Ultra-short-term wind power prediction method based on DCCSO optimization deep learning model

Abstract

The invention relates to the technical field of wind power prediction, in particular to an ultra-short-term wind power prediction method based on a DCCSO optimization deep learning model. The method comprises the following steps: collecting original wind power data, preprocessing the original wind power data, and establishing a time sequence attention-gating cycle unit deep learning prediction model; then optimizing an initial weight and a threshold value of the sequential attention-gated loop unit deep learning prediction model based on an improved crisscross algorithm, so that the convergence speed and the generalization performance of the sequential attention-gated loop unit deep learning prediction model can be effectively improved, wherein in the time sequence attention-gated loop unit deep learning prediction model, the time sequence attention can improve the sensitivity of the model to the input time, and the gated loop unit can further mine the hidden time correlation in the input time sequence; besides, the combination of the time sequence attention and the gating circulation unit is of great significance for improving the wind power prediction precision.

Description

technical field [0001] The present invention relates to the technical field of wind power forecasting, and more specifically, to an ultra-short-term wind power forecasting method based on a DCCSO optimized deep learning model. Background technique [0002] Existing wind power prediction methods can be divided into physical methods and data-driven prediction methods. The modeling of the physical prediction method is relatively complicated and the calculation cost is high. At the same time, the model is easily affected by the actual environment and has the problem of poor anti-interference ability. Data-driven forecasting methods can be subdivided into statistical methods and data-driven forecasting methods using artificial intelligence algorithms. Statistical methods obtain corresponding statistical relationships through research on historical data to achieve future wind power forecasting, but there are assumptions The data is linearly related, so it cannot reflect the nonli...

AI Technical Summary

Problems solved by technology

For example, Chinese patent CN200910193820.6 discloses a wind power prediction method based on genetic algorithm to optimize BP neural network. When genetic algorithm is used for ultra-short-term wind power prediction, the running time is too long; Chinese patent CN107274012A discloses a particle swarm algorithm based on cloud evolution The short-term wind power forecasting method, when used for ultra-short-term wind power forecasting, the particle swarm optimization algorithm is easy to fall into the local optimal solution

Images