Clustering analysis of the wind power output based on similarity theory

The most important difference of wind power generator sets is the uncontrollability of their output compared to traditional ones, and it is an important issue to describe the fluctuation that affects on power system and to development a corresponding planning strategy. This paper uses a fuzzy clustering method and similarity theory to classify different periods into different time category and then choose a fixed output value to represent the output of wind power in category respectively. The wind speed distribution function is used to describe the characteristics of classified wind speed data, and expected output of wind power is obtained via a typical wind turbine power curve. The results clustered can provide a theoretical basis for the production simulation and forecasting analysis of wind power integrated system.