Occupant-Oriented Economic Model Predictive Control for Demand Response in Buildings

The present paper develops an Economic Model Predictive Control (EMPC) framework to provide Demand-Response (DR) for supporting the power grid stability while also maintaining Occupants' Thermal Satisfaction (OTS) in buildings. Our controller combines economic and occupant-oriented aspects by simultaneously optimizing two conflicting control goals, namely grid stability and OTS in buildings. We represent grid stability with Grid Costs (GC) based on a real-world dynamic electricity price and OTS with a reference indoor temperature, respectively. In the literature, there exists no study about occupant-oriented DR where the Model Predictive Control (MPC) is based on Resistor-Capacitor (RC) models identified from real measurements that also includes an attendance schedule for DR. For this, the EMPC uses a grey-box thermal building model that is designed, identified, and validated with real-world measurement data. For evaluation, we compare the EMPC with a well-tuned conventional Proportional-Integral (PI) controller. The results show that the EMPC significantly outperforms the PI controller in terms of GC, while it respects OTS.