This work shall lay the foundations for improved thermal comfort in climatecontrolled rooms. A methodology is developed which makes it possible to estimate the airflow process in real-time. Based on mathematical modeling, the air temperature and velocity are calculated by virtual sensors at given positions in the room. Therefore, a computationally efficient, data-based airflow model is developed. When applying the virtual airflow sensors, a particular focus is set on cooling with mechanical ventilation, which can produce spacedependent temperature and velocity profiles. These critical thermal conditions in the occupied zone cannot be captured sufficiently by standard room sensors which are typically located on the boundaries. The benefit of the new virtual airflow sensors is demonstrated by controlling the Predicted Mean Vote (PMV according to ISO 7730 standard) in a real conference room. The results are compared with those of a standard state-of-the-art temperature control. The evaluation illustrates that the PMV control via virtual sensors can establish comfortable conditions much more reliably.weiterlesen