The WG Numerical Optimization focuses on different kind of optimization problems, namely
- Nonlinear Optimization
- Multiobjective Optimization
- Nonsmooth Optimization
- Discrete Optimization
- Stochastic Optimization
- Optimal Control of Complex Dynamical Systems
The goal is to develop, analyze and validate new algorithmic approaches, in particular in cases, where the optimization problems are governed by partial differential equations.
Since the underlying numerical solution techniques lead to large-scale problems, which are heard to tackle numerically, reduced-order methods (especially Proper Orthogonal Decomposition and Reduced-Basis) are applied to replace the large-scale problems by smaller ones. Then an a-posteriori error analysis is developed and its computationally efficient implementation is realized.
Another topic is the optimization of decentralized energy supply systems where we are currently working with local decision makers, but also with other departments of the University of Konstanz in order to implement locally the climate protection in an optimal way.
Our research is also mainly driven by our current projects.