Thermal Comfort in Air Conditioning Based on Bayesian Inference Using the EAP Estimation

On the basis of the dynamical Bayesian inference using the expected a posterior (EAP) estimation, we investigate a role of the thermal comfort using the temperature-humidity index(THI) in the model system of air conditioning via air cooling or dehumidification and the sensor network system composed of the set of data loggers for temperature and relative humidity. For this purpose, we estimate time evolution of the environmental variables, i.e. the temperature, the relative humidity and the THI at each sampling point, by making use of the dynamical Bayesian inference using the EAP estimation with the use of the master equation and the mean-field theory in statistical physics. We find that the air cooling or dehumidification in small-scale room succeeds in realizing comfortable environment by utilizing the term of the thermal comfort using the THI enhancing an optimum of the THI at each sampling point, although the passes from the observed state to the optimal state are respective of the air cooling or dehumidification.