Clustering wind turbines for a large wind farm using spectral clustering approach based on diffusion mapping theory

Increasing number of wind farms connected to power system calls for a simplified equivalent model of wind farm so that it can be represented by only a few equivalent wind turbines for dynamic system studies. In order to aggregate wind turbines in complex terrain or irregular layout, this paper presents a new wind turbine clustering method based on spectral clustering algorithm. The clustering principle relies on the same or similar operating points of selected wind turbines. The proposed method is used to capture the similarity of characteristics of wind turbines by analyzing real-time data sets of wind turbines in a wind farm. Thus wind turbines of a wind farm can be clustered into different groups. The clustering method is demonstrated and tested in an actual wind power system. A good performance is achieved for all simulation scenarios while comparing the aggregated and detailed models.