Power system inertia estimation during normal operation using adaptive ensemble empirical mode decomposition

Power system is a critical parameter for assessing anti-disturbance capabilities and ensuring stability. Accurate estimation of power system inertia holds significant importance for scheduling operators. The current methodology for estimation remains perturbation-based, which is constrained by the occurrence of significant disturbances in practical power grids, thereby limiting the applicability of these approaches. To address the challenge of extracting effective features during normal operation, this paper proposes the adaptive ensemble empirical mode decomposition (AEEMD) method for inertia estimation under normal operating conditions. The approach involves adaptively computing white noise for the system, utilizing AEEMD to extract frequency characteristics, and estimating the effective inertia constant by assessing the maximum rate of change of the intrinsic mode functions (IMFs). The paper discusses the impact of primary frequency regulation on the estimation of effective inertia. It not only considers the inertia provided by synchronous generators but also takes into account the scenarios where loads can provide inertia, as well as the operational conditions where inverter-based resources (IBRs) provide inertia support. The efficacy of the proposed method is validated through simulations of the Kundur's four-generator two-area system in Simulink, as well as measurements from the power grid.