In the growing interest in renewable energy, the role of power converters has become more important. However, as the use of power converters is elevated, the negative effects of power conversion also increase because the grid-connected inverters not only convert energy but also generate harmonics that degrade grid quality by their switching behavior. Using an LCL filter is a cost-effective solution to mitigate switching harmonics in the grid-connected inverters especially compared with conventional L filters.
Although the LCL filters have superior attenuation ability, the entire system could be unstable because of the LCL filter’s resonance characteristic. Various passive or active damping methods have been proposed to solve this problem. Digital notch filter based methods are simple ways to stabilize the system among the methods. The methods inhibit the system resonance phenomenon by simply aligning the notch frequency to the resonance frequency. However, the system is only stable where the system resonance frequency is matched with notch frequency. In the real systems, there is a chance of resonance frequency variation caused by filter aging or grid impedance changing, therefore, the resonance frequency estimation is necessary to avoid the stability problem excited by variation of the system parameters. Some of the estimation methods were proposed to solve the problem. However, The conventional resonance frequency estimation methods have some drawbacks such as unavailability during operation state or very slow estimation time.
This paper proposes a new adaptive digital notch filter that has the ability to track the system resonance frequency to maintain the system stable even the system parameters are changed where the inverter side current is controlled. The system resonance frequency variation tracking speed is considerably improved compared to the conventional methods. The case that the system resonance frequency and the notch frequency are not matched is simulated and experimented to verify the validity of the proposed method. It is expected that the proposed method can be used in the field of the grid-connected inverter system with improved stability.