Thermal comfort prediction using normalised skin temperature of different age group as physiology parameter in the uniform air-conditioned space

Abstract

By using the predicted thermal comfortability of individual inside the building, thermal comfort prediction could be a good tool for energy efficiency as well as to create the efficient cooling system in buildings. This study will be presenting on the Thermal State Prediction Model (PTS) by using the skin temperature and the skin temperature gradient as a factor as well as evaluate the thermal comfortability state based on different age group. By using the surface area of the body and clothing insulation, normalisations process was introduced in this study to solve the issue of different in each individuals. Experiments were carried out on different human and skin temperature and thermal sensation level were recorded at every minute throughout the experimental procedure. The experimental location was simulated with about 18°C to 27°C to create the cooling, cold to room (natural) environment. It was observed that the thermal comfortability/state is able to be established by the combination of skin temperature and skin temperature gradient. Support Vector Machine(SVM) and Extreme Learning Machine (ELM) were used to classify the data with four model input situation. Normalised skin temperature was able to predict with a higher accuracy than the non-normalised skin temperature. It was also observed in this study age could also affects the thermal comfortability and skin sensation.