The goals of the GridAnalysis research project are the development of new models and methods for quasi-stationary and time-series-based simulation of power distribution grids in combination with classical grid calculation and machine learning methods. Furthermore, the partners will develop a guideline, which enables a (partly) automated operation of power distribution networks with the possibility of a topology change while ensuring a high supply quality and maintaining the selectivity of the network protection concept.
In cooperation with the project partners, the aim of this subproject is two-fold. (a) The conception and development of AI-based methods for the system analysis of power distribution networks in rural, suburban, and urban areas in normal and short-circuit operation. (b) The conception and development of an assistance system for non-linear systems, which supports a user in his decision-making. If required, the assistance system visualizes the results and evaluations determined by the AI-based network simulation in an understandable and comprehensible way to increase the user's acceptance in the AI-based system analysis.
