A hybrid physics-based/data-driven model for personalized dynamic thermal comfort in ordinary office environment

Abstract Occupants’ thermal comfort plays a critical role in building operation and its energy management. However, personal differences in psychology and physiology may cause diversity and uncertainties of personal thermal comfort. In order to perceive the individual difference and make a personalized prediction for thermal comfort, based on the cyber-physical system framework, we propose a hybrid physics-based/data-driven model to dynamically predict the personalized thermal comfort through online learning and computation. This model consists of a physical part and a data-driven part. The physical part is developed based on the traditional heat balance equation. Since in the physical part there are some parameters (such as metabolic rate and skin temperature) which are costly to be measured in practice, a data-driven part is thus developed based on the regression model to estimate the uncertain parameters with the feedback of occupants. By integrating these two parts, the developed model is capable of taking both advantages of the physics-based and data-driven methods. The model performance is evaluated by a field experiment study and the results demonstrate that the model could achieve performance improvement compared with the PMV model and data-driven model, even with insufficient data.