Data-Light Oscillation Mode Identification for Fast Stability Assessment of Grid-Tied Converters

Dynamic mode decomposition (DMD) is an effective data-driven oscillation mode identification method for instability identification in grid-tied converters. However, time-consuming computation and empirical setting of parameters limit its practical application in rapid stability assessment. This paper introduces data-light dynamic mode decomposition (DL-DMD) by integrating the conventional dynamic mode decomposition (DMD) algorithm with harmonic state space (HSS) theory, which significantly improves computational speed while providing a bandwidth-oriented interpretation for parameter selection in correspondence with the identifiable frequency range of modes. The effectiveness and advantages of the proposed method are verified through offline analysis and online tests based on an experimental platform.