Building Energy SystemsCopyright: © EBC
Our Vision is the Optimization regarding building energy systems (BES) of tomorrow by systematically linking experimental and simulative methods.
We, the team BES, conduct research on the holistic optimization of sustainable energy systems in terms of design and operation. For this purpose, we couple experimental and simulative methods to bring innovative systems into practice. In our team, we deal with both new and existing buildings in terms of single and multi-family houses as well as non-residential buildings. We focus on the integrated optimization of technologies related to heating, cooling and electricity used for generation, distribution, storage and transfer to the occupants and users.
In the area of simulation, we rely on both data-driven models through machine learning algorithms and physics-based models. The two at the Institute developed open-source model library AixLib and BESMod enable us to tailor dynamic, coupled building performance and HVAC simulations. Using digital twin concepts, we obtain calibrated dynamic models of building energy systems. Increasingly, we use Building Information Modeling (BIM) to gain simulation models and with creating the Tool bim2sim we generated the possibility to automatize the process.
In the field of experiments, we use programmable logic controllers (PLC) to allow analog signals from high-resolution measurement equipment and various bus systems (sensors and actuators) to communicate with cloud-based signals. We design our dynamic experiments using methods of statistical and optimal experimental design. This serves in particular to reduce experiment duration and costs.
Using the hardware-in-the-loop concept, we combine experimental investigations of typical system components (heatpumps, storage tanks, radiators, valves, etc.) with dynamically simulated buildings. This allows new products and control concepts to be investigated reproducibly under realistic conditions. For this purpose, we use state-of-the-art tools from the field of software and the Internet of Things (IoT). We perform evaluations with high-performance, statistical methods to derive scientifically well-founded statements from experiment and simulation for practical applications.
We use mathematical optimization methods to determine ideal system designs as well as cost-efficient and sustainable operation of building energy systems. For these optimizations, we use mixed-integer linear programs in combination with dynamic simulation models. Furthermore, we develop methods to optimize renovation roadmaps within multi-period systems.