Energy optimization of a residential building using model predictive control - A case study in temperate oceanic climate

This study presents a Model Predictive Control (MPC) framework for energy optimization in a residential building. A Dynamic control of the building service system is designed using a coupling between TRNSYS software for building simulation and Python for optimal control. Firstly, a predictive model is developed for building thermal load considering boundary conditions and time-varying uncertainties. An MPC-based supervisory controller in Python script interacts with local controllers in TRNSYS at the beginning of each time step. The supervisory controller calculates the control variable, corresponding to the minimum energy demand and desired thermal comfort. A geothermal brine-to-water heat pump is used with thermal energy storage and cooling water coils to serve the thermal load of the building. The optimization framework is employed in a single-family household in Saarbrücken, a temperate oceanic climate, for three months (January, June, and October). MPC calculates the optimal values of window shading fraction and reference set points of local PID controllers in TRNSYS. The simulation results show that MPC-based dynamic control of the building service system significantly improves the energy performance of the building without compromising thermal comfort.