Interactions between grid-connected inverters and the equivalent grid impedance seen at their point of common coupling have been identified as one of the main causes of instability problems. Therefore, the need for real-time monitoring tools to identify grid impedance variations cannot be underrated. Even though existing techniques have shown to estimate the grid impedance accurately, several concerns have remained unsolved. These include the trade-offs between the estimation accuracy and the magnitude and the time duration of the injected disturbance(s). Hence, this paper proposes a fast and accurate grid impedance estimation approach for the application of online impedance-based stability analysis of grid-connected inverter systems. First, the wideband grid impedance is estimated online based on the injection of two triangular pulses, allowing fast and multiple-input multiple-output (MIMO) identification in the synchronous (dq) reference frame. Then, curve fitting of the estimated wideband grid impedances is applied to obtain their parametric models. Afterward, the stability analysis is performed in the dq reference frame based on the generalized Nyquist criterion. The proposed algorithm allows accurate stability analysis over all frequencies as it relies on the parametric impedances. Most importantly, to optimize the trade-offs between the power quality and accurate stability analysis, the proposed algorithm is executed again and injects the two pulses disturbances to the grid only if the system operating conditions change (e.g., variations of the grid impedance). It is also applicable to perform the stability analysis while the inverter is still in standby mode. The simulation and experimental results of a 3-phase grid-connected inverter are presented to validate the efficacy of the proposed technique.
Read full abstract