Abstract: The large number of wind turbine generators (WTG) in power systems with large-scale wind power integration, high model order, complex solution, and the existence of structural parameter privacy problem of WTGs lead to the limitation of system frequency response modeling. For this reason, this paper proposes a practical frequency response model and its parameter identification method in the power system with large-scale wind power integration, considering information privacy. First, the frequency-active transfer function of a doubly-fed induction generator (DFIG) with synthetic inertia control is derived, and the frequency response model of DFIG is constructed; then, for the power system containing multiple wind farms, the system is aggregated and equated into a two-machine model based on the capacity-weighted averaging method, and the order of approximate frequency response model for power system with large-scale wind power integration is determined. Further, to address the problem of incomplete information on the structural parameters of DFIG, the fourth-order model of the output error considering the influence of noise is constructed to fit the system frequency response model, and the parameters of the frequency response model are further identified based on the least squares iterative algorithm of the auxiliary model. Finally, the accuracy and practicability of the proposed fourth-order SFR model and parameter identification method are verified by a 5-bus test system with a wind farm and a 220kV real power system simulation in the western part of a province.