Streamlined Model Predictive Control for a Triple-Optimization of Thermal Comfort, Heating Costs, and Photovoltaic Self-Consumption

This paper introduces a straightforward Model Predictive Control (MPC) method for residential buildings equipped with thermally activated building structures (TABS), photovoltaics (PV), and heat pumps (HP). The main goal is a triple-optimization of thermal comfort, heating costs, and PV self-consumption. Building on existing low-tech MPC developments, this study extends them to enhance PV self-consumption optimization. The proposed streamlined MPC algorithm relies solely on weather forecast data and real-time electricity market prices to determine the required heating power input. The developed and extended MPC algorithm is simulated and validated using Matlab/Simulink. Various cost function variants of the MPC algorithm are analyzed and calculated with both fixed and flexible price signals. The results demonstrate that different cost function variations significantly impact energy consumption, costs, and comfort levels. With flexible day-ahead prices, the cost function incorporating comfort, electricity costs, and PV signals performs comparably to the variant focusing solely on comfort and PV optimization. However, for fixed electricity prices, the results indicate that the simplest cost function of the MPC achieves the best outcomes. These findings suggest that further simplifications of the MPC approach can effectively reduce heating costs while maintaining comfort, making it more practical for implementation in residential buildings.