Large-Signal Stability Analysis of Inverter-Based AC Microgrids: A Critical and Analytical Review

With the ever-increasing development of microgrids to improve power supply reliability and resilience, large-signal stability analysis of microgrids is needed to provide accurate insights into system dynamics during significant disturbances and to offer a robust and comprehensive assessment of microgrids’ ability to return to normal operation after contingencies. This paper presents a comprehensive analytical and critical review of large-signal stability analysis methods for inverter-based AC microgrids supported by both theoretical analyses and numerical case studies. These analyses and case studies are developed and implemented to analyze, compare, and identify the gaps in existing large-signal stability assessment methods, which have not been investigated in the literature. More specifically, the recent developments in large-signal stability analysis techniques for microgrids, including Lyapunov-based methods and energy function analysis, are reviewed and discussed. Also, as the accuracy of the dynamic models of inverter-based resources (IBRs) is a crucial factor for authentic and accurate stability assessment, the impacts of full-, reduced-, and second-order dynamic models of IBRs on the accuracy of large-signal stability assessment in microgrids are numerically scrutinized. The study concludes by identifying the applicability of existing stability analysis methods for microgrids (e.g., Krasovskii’s, Popov-Lure, and sum of squares (SOS)-based methods) and presenting their challenges for further investigation in the field of large-signal stability of microgrids using numerical case studies. This comparative assessment of large-signal stability assessment provides an informative analysis of the system’s capability to operate under contingencies and helps to set the protection system efficiently and prevent unnecessary outages and trips. The numerical assessment shows that SOS-based stability assessment can provide a more realistic and less conservative stability region. It is worth mentioning that other approaches are computationally more efficient and can be applied for online applications and control objectives.