Volume 62

Data-Driven Modeling of Indoor Thermal Comfort for Smart Cooling Systems Xiaohan Li, Yanjuan Liu, Wenjing Yang

Abstract

This paper presents a systematic study of data-driven modeling of indoor thermal comfort for smart cooling systems, establishing an analytical framework encompassing three main aspects: data collection, modeling analysis, and control execution. The findings indicate that existing studies in data collection exhibit the characteristics of integrating multi-source data, combining objective and subjective data, and employing multi-technology collaborative collection. During the modeling and analysis phase, machine learning algorithms play a dominant role in indoor thermal comfort modeling, achieving higher predictive accuracy than the traditional Predicted Mean Vote model. These models can be divided into two categories: group and personal models, with the latter showing greater potential for personalized predictions. In terms of control execution, the primary strategies can be categorized into two types: model-based optimization control and adaptive control that integrates user feedback. Their effectiveness in improving comfort, reducing energy consumption, and dynamically responding to personalized needs and environmental changes has been preliminarily validated. Furthermore, this paper outlines critical challenges at different stages and systemic levels in this field and proposes directions for future research.

Keywords thermal comfort, smart cooling system, energy consumption, data collection, machine learning, control execution