Modeling local thermal responses of individuals : validation of advanced human thermo-physiology models

Human thermo-physiology models (HTPM) are useful tools to assess dynamic and non-uniform human thermal states. However, they are developed based on the physiological data of an average person. In this paper, we present a detailed evaluation of two sophisticated and well-known models, JOS3 and ThermoSEM, with the objective to evaluate their capabilities in predicting the local skin temperature of individual people as both models use individual parameters such as sex, height, weight, and fat percentage as input. For the purpose of validation, controlled experiments were conducted with six human subjects (3 males, 3 females) at different environmental conditions (22–28°C). The measured core temperature and the local skin temperature at 14 locations were used to evaluate the predicted values. Outputs from both HTPMs followed the dynamic trend of the experiments, with a root mean squared error (RMSE) of 0.9–0.3°C for core temperature and 1.3–0.9°C for mean skin temperature from both ThermoSEM and JOS3 correspondingly. However, the main errors came from the body extremities. The RMSE was different for each subject, and both models showed lower errors in the warmer environment. The average RMSE for the hands of all subjects was 2°C from ThermoSEM and 1.9°C from JOS3, while it was 0.8°C for the forehead in both models. The paper highlights the capabilities and limitations of the selected HTPMs and, furthermore, discusses the application of HTPMs in the field of personalized thermal comfort