Introduction to Koopman Mode Decomposition for Data-Based Technology of Power System Nonlinear Dynamics

Data-based technology of power system dynamics has attracted a lot of interest in the modern power system with the so-called Wide-Area Measurement System (WAMS). Concerted efforts are currently being made to develop the technology based on linearized dynamic models. In contrast, the full nonlinear models have received less attention in this research community. We present the research challenges to go beyond the conventional linear framework by focusing on Koopman Mode Decomposition (KMD), which is a nonlinear generalization of linear oscillatory modes guided by operator theory of nonlinear dynamical systems. Our discussion begins with a review of the main tools—Koopman operator and Prony approximation of KMD. We pose several distinct problems on data analysis, computation, monitoring, situational awareness, and control for the future power system architecture. Our idea is then illustrated on how the Prony approximation of KMD is distributed in WAMS.