The invention provides a method for processing data and identifying fluctuations of historical wind power output, which consists of the steps of: collecting historical wind power output data and analyzing & arranging the output data; carrying out the normalization processing and wavelet filtering processing of the analyzed and arranged historical wind power output data and then carrying out fluctuation division; checking the historical wind power output data and the wind fluctuation clustering effect by means of a visualization method. The invention provides the method for processing data and identifying fluctuations of historical wind power output, which can be used as the pre-stage basis of long-term output time series modeling of wind power and can also be used for the planning and operation analysis of the power system including wind power.