Competencies
Our core competence is the combination of numerical simulations with data science and scientific machine learning. In many of our research projects, we realize model-based controllers, system optimizations, and physical emulators through this synergy.
Physical modelling
Our research group specializes in the physical modeling of a variety of problems from industry and energy technology. We support our research projects with simplified 1-D modeling up to complex 3-D flow simulations (CFD), acoustic simulations, structural mechanics (FEM), and general thermodynamic simulations. A specialty of our research group is the numerical simulation of thermal arcs in high, medium, and low voltage ranges. The complex and coupled simulation technique is not limited to arcs but is also used, for example, for investigating the thermal runaway of battery cells in electric vehicles.
Data Science
Digitalization in society and industry requires competence in data science at many levels. We apply our expertise in machine learning for open-loop predictions, system monitoring, data analysis, and data-driven emulators for industrial applications as well as in public sector applications such as drinking water supply.
Scientific Machine Learning
Scientific machine learning is not just a purely data-driven method but also incorporates physical modeling. This adds additional information to the model, which is indispensable for certain applications or expands the range of possibilities. We use scientific machine learning for example in model-based control, model identification, as well as applications with soft sensors.