Office Heating Modeling Using an HVAC Power Step Response Experiment

Abstract

Indoor air quality is an important aspect in office buildings. Poor air quality leads to people feeling unwell and working inefficiently. However, heating, ventilation and air conditioning cause a significant expenditure in large building complexes. Automation systems that are widely installed in buildings in many places around the world allow control of the ventilation, which can be utilized in reduction of energy consumption. Still, utilization of the control systems needs knowledge of the system dynamics, which may be a troublesome task to accomplish with insignificant return. This paper presents a method for indoor temperature modeling using a gray box model and a step response experiment for data gathering to obtain a means of predicting indoor air temperature with adequate accuracy with low effort compared to more complex modeling endeavors. When testing the model, the maximum error between the actual and predicted air temperature was 0.270∘C and the absolute error value of the prediction was 0.112∘C.