Wind power prediction method based on combined model

The invention discloses a wind power prediction method based on a combined model, and the method specifically comprises the following steps: S1, data collection: the obtained data comprises the actual measurement data of a wind power plant, the data of numerical weather forecast and the start-up capacity of a wind turbine generator, and the actual measurement data of the wind power plant comprises the data of a wind measurement tower of the wind power plant and the characteristics of the wind power plant; the data of the numerical weather forecast comprises the data of the numerical weather forecast; s2, data preprocessing: carrying out smoothing processing and noise filtering on the collected data to realize data preprocessing; and S3, feature extraction: extracting statistical features of the data by using a statistical method, including the average value, variance and kurtosis of the data, the wind power can be better predicted, and the prediction precision of short-term power prediction is effectively imp