[Back]

M. Killian, A. Leitner, R. Goldgruber, M. Kozek:
"Short-term forecasts for optimal model predictive control in smart homes";
Talk: e-nova 2017: Zukunft der Gebäude: digital - dezentral - ökologisch, Pinkafeld, Burgenland; 2017-11-23 - 2017-11-24; in: "Zukunft der Gebäude: digital - dezentral - ökologisch", Science Research Pannonia / Leykam Verlag, 21 (2017), ISBN: 978-3-7011-0399-7; 61 - 68.

Buildings are dynamical systems with several control challenges: large storage capacities, switching aggregates, technical and thermal constraints, and internal and external disturbances (occupancy, ambient temperature, solar radiation). Model predictive control (MPC) can achieve considerable reductions in energy consumption and increase the thermal comfort, however, a prerequisite are reliable forecasts for both weather and occupancy. Thus, with an intelligent short-term forecast model of building occupancy (and weather, especially solar radiation) the extensive skills of MPC become more efficient in case of smart home automation. In this research a prediction model for occupancy in smart homes is designed. The prediction model is embedded in the MPC framework and tested in different simulations cases. To get the highest possible performance with the least effort, beside the MPC concept, the forecast model for generating the prediction is a self-adaptive model. The combination of MPC with self-adaptive forecast models for occupancy (and solar radiation) is innovative and the potential is demonstrated in simulations.