The fast-full coverage of regional wind farms of wind power prediction is very important for optimal scheduling of wind farms;the existing prediction algorithms are unable to solve this problem.Consequently

we adopted fuzzy clustering analyses to achieve an effective recognition of the wind power characteristics and a reasonable division of wind farm clusters.The principal component analysis was used to extract the region weather characteristic parameters

and a framework of wind power prediction method was established based on clusters.The results show that the prediction accuracy of a single cluster is slightly lower than the traditional method which models for each wind farm

with the expansion of the forecast range

the prediction accuracy will reach the level of the traditional method.The prediction method based on wind farm clusters can ensure the prediction accuracy

at the same time

achieve a significant efficiency improvement compared to the traditional method.