The simulation of high-dimensional machine tool models, as well as their integration into hardware controllers in thermal real time requires model reduction techniques. The main focus is to develop model order reduction methods for (potentially nonlinearly coupled) network models of entire machine tools (in cooperation with A05 and A07) with respect to given accuracy demands. Furthermore, the application of cooling lubricant gives rise to models with uncertainties. That is, specially tailored model order reduction methods need be developed and adapted to these new model properties. An other important working package is the parametric model order reduction for the use in parameter identification and uncertainty quantification (with e.g., B08). Finally, the consideration of nonlinearities, like e.g. hysteresis effects in C02, requires specially tailored reduction methods for nonlinear models.
DFG Programme
CRC/Transregios
