On-line Coherency Analysis based on Sliding-Window Koopman Mode Decomposition

Nowadays, power systems complexity requires of innovative methods to monitor and provide an adequate online assessment. Coherency identification (based on data-driven methods) is a potential tool that can be integrated into the system infrastructure for the protection and resilience of the power grid. This work presents a modification of the Koopman Mode Decomposition (KMD) by adding a sliding-window to emulate the processed system's signals and to visualise the data concentration as a Transmission System Operator (TSO). Finally, we present a study of a data-set of rotor angle observables from the Nordic 32 test system after a disturbance to observe the rapid coherency at specific time-shots. This study provides evidence that the proposed modified KMD is a fast and robust approach to analyze large time-domain simulation data.