Large-scale doubly-fed wind farms are usually built in remote areas. The adoption of series capacitive compensation of transmission line can reduce energy losses. However

it can cause sub-synchronous oscillation in the system

which makes it difficult to conduct equivalent modeling of the doubly-fed wind farms. In this paper

an equivalent modeling method of doubly-fed wind farm based on dominant characteristic identification is proposed. First

a small signal model is set up based on the mathematical equation of doubly-fed wind power generation system

and the sub-synchronous oscillation mode of the doubly-fed wind farm is identified by the characteristic analysis method

and the influence of dominant characteristic variables on the sub-synchronous oscillation mode is verified by hardware-in-the-loop simulation. Then

the dominant characteristic variables of sub-synchronous oscillation mode is taken as the variable index of the doubly-fed wind farm

and doubly-fed fans are grouped in accordance with the hierarchical clustering algorithm. After that

the parameters of equivalent model are calculated by capacity weighting method. The equal power loss method is used to obtain the equivalent parameters of complex centralized network. Finally

the effectiveness of the equivalent method is verified by setting up the detailed and equivalent model of wind farm with 36 doubly-fed fans.