Surrogate models, inverse solvability and multi criteria optimization

The working group deals with the development and application of surrogate models for more efficient optimisation in production engineering. The aim is to describe complex physical processes using simplified surrogate models in such a way that computing times are significantly reduced. Various types of models are being investigated – from purely data-based approaches to hybrid data-based models with physically-informed components to accelerated purely physics-based models. Another research focus is on inverse solvability, i.e. the backward calculation of optimal process parameters for individual process steps and complete process chains. In addition, advanced methods of multi-criteria optimisation are analysed and further developed in order to investigate the optimisability of complete process chains.