Our mission is to develop fast numerical methods for feedback control of large-scale, real-world processes in real-time. Our efforts are focused on the challenges of processes that
- are modelled with nonlinear parabolic PDEs,
- involve difficult boundary conditions, for example periodicity in time,
- involve switches in the dynamics,
- need to be optimized with respect to economic objective functions,
- are subject to uncertainties.
In a genuinely interdisciplinary approach, the development, analysis, and implementation of new algorithms addresses important questions from applications, especially chemical engineering.
Our research belongs to the following mathematical areas:
- Preconditioning of saddle-point systems
- Globalization of Newton-type methods
- Linear, quadratic, and nonlinear programming
- Numerical methods for differential equations (ODE, DAE, PDE)
- Model reduction
Last Update: 04.09.2017 - 16:37