This paper presents a method for utilizing sensor networks to predict human’s thermal comfort and sensation. A neural network is dynamically organized on the basis of correlations between the thermal sensation of office occupants and a number of values measured by the sensor network, and the structure of the neural network is cyclically updated. By way of an example, the air-conditioning system in an office is used. We place a number of temperature sensors and participants in this indoor environment, and conduct an experiment where the sensor readings and the thermal sensation of the participants are monitored concurrently. From the experimental results it can be seen that the various sensors are selected correctly and can be used to predict the desired system behavior.
For more details: omicsonline.org/sensor-networks-data-communications.php