Human Thermal Comfort Assessment Model under Lower-Pressure Environment Based on BP Network

Most of the studies on thermal comfort are carried out under normal pressure condition. Evaluation models of human thermal comfort are also limited to normal atmosphere. However, under other special conditions, e.g. lower-oxygen lower-pressure altitude environment, is worthwhile and necessary to do more work .In this paper, based on artificial neural network (ANN), the relation between the main influencing factors (such as temperature, relative humidity, air velocity, mean radiant temperature, air pressure, clothing insulation) and thermal sensation is made. Moreover, back propagation (BP) neural network evaluation model of human thermal comfort is put forward by Matlab under low pressure environment and learning algorithm of network is given. The result shows that prediction of network model is consistent with experimental results closely.