01.01.2023 − 31.12.2024
The transition of the energy system by including renewable sources and the coupling of different energy sectors make the dynamical models large and highly volatile. To prevent system failures and to reduce costs, the operation of systems needs to change to real-time. To achieve this we will use the energy-based framework of constrained port-Hamiltonian PDEs to construct model hierarchies for every network node, ranging from PDEs to data-based models, in combination with structure preserving model order reduction and space-time-model-adaptive methods.
At the top of the hierarchy of models are the generators. These electrical machines are complex technological devices, as they intrinsically couple electromagnetism and mechanics during their operation. A model problem for this coupling is represented by the equations of magnetohydrodynamics (MHD), that describes the physics of ionized fluids such as plasma. Structure-preserving numerical schemes for MHD are currently under investigation. These algorithms can then be extended to electrical machines by replacing the material behaviour of the fluid to that of a solid. Once an high-fidelity description of electrical machines is available, simplifications can be introduced to obtain coarser models for other components of the electrical grid.