Grey box aggregation modeling of wind farm for wideband oscillations analysis

The power electronic converter-based wind farms have imposed great risks on the small-signal stability of the modern power system, mainly characterized by wideband oscillations ranging from one Hertz to several thousand Hertz. To analyze such wideband oscillations problems, a consensus has been reached that the aggregation of the wind farm is indispensable. However, most of the present aggregation studies are focused on the time-domain consistency (e.g., output active/reactive power) before and after aggregation, while the frequency-domain consistency, closely associated with the accuracy of wideband oscillations analysis, has not been paid much attention. To fulfill this gap, in this paper, a wideband grey-box aggregation model of the wind farm with its parameter identification method are presented, which can be acquired without knowing the details of the wind farm. The identified aggregation model shows a good conformity with the large scope frequency characteristics (e.g., 1–2500 Hz) of the detailed wind farm, suitable for the wideband oscillation analysis. Firstly, the trajectory sensitivity and wideband characteristics of the wind farm are studied, based on which a reduced-order grey-box aggregation model is established in low-medium and high frequency band, respectively. Secondly, a vector fitting-based parameter identification method is presented for the established grey-box aggregation model. At last, the influences of the wind speed distribution and transmission line length on the accuracy of the proposed aggregation method are studied, where a comprehensive deviation analysis is carried out statistically.