A thermal model for processing data from undergarment sensors in automatic control of actively heated clothing

Abstract

Despite its recent growth in popularity, actively heated clothing still lacks the ability to cope with demanding user scenarios. As many of these deficiencies stem from an absence of automatic control, the authors propose a novel approach using a set of sensors embedded in the clothing to provide data about thermal comfort. Available sensors suffer from a lack of accuracy, as for prac-tical reasons, they cannot be attached to the skin, whose temperature is usually used as a comfort indicator. To determine the magnitude of the problem, the authors conducted experiments, and a thermal model was proposed based on experimental findings; the output from the model was compared with the experimental reference data for three different upper body undergarments. The over-all accuracy was found to be good: in most cases, the difference between the computed and reference skin temperatures did not exceed 0.5 °C. Furthermore, the model does not rely on unrealistic as-sumptions regarding the availability of parameters or measurement data. Our findings demonstrate that it is possible to create a thermal model that, when used for input data processing, allows undergarment temperature to be converted to skin temperature, allowing for automatic control of heating insets.