement AI solutions in operational buildings. Based on the current study, from 2000 to 2026, the application of AI techniques and thermal comfort models has enabled energy saving on average 21.91 and 44.76% and comfort improvement on average between 21.77 and 85.95%. Finally, this paper discusses the challenges faced in the use of AI for energy productivity and comfort improvement and opens main future directions in relation with AI-based HVAC control systems for human comfort and energy-efficiency management. © 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

키워드