Wind power prediction method based on morphological filter and similar section search algorithm

Abstract

The invention discloses a wind power prediction method based on a morphological filter and a similar section search algorithm, comprising the following steps: (1) designing a high-frequency filter; (2) for an average trend component m(t) which is a low-frequency component, replacing global prediction with a local prediction method, searching for similar sections based on Euclidean distance, and using the similar sections to train a least squares support vector machine model so as to perform prediction; (3) using a local prediction method (namely, according to certain principles) to select part of data from historical data for modeling and prediction, and adopting a similar section search algorithm (namely, a 'secondary similarity' algorithm); and (4) synthesizing the prediction results of two independent components to get a final result of wind power prediction. The method has the advantages of clear physical significance of wind plant prediction power prediction, stable prediction result, and high prediction precision.

Description

technical field [0001] The invention relates to the technical field of wind power generation power forecasting, in particular to a wind power forecasting method based on a morphological filter and a similar segment search algorithm, and is a renewable energy power generation power forecasting technology. Background technique [0002] With the scarcity of resources and the growing call for environmental friendliness of human beings, the development and utilization of wind energy has been paid more and more attention. However, due to its strong randomness and intermittent nature, it brings great difficulties to wind energy prediction, which directly limits the application of wind energy in large power grids. It is of great significance to carry out wind energy forecasting. First of all, for wind farms, the evaluation and prediction of wind energy is an important part of the work to assess whether large-scale wind power projects are feasible. Furthermore, for the entire power...

AI Technical Summary

Problems solved by technology

[0004] (1) The prediction accuracy of existing prediction models largely depends on the user's prior knowledge, and the predictions have relatively large limitations;

[0005] (2) The inherent laws of wind energy changes and the physical meaning or characteristics reflected in the data are not fully explored;

[0006] (3) The prediction accuracy is not high, and the stability needs to be strengthened

Images