Virtual inertia emulation in islanded microgrids using a time-series disturbance observer for frequency stabilization

Virtual inertia emulation (VIE) regulates the power exchange of energy storage devices (ESDs) to suppress the rate of change of frequency (RoCoF) and enhance frequency stability in low-inertia power systems. However, existing VIE strategies predominantly rely on damping augmentation or tuned controllers, which exhibit limited capability in rejecting fast-varying and aggregated disturbances arising from renewable intermittency and load fluctuations. To address this limitation, this paper primarily proposes the integration of a time-series disturbance observer (TSDOB) into the VIE loop of an islanded microgrid. The TSDOB estimates the total power disturbance as a time-series polynomial, enabling real-time disturbance reconstruction and compensation. To facilitate effective cancellation of the estimated disturbance, a PD-based dynamic inversion of the ESD dynamics is employed as a supporting feedforward mechanism. A baseline tilt–integral–derivative (TID) controller, optimally tuned using a hybrid PEO-GA algorithm, is incorporated to ensure nominal frequency regulation, resulting in a two-degree-of-freedom control architecture. Simulation results demonstrate that the proposed TSDOB-aided VIE framework significantly enhances frequency stabilization without increasing ESD energy throughput. Compared with the PEO-GA-TID and H∞ controllers, the proposed scheme reduces the maximum frequency deviation by 93.46% and 99.55%, respectively, and achieves 99.57% and 99.97% reductions in the integral of absolute error (IAE) of frequency deviation under real wind and solar power variations. Furthermore, under high-frequency disturbances up to 10 Hz, the proposed method attains at least a 79.12% reduction in frequency IAE relative to the baseline PEO-GA-TID controller, highlighting the effectiveness of the TSDOB in mitigating both fast- and slow-varying disturbances.