Distributed Adaptive Under Frequency Load Shedding of Islanded Microgrid Based on Local Information

Under frequency load shedding (UFLS) scheme is particularly important for islanded microgrid (IMG) which has limited resources and low inertia. The existing research on adaptive under frequency load shedding (AUFLS) schemes mainly focuses on improving the performance. However, they are highly dependent on centralized communication networks, which makes them vulnerable to communication link failures. Meanwhile, traditional multi-stage UFLS schemes eliminate communication dependencies but suffer from substantial frequency deviations due to fixed threshold settings under large power imbalances. To address these challenges, this paper proposes a novel AUFLS scheme that dynamically estimates the system's power deficit solely through local load measurements, thereby minimizing reliance on communication infrastructure. Furthermore, the proposed scheme utilizes the initial rate of change of frequency (RoCoF0) before communication network failure as a triggering threshold, significantly accelerating the load-shedding response. First, a comprehensive system frequency response (SFR) model incorporating load regulation characteristics is established, analytically linking RoCoF0 to the minimum transient frequency. Then, a virtual equivalent inertia calculation method based on post-disturbance load dynamics is derived to quantify localized active power deficits. Finally, by integrating the pre-fault SFR, the activation thresholds are optimized to realize a distributed AUFLS framework requiring no central controller intervention. Simulation results demonstrate the proposed scheme's superior performance over traditional approaches in both response rapidity and frequency nadir preservation.