Application of model based predictive control for water-based floor heating in low energy residential buildings

A model based predictive control method is applied in order to determine the optimal supply fluid temperature in the case of concrete embedded water-based floor heating in low energy residential buildings. The aim of the control is to keep the indoor temperature within a defined comfort interval. The forthcoming supply fluid temperature is obtained through a numerical optimisation based on a prediction of the upcoming heat demand. The elementary response function, which is the basis for the method, is obtained from a numerical control volume model, and as an alternative, from a simplified 2-node lumped model. The accuracy of the results obtained from the simplified model is surprisingly good in comparison to the detailed numerical model. The control method is applied for a single room for which a perfect prognosis of the heat demand exists. The results show a fairly steady optimised supply fluid temperature.