Extremum Seeking Control based Resonant Frequency Estimation for a Grid-Tied Inverter with LCL Filter

The resonant frequency of the LCL filter can drift away from its nominal value due to the circuit components parameters uncertainty and the grid-side impedance, which will further deteriorate the robustness of the system and damping performance. Conventional online resonant frequency estimation schemes, such as discrete Fourier transformation and recursive prediction error, require a significant amount of computational power to implement, which may even need to reduce the sampling and control frequency of the inverter controller. In this work, an extremum seeking control based resonant frequency estimation scheme is proposed, which can provide an accurate estimate of the resonant frequency with notably reduced execution time. Simulations, hardware-in-the-loop (HIL) tests, and experimental tests on a single-phase grid-tied inverter are conducted to validate the effectiveness of the proposed scheme.