The invention relates to a wind power prediction method based on a fuzzy rough set and a radial basis network. The method comprises the following steps: S1, constructing a meteorological matrix and afan historical output matrix; S2, performing standardization processing on the meteorological matrix and the fan historical output matrix by adopting a normalization method; S3, performing attribute reduction on the meteorological matrix by using a fuzzy rough set method; S4, building a wind power prediction model based on the radial basis network; Establishing a function relationship between thereduced meteorological value matrix and the wind power standardized value according to the mapping relationship of the radial basis network; S5, verifying the effectiveness of the wind power prediction model built in the step S4 by using the collected and arranged data, and carrying out inverse standardization on the predicted standard power value to obtain an actual fan output prediction value. According to the method, redundant parts in meteorological physical value attributes are removed, the wind power prediction precision is improved, and the calculation speed of the model is increased.