Moving horizon estimation of lumped load and occupancy in smart buildings

Abstract

Model-based control strategies are widely used to improve the energy efficiency of building systems but they require accurate dynamic building models. Information about disturbances such as lumped load and occupancy is necessary in order to cover the mismatch between the simulations and reality, which leads to determining correct energy predictions of the building. Rigorous models that are applied in simulators such as EnergyPlus and Trnsys are difficult to be calibrated due to the high number of parameters and computational effort. Reduced models along with online estimators may be applied to overcome this issue. In this work, moving horizon estimation (MHE) is implemented for estimating the lumped load and number of occupants of an office whose required heat load is supplied by a hydronic heating system. MHE performance shows that its online estimations can be used along with model predictive control to improve output predictions.