New Method for Accurate Prediction of CO_2 in the Smart Home

摘要

This article describes new method for accurate prediction of CO_2 in the Smart Home calculated from the temperature and relative humidity in application of the decision tree regression method. The measured data are loaded from the individual BACnet technology sensors by means of the Desigo Insight visualization tool. The individual BACnet technology components are used to control the heating, cooling and ventilation in Smart Home. The measured temperature (T) and humidity (rH) values are then used as input parameters for prediction of CO_2 content in the air of selected rooms in the Smart Home by application of decision tree regression. As described in the article, the method can determine the CO_2 content with the accuracy of 46.25 ppm. The obtained information can be used for monitoring the residents' life activities, optimizing the technical service system for reduction of the building's operating costs or automation of its responses to changes of the environment or the residents' activities.