Laurent Pagnier – The University of Arizona, United States
With the increasing penetration of renewables, power systems are getting more prone to significant disturbances. It is therefore paramount to have efficient tools to assess their vulnerabilities and enhance their resilience. At the same time, data concerning their operation are increasingly recorded and made available. In this seminary, we show different applications of machine learning for power systems. In particular, we present effective reduced models of large electric transmission grids that take the form of PDE equations over the areas that are covered by them. We show how such models can be calibrated from recorded data.