Thermal sensation models: Validation and sensitivity towards thermo-physiological parameters

Abstract Thermal sensation models are commonly used to assess thermal perception in various indoor environments. However, using different models to evaluate the same environment can result in high discrepancies between the models' predictions, as they have been developed based on experimental data from various populations, use diverse input parameters, have different ranges of applicability and use different output scales. In this study, a validation of seven existing thermal sensation models has been performed based on literature data with regards to uniform steady-state and transient indoor environments. The environmental and personal parameters were selected following the experimental protocol and the needed thermo-physiological input parameters were obtained from a thermo-physiological model. Six models showed a good performance for the analyzed range of conditions, with a mean root-mean-square deviation equal to or lower than 1 thermal sensation unit, even beyond their original range of application. A sensitivity study towards thermo-physiological parameters was also performed, showing that the models are not equally influenced by some inaccuracies in these input parameters. Since thermal sensation models are often associated with different thermal sensation scales, the possibility of applying a scaling to the predictions has been considered. However, the scaling did not consistently improve the predictions accuracy.