Conference Year



Agent-based approach, Plug and play, MPC, High performance building, Radiant floor system


This study presents an optimal control formulation for the operation of a radiant floor system in an open plan office space with an air-cooled chiller as a source. A simulation case study with different control schemes is used to evaluate the potential of the model predictive control for the radiant floor as well as the optimal control coordination of a radiant and air comfort delivery system. The comparison with a reference case of proportional control shows a saving potential for the radiant floor of around 10 to 15.8 %, which results from maintaining the temperature at upper bound and precooling or load shifting. Optimal control coordination of radiant floor and air system yields additional saving of around 2 %. The proposed intuitive formulation of linear programming can be implemented to other control problems with a linear building model and known COP with respect to weather prediction. The formulation is applicable to other complex systems with two or more control systems such as open-plan spaces with several control units or multiple zones (or buildings) with centralized plant.