Real-time execution of linear ringdown analysis methods for identifying dominant modes

Abstract

This paper specializes in the real-time execution of three mature linear ringdown analysis methods for modal identification in power systems. Data-based identification methods such as Prony’s method, eigensystem realization algorithm, and matrix pencil are embedded into a Matlab & Simulink-powered real-time simulation environment. Their implementations are achieved by using a sliding window approach, are profited by the inherent features of a parallel computing software architecture, reduce the computational complexity of the methods through shorttime windows length, and provide instantaneous modal information (damping and frequency). These enhancements make part of a user-friendly tool to effectively deal with the modal identification issue. This tool is successfully tested using two test power systems: the two-area Kundur system and a reduced-order representation of the New England power grid. Its effectiveness and performance are demonstrated with the attained results and its validation.