Predicting Human Thermal Sensation Based On Forehead Temperature In An Office Environment

With the rapid development of society,people are working longer hours,putting forward higher requirements for indoor comfort.According to the thermal adaptation perspective,people in actual buildings are not only passive recipients of the environment,but also active adapters of self-regulation.The perception of human thermal comfort is not static and at the same time each individual has different thermal preferences.Therefore,there is a need to propose a research method that enables unsupervised,non-contact personalised thermal comfort models.Personalised thermal comfort modelling relies on the physiological parameters of the subject and current skin temperature based methods require the test instrument to be attached directly to or on the body.This approach is clearly not widely available and can cause inconvenience to building users.In order to be practical and acceptable to building occupants,the use of physiological parameters collected by non-invasive sensing techniques should be considered for the development of thermal comfort prediction models.To address these issues,this paper develops a personalised predictive thermal sensory study based on forehead temperature through the design of field experiments,which include the following.First,the changes in skin temperature and thermal comfort of various parts of the human body surface in an office environment are analysed.By correlating the skin temperature of each part and the average skin temperature with thermal sensation,it is concluded that the skin temperature of the forehead has a significant correlation with the human thermal sensation.Second,a multiple linear regression model for predicting thermal sensation based on human skin temperature was developed using multiple linear regression methods.The effect of each independent variable on the dependent variable TSV was also obtained by standardising the regression coefficients.Third,a field experiment was then conducted in Dalian,Liaoning Province,to investigate the relationship between forehead temperature and human thermal comfort,using both noncontact measurements(infrared thermography)and contact measurements(eTact).The effect of solar radiation on human thermal sensation was also considered and a control group was designed to investigate whether solar radiation had a significant effect on forehead temperature and human thermal comfort.Finally,a logistic regression method based on disordered multi-classification was used to develop a personalised thermal sensation prediction model based on the forehead temperature measured by the infrared thermographic camera.In addition,in a real-life scenario where there are multiple people with different thermal comfort preferences in an indoor environment,a multi-person fused thermal sensation prediction probability model is developed.