Enhanced building thermal model by using CO2 based occupancy data

Abstract

Prevailing low energy buildings attracts lots of attention in the world. Many studies have contributed in introducing higher thermal efficiency towards rooms with low energy heating, ventilation, and air-conditioning (HVAC) systems. However, current HVAC systems do not consider CO2 concentration change and thermal contribution towards human bodies in a room. This paper presents a novel method to predict thermal dynamics, including person count. Occupancy data are dynamically estimated by CO2 concentration and thermal contribution from the human bodies. The model is formulated as a resistor-capacitor circuit (RC circuit) in the Modelica modeling language. All parameters in a simulation are identified using actual building data during the winter season in Japan. Results are validated using measured information of actual building environment, and the test results concluded an improvement of absolute percentage error by 0.16 % over the conventional model. From the test results, it was concluded that the moving average filter of 20 minutes was an appropriate mean time to represent the time delay value.