Comparing State Estimators for Robust Fast Extremum-Seeking Control for Frequency Support of Low-Inertia Power Systems

Low inertia power systems such as microgrids are vulnerable to frequency stability issues which can lead to blackouts. Existing control strategies to support the frequency require accurate models. Fast extremum-seeking control (FESC) is emerging as a solution to provide fast frequency support (FFS) due to the lower dependency in the model accuracy. However, the performance of FESC relies on reliable state estimation, which is challenging when estimators experience a mismatch between system models used in the design stage and its actual dynamics driven by disturbances and uncertainties. Thus, finding an estimator that depends less on the system model and provides reliable state estimation during dynamic changes is crucial to match the benefits of FESC. This paper investigates the performance of prominent state estimation methods (i.e., extended Kalman filter (EKF), ensemble Kalman filter (EnKF), cubature Kalman filter (CKF), and moving horizon estimation (MHE)) under changing system dynamics against the system consideration for the estimators for robust implementation of FESC in microgrids for FFS application. Matlab/Simulink simulations on a microgrid test system show the effectiveness of the ensemble-based EnKF in achieving close approximations of the system states with minimal model reliance when the system undergoes dynamic events and its capability to enable FESC to maintain robustness in providing optimal control input to the system. The proposed research explores prominent solution for applications where dynamic estimation and control are indispensable.