[Back]


Talks and Poster Presentations (with Proceedings-Entry):

M. Killian, M. Kozek:
"Short-term occupancy prediction and occupancy based constraints for MPC of smart homes";
Talk: 2019 IFAC Workshop on Control of Smart Grid and Renewable Energy Systems, Jeju, Korea; 2019-06-10 - 2019-06-12; in: "2019 IFAC Workshop on Control of Smart Grid and Renewable Energy Systems", (2019).



English abstract:
The implementation of a smart home product should be simple and possibly done by craftsmen. However, the important occupancy prediction currently requires considerable expert knowledge. In this paper a model predictive control (MPC) scheme for smart homes using an unsupervised occupancy prediction method is presented. The developed MPC is able to exploit the full potential of energy storage in a smart home with a minimum risk of comfort violations. Occupancy data are first transformed by proper orthogonal decomposition and then clustered in a low dimensional parameter space. The proposed MPC scheme uses the set of extracted occupancy profiles together with a suitable hypothesis test to predict the occupancy of the smart home. This prediction is the basis for smart thermal comfort constraints effective in the MPC optimization step. Simulation results of the proposed methods coupled to an MPC for the smart home and a grid profile show the versatile and effective performance.

Keywords:
Predictive control, model-based control, smart home automation, occupancy prediction, smart grid constraints

Created from the Publication Database of the Vienna University of Technology.