Multi-wind-field wind speed space-time prediction method based on graph convolution and recurrent neural network

Abstract

The invention discloses a multi-wind-field wind speed space-time prediction method based on graph convolution and a recurrent neural network. Aiming at the problem that a convolutional neural network commonly used in an existing wind speed space-time prediction method is difficult to effectively analyze wind speed data of multiple wind fields presenting non-grid distribution in reality, a graph convolutional long-short-term memory neural network is provided to process the data. The method comprises the following steps: firstly, performing graph modeling on wind speed data of multiple wind fields based on a Pearson correlation coefficient to construct a wind speed graph signal sequence; then, using graph convolution to replace multiplication in the long short-term memory neural network, and constructing a graph convolution long short-term memory neural network; and finally, constructing a multi-wind-field wind speed space-time prediction model based on the graph convolution long-short-term memory neural network and a transfer learning principle. The space-time prediction model has good point prediction and probability prediction performance, it is verified that the accuracy of wind speed point prediction and probability prediction can be improved by fusing historical wind speed information of adjacent wind fields, and a new thought is provided for short-term wind speed prediction of multiple wind fields.

Description

technical field [0001] The invention relates to a wind speed prediction method, in particular to a multi-wind field wind speed spatio-temporal prediction method based on graph convolution and cyclic neural network. Background technique [0002] Wind energy is a clean, non-polluting renewable energy source, which has the characteristics of abundant resources and clean power generation process, so it has been widely promoted and applied. However, due to the fluctuation and intermittence of wind speed, wind power has strong fluctuation and randomness. At present, the spatio-temporal prediction methods of wind speed usually include the spatio-temporal prediction model based on correlation analysis and the spatio-temporal prediction model based on convolutional neural network. Among them, the spatio-temporal prediction model based on correlation analysis uses correlation analysis methods such as Pearson coefficient and mutual information to judge the degree of correlation betwee...

AI Technical Summary

Problems solved by technology

[0003] Purpose of the invention: The present invention aims at the problem that the current convolutional neural network can only process data in Euclidean space, but it is diff

Images