Study of data-driven thermal sensation prediction model as a function of local body skin temperatures in a built environment

Current thermal/sensation models primarily rely on predefined formulas or empirically defined recommendations, but fail to consider each individual's physiological characteristics. Such models frequently ignore occupants' diverse physical conditions and, therefore, have critical limitations in estimating each individual's thermal sensation levels. Since the human body is governed by the thermoregulation principle to balance the heat flux between the ambient thermal condition and the body itself, skin temperature has a significant role in maintaining this physiological principle. Therefore, this study investigated the potential use of skin temperature and its technical parameters in establishing a thermal sensation. By using advanced modern sensing technologies, and existing thermal regulation model research, this study selected and validated seven body areas as significant local body segments for determining overall thermal sensation. A series of environmental chamber tests were conducted for 2 h. While the indoor temperature fluctuated between 20 degrees C and 30 degrees C, skin temperatures of the seven selected body areas were measured in conjunction with a thermal sensation and comfort survey. Results of this study revealed that combinations of skin temperatures for the arm, back, and wrist provided the significant information needed to accurately estimate the thermal sensations of each user. Most of all, both sides of the wrist generated accurate data more than 94% of the time. Therefore, considering the modern advanced wearable sensing technologies, results of this study confirmed that optimum combinations of skin temperature information from selected body areas, is reliable and generally applicable for estimating individual thermal sensations. (C) 2017 Elsevier Ltd. All rights reserved.