A Wind Power Combination Prediction Method Based on Wind Speed ​​Fluctuation Feature Extraction

Abstract

The invention discloses a wind power combination prediction method based on wind speed fluctuation feature extraction, comprising the following steps: normalizing the wind speed data collected by training samples; establishing a time window for the normalized wind speed, and Perform multi-fractal spectrum analysis within; analyze and compare the width ω of the value interval of the singular exponent α in each time window, the peak difference Δf(α) of the singular spectral function f(α), and the symmetry parameter S of f(α). The wind speed is classified according to the parameters [ω, Δf(α), S], and the size of the time window is further adjusted. Use extreme learning machine, support vector machine and optimized regression power curve methods to train the divided categories in turn, and compare the monthly average accuracy of the generated prediction results, and choose one of the methods as the optimal single algorithm for this category, Get a trained model. Carry out the same classification and modeling on the test samples, and predict the optimal single algorithm corresponding to different models, and finally combine to obtain the final prediction result.

Description

technical field [0001] The invention relates to a wind power combination prediction method based on wind speed fluctuation feature extraction, which belongs to the field of power system operation and control. Background technique [0002] With the gradual increase of the penetration rate of wind power in the entire power system of our country, the problems of voltage control, active power dispatching and system stability caused by its volatility, intermittent and randomness have become more and more prominent. Accurate wind power forecasting can not only reduce the System reserve capacity and energy storage reduce system operating costs, and at the same time help to reduce the impact of wind power access on the grid and improve grid operation reliability. [0003] The intermittent nature of wind energy in nature determines that wind power has strong fluctuations. As the number and installed capacity of wind farms continue to increase, once wind power is integrated into the g...

AI Technical Summary

Problems solved by technology

Therefore, the prediction accuracy of a single prediction model under different wind speed conditions needs to be improved

[0006] In addition, when the wind speed changes drastically, the frequency and intensity of the wind speed fluctuations in this time period will also increase

As time goes by, only building a model for historical data in a specific period of time cannot satisfy the dynamic characteristics of wind speed fluctuations on all time series

Moreover, the size of the time window has a crucial impact on the extraction and classification of data features. Too large and too small windows are not conducive to feature analysis.

Images