Robust active damping in LCL-filter based medium-voltage parallel grid-inverters for wind turbines

Peña-Alzola, Rafael and Roldán-Pérez, Javier and Bueno, Emilio and Huerta, Francisco and Campos-Gaona, David and Liserre GAE, Marco and Burt, Graeme (2018) Robust active damping in LCL-filter based medium-voltage parallel grid-inverters for wind turbines. IEEE Transactions on Power Electronics, 33 (12). pp. 10846-10857. ISSN 0885-8993 (

[thumbnail of Pena-Alzola-etal-IEEE-TOPE-2018-Robust-active-damping-in-LCL-filter-based-medium-voltage]
Text. Filename: Pena_Alzola_etal_IEEE_TOPE_2018_Robust_active_damping_in_LCL_filter_based_medium_voltage.pdf
Accepted Author Manuscript

Download (8MB)| Preview


LCL-filter based grid-tie inverters require damping for current-loop stability. There are only software modifications in active damping, whereas resistors are added in passive damping. Although passive damping incurs in additional losses, it is widely used because of its simplicity. This article considers the active damping in medium-voltage parallel inverters for wind turbines. Due to cost reasons, only minimal software changes are allowed and no extra sensors can be used. The procedure must be robust against line-inductance variations in weak grids. Double-update mode is needed so the resonance frequency is under the Nyquist limit. The bandwidth reduction when using active damping is also required to be known beforehand. Moreover, the design procedure should be simple without requiring numerous trial-and-error iterations. In spite of the abundant literature, the options are limited under these circumstances. Filter-based solutions are appropriate and a new procedure for tuning the notch-filter is proposed. However, this procedure requires that the resistance of the inductors is known and a novel filter-based solution is proposed that uses lag-filters. The lag-filters displace the phase angle at the resonance frequency so that the Nyquist stability criterion is fulfilled. Simulations and experiments with a 100 kVA prototype validate the analysis.