Approximately 40% of the global final energy consumption is used for air conditioning and heating of buildings. The building sector is thereby responsible for nearly 40 % of direct and indirect CO₂ emissions. The increase in energy efficiency and reduction of energy consumption of buildings is therefore an important challenge in respect of the climate protection goals. Inefficient or inadequate control of heating, ventilation and air conditioning (HVAC) systems can significantly increase the energy consumption, especially of modern and highly insulated buildings. 

In this project the application of model predictive control (MPC) in a large-sized office building in Hamburg is investigated. This control strategy makes use of forecast data, e.g. weather forecast or occupancy forecast and a dynamic model of the building to obtain the optimal trajectory of the control variable of a given time horizon by solving a dynamic optimization problem.

The goal of MPC is the reduction of energy consumption while at the same time the comfort of building occupants should be improved. Different modelling approaches like grey-box or black-box modelling (e.g. neural networks), are investigated. MATLAB and the modelling language Modelica^® are used for this purpose.

Contact: Svenne Freund