Buildings are more than just Architecture

The building sector accounts for about 40% of the total energy consumption worldwide. In order to guarantee maximisation of user comfort while minimising building energy consumption, it is necessary to both invest in passive energy-saving measures (insulation, external shading, green energy, etc.) and in intelligent automation schemes. An additional automation challenge is posed by advanced alternative building systems like heat pumps, free cooling, photovoltaics, etc.

Nowadays, many modern buildings are not only energy consumers but also energy producers. Therefore, solutions for energy-efficient buildings and grids are a major topic of research.

energy efficient modern building

© Luigi Caputo

Building Models: The Key to Control Performance

Model predictive control (MPC) requires building models for online optimisation of both comfort and energy consumption. These models need to cover nonlinear building dynamics with minimum effort.
Black-box (based on measured data only) and
grey-box models (utilising also expert knowledge) fulfil these requirements. An efficient way to approximate the overall nonlinear building behaviour is given by local linear models (LLM). These LLMs enable the use of a special type of control, the so-called fuzzy control. Moreover, building supply systems are typically represented by hybrid systems (both switched and continuous dynamics). Both building and supply face an abundance of constraints, which must be respected by the control design.

Pushing the Limits: Smart Building Control by MPC

Modern office buildings with large thermal mass have high load-shifting potential and can afford to shift their heating and cooling schedule for several hours without thermal comfort loss. Moreover, predictions of stochastic disturbances like weather forecasts (radiance and ambient temperature) and occupancy are explicitly incorporated in the MPC optimisation. The workgroup‘s research focus is placed on two types of model predictive controllers, such as on a fuzzy MPC for comfort maximisation of the building's indoor temperature and also on a mixed-integer MPC for supply optimisation. In order to individually regulate different building zones, a new hierarchical concept as well as a cooperative MPC scheme have been developed.

However, not only individual buildings need intelligent automation systems, but interaction in smart grids also requires a hierarchical MPC structure for global optimisation. A patent for hierarchical predictive load control in smart grids has been filed.


Fallmann, Markus, Agnes Poks, and Martin Kozek. "Control-oriented hybrid model of a small-scale refrigerated truck chamber., opens an external URL in a new windowApplied Thermal Engineering 220 (2023): 119719, ISSN 2451-9049

Poks, Agnes, Elisabeth Luchini, Markus Fallmann, Camillo Signor, Andreas Wurzinger, Dominik Radler, Stefan Jakubek, and Martin Kozek. "Distributed hierarchical control for multiple refrigeration units., opens an external URL in a new windowThermal Science and Engineering Progress 33 (2022): 101319, ISSN 2451-9049

Fuhrmann, Florian, Alexander Schirrer, and Martin Kozek. "Model-predictive energy management system for thermal batch production processes using online load prediction., opens an external URL in a new windowComputers & Chemical Engineering 163 (2022): 107830, ISSN 0098-1354

Fuhrmann, Martin, Nikolaus Euler-Rolle, Michaela Killian, Markus Reinwald, and Stefan Jakubek. "Longitudinal tunnel ventilation control. Part 2: Non-linear observation and disturbance rejection., opens an external URL in a new windowControl Engineering Practice 63 (2017): 44-56, ISSN 0967-0661

Euler-Rolle, Nikolaus, Martin Fuhrmann, Markus Reinwald, and Stefan Jakubek. "Longitudinal tunnel ventilation control. Part 1: Modelling and dynamic feedforward control., opens an external URL in a new windowControl Engineering Practice 63 (2017): 91-103, ISSN 0967-0661

Killian, M., B. Mayer, and M. Kozek. "Effective fuzzy black-box modeling for building heating dynamics, opens an external URL in a new window." Energy and Buildings 96 (2015): 175-186.

Mayer, Barbara, Michaela Killian, and Martin Kozek. "Management of hybrid energy supply systems in buildings using mixed-integer model predictive control, opens an external URL in a new window." Energy conversion and management 98 (2015): 470-483.

Brandstetter, Markus, Alexander Schirrer, Maja Miletić, Sawsan Henein, Martin Kozek, and Friederich Kupzog. "Hierarchical predictive load control in smart grids, opens an external URL in a new window." IEEE Transactions on Smart Grid 8, no. 1 (2015): 190-199.

Research Projects at our Institute


Ao.Univ.Prof. Dipl.-Ing. Dr.techn. Martin Kozek

Send email to Martin Kozek