Thermal comfort of naturally ventilated houses in countryside of subtropical region

Based on field study and Site investigations, this paper obtains the relationship between thermal comfort and air temperature, air velocity, relative humidity and means radiant temperature of naturally ventilated houses in countryside of subtropical region. According to Fanger, one will feel good when the thermal comfort vote is in the interval [-1, 1]. Thus, this thesis gains the reasonable interval of temperature: [24.7°C, 31.2°C], air velocity: [0.5m/s, 1.1m/s], relative humidity: [65%, 85%] and mean radiant temperature: [17.2°C, 32.8°C]. Using the seven scale comfort index, the artificial neural network is designed to predict the thermal comfort. We gain six indexes in the field research and they are temperature, air velocity, relative humidity, mean radiant temperature, metabolic rate and clothing thermal resistance. These six indexes can be used as the inputs of the network, meanwhile, the output of the networks is the thermal sensation vote. In addition, forty groups of data can be used to train the network and the other seventeen groups are used to predict. Residuals are small and it proves that the effectiveness of the network is excellent.