Ultra-short-term wind power combined prediction method and system

Abstract

The invention relates to an ultra-short-term wind power combined prediction method and a system thereof. The method comprises the following steps: training a long-short-term memory network by utilizing wind power output historical data; training an extreme learning machine by using the numerical weather forecast data; according to the predicted value of the long-short-term memory network, the predicted value of the extreme learning machine, the wind power, the numerical weather forecast data, the wind speed variation and the rolling prediction step number of the long-short-term memory network, determining historical similar moments of which the similarity with the numerical weather forecast data at the moment to be predicted is greater than a similarity threshold by adopting a weighted grey correlation algorithm; according to the predicted values of the two models corresponding to the historical similar moments, the wind power, the numerical weather forecast data and the actual wind power output data, adopting a time-varying adaptive coefficient method to determine combined prediction weight parameters of the two models, and then determining a combined prediction model; and predicting the wind power at the prediction moment according to the combined prediction model. The method can improve prediction error distribution and improve prediction precision.

Description

technical field [0001] The invention relates to the field of wind power forecasting, in particular to an ultra-short-term wind power combination forecasting method and system. Background technique [0002] As an important part of renewable energy, wind power is developing rapidly in China, but the inherent randomness and intermittency of its natural energy properties also pose a huge challenge to ensure the stability and reliability of the power system. Accurate and reliable wind power forecasting can not only improve the penetration rate of wind power, but also play an important role in the reliable and stable operation of the power system and the improvement of the economical operation of the power grid. [0003] The wind power forecasting methods at home and abroad can be mainly divided into physical methods and statistical methods. The physical method usually obtains meteorological information such as forecast wind speed, wind direction, and temperature in numerical wea...

AI Technical Summary

Problems solved by technology

Due to the accumulation of rolling forecast errors and sudden changes in the trend of wind power changes, as well as the fact that the input features cannot be updated in time during the long-term short-term memory network forecasting process, the input features cannot track the trend of wind power changes.

Images