Abstract: In this paper, a double-fed wind farm polymerization method based on similar coherency is proposed. Firstly, Prony algorithm is used to extract the dynamic characteristics of doubly-fed induction generators （DFIG） in wind farms. Similarity theory is used to find the similar parts of oscillation modes between doubly-fed induction generators. And the frequency, damping, amplitude and other information in the similar oscillation mode are used to define the similar metrics of the coherent discrimination of the DFIG and determine the coherent cluster. Then, according to the external characteristics of the DFIG, a power aggregation algorithm is proposed. The transient internal potential of the equivalent generator is obtained using the active power output of the DFIG as the weight, and the parameters of the equivalent bus and the equivalent generator are obtained. This method systematically solves the problem of coherent aggregation of different doubly-fed wind farms from the aspects of coherent cluster identification, coherent bus aggregation, and coherent cluster parameter aggregation. It effectively solves the complexity and limitations of clustering and proposes a method for simplification of buses for coherent doubly-fed clusters. The parameter aggregation method effectively preserves the electrical and structural characteristics of the original system. Finally, the simulation is conducted with the example of wind farms in an area of Yunnan Power Grid.

Key words: similar coherency, doubly-fed wind farm, the similar metrics of the coherent discrimination, power aggregation algorithm