Adaptive digital notch filter based on online grid impedance estimation for grid-tied LCL filter systems

Abstract

In grid-connected systems, LCL filters are playing an important role in mitigating the current harmonics as compared with the traditional L filter. Nevertheless, the performance of the LCL filters might deteriorate under the variations in grid impedance conditions resulting in excessive resonance phenomenon and even system instability hazards. The digital notch filter is an effective solution to tackle the resonance frequency problem by eliminating the resonant peak value; however, the prior art is only designed under the consideration of unknown grid impedance. This paper presents a novel implementation of an adaptive digital notch filter based on on-line grid impedance estimation to ensure the robustness against grid impedance variations. Stability analysis to confirm the relationship between the system parameters and grid impedance value is carried out. Furthermore, an extended Kalman filter is applied to design an observer which can estimate the grid impedance accurately. Considering the uncertainty of the real-time system, a lookup table based on the estimated grid impedance value and the control gain is proposed to tune the parameters of the digital notch filter to increase the relative stability of the system. Finally, the effectiveness of the developed adaptive notch filter is verified by both simulation and experimental results under varying grid impedance conditions.