A data-driven model on human thermophysiological and psychological responses under dynamic solar radiation

Excessive solar radiation can also cause thermophysiological and psychological discomfort in the human body. In real-world environments, solar radiation is highly variable. Moreover, the amount of solar radiation exposure that people receive due to their outdoor activities can have large fluctuation. The physiological and psychological responses of the human body under this dynamic solar radiation exposure are quite different from those under steady-state solar radiation exposure. Therefore, this paper studies the physiological and psychological responses of the human body under the condition of dynamic solar radiation through the method of field experimental research and explores the applicability of the existing indicators of PET and UTCI in the dynamic radiation environment. A novel recurrent neural network model was used to predict skin temperatures and thermal sensations, in which genetic algorithm was applied to tune hyper-parameters. It is noteworthy that both thermal sensations and skin temperatures at exposed body segments show a stronger correlation to solar radiation, while there exists a time lag between solar radiation and skin temperature when people are exposed to solar radiation. Compared with thermophysiological models (DTS), the GA-LSTM model has a better prediction accuracy in thermal sensations.