Wind power interval prediction combination method based on signal decomposition

Abstract

The invention discloses a wind power interval prediction combination method based on signal decomposition, and belongs to the technical field of wind power generation output prediction. According to the method, interval prediction combination is achieved through three steps of data preprocessing, interval prediction and subinterval synthesis. According to the prediction model, firstly, naive Bayesian point prediction is carried out on sub-component signals, a kernel extreme learning machine weighted by an entropy weight method and a kernel density estimation combined model are established, and point prediction errors are input into the combined model to obtain prediction sub-intervals; and finally, the sub-intervals are synthesized to obtain a final prediction result. According to the method, the non-stationarity of a wind power signal is considered, and signal decomposition is carried out before prediction. The influence of an unstable signal on the prediction precision is reduced. Aiming at the inherent defects of a single model, different models are combined, complementation is formed to a certain extent, and the prediction precision is improved.

Description

Technical field[0001]The present invention belongs to the field of wind power, and is specifically involved in a wind power interval prediction combination method based on signal decomposition.Background technique[0002]In order to cope with the gradual depletion of fossil energy and the environmental pollution problem brought by fossil energy, the world is vigorously developing new energy industries. Wind power generation has been vigorously developed by countries due to low cost, clean and renewable advantages. However, due to the uncertainty of wind power itself and the anti-tone peak characteristics, the scheduling is difficult, the large-scale wind power access power system will bring a serious impact on the grid. Therefore, the wind power that is highly efficient and large-scale entry is the urgent task of electricity development. For my country's electricity production and consumption layout, the wind power base is mainly concentrated in the "three-north" area, and electric en...

AI Technical Summary

Problems solved by technology

On the one hand, due to the algorithmic defects of various prediction models, errors are inevitable; on the other hand, the wind power signal is a non-stationary signal. When a single model is used for prediction, the non-stationary signal will significantly reduce the prediction accuracy.

Images