A fuzzy control adapted by a neural network to maintan a dwelling within thermal comfort

This paper presents the results of a neuro fuzzy control setting the fan-coils air flow rate of three zones of a dwelling to improve the comfort. The zones temperature and humidity are the input variables from which the value of Fanger’s thermal comfort index PMV (predicted mean vote) is calculated and used as comfort variable. The TRNSYS simulation program has been used to simulate the building under different weather conditions. Under winter conditions in the heating mode hot water circulates through the fan coils and under summer conditions in the cooling mode, chilled water.