Conference article

The management of modern large buildings’ energy supply systems (ESS) demands the maximal usage of renewable energy sources at minimum energy costs while meeting the energy demand of the consumption zone. Building ESS for heating and cooling usually consist of various supply circuits with several energy sources and different physical characteristics, possibly incorporating switching aggregates (heat pump, chiller) with latency times and strati?ed storage which change their operating state in a discontinuous fashion. Hence, these circuits can be seen as hybrid systems whose modelling as well as optimisation are demanding. Model predictive controllers (MPC) are an effective means for the optimisation of such problem formulations with divergent goals. The proposed modular predictive control concept (MPCC) is designed for a ?exible usage in ESS building automation adding one appropriate MPC for each supply circuit including mixed-integer MPCs to the individual building’s control structure. The resulting MPCC is capable of maximizing the usage of renewable energy sources at minimum cost as well as ef?ciently managing switching aggregates with active storage. Suitable modelling of the linear and hybrid systems is demonstrated and validated on the example of a large of?ce building in Austria. Furthermore, a simulation study shows the performance of the resulting MPC concept compared to a rule-based controller.

building energy supply system, model predictive control