Wind power prediction method and system for optimizing deep Transformer network

Abstract

The invention discloses a wind power prediction method and system for optimizing a deep Transformer network based on a whale group algorithm. The method comprises the steps of taking the sequence data of wind power and related influence factors as sample data, and dividing the sample data into a training set and a test set, establishing a Transformer network model according to values of the initialized hyper-parameters, respectively training and predicting data in the training set and the test set, and taking an average absolute error of wind power prediction as a fitness value of each whale group, determining a local optimal position according to the initial fitness value of the individual whale group, updating the current optimal position by utilizing whale group optimization, and obtaining the best prediction effect by comparing the local optimal solution with the global optimal solution, and obtaining an optimal hyper-parameter combination in the Transformer network after a plurality of iterations of the whale group algorithm, and predicting the wind power. According to the method, the optimization algorithm and the deep learning prediction algorithm are combined, so that the wind power prediction accuracy is greatly improved.

Description

technical field [0001] The invention belongs to the field of power system planning, and more specifically relates to a wind power prediction method and system based on a whale group algorithm to optimize a deep Transformer network. Background technique [0002] Due to the shortage of fossil fuels, the problems of environmental pollution and greenhouse effect are becoming more and more serious, the development and utilization of renewable energy has gradually attracted the attention of the whole world. Wind energy is a clean energy with huge reserves and high development potential. Wind power generation is one of the main forms of wind energy utilization, and has received more and more attention from researchers in recent years. When wind energy is used for power generation on a large scale, it is necessary to predict the wind power to ensure the reliability, stability and economy of the grid. Wind power forecasting is an indispensable key link in wind farm integration, whi...

AI Technical Summary

Problems solved by technology

Traditional neural networks with recurrent structure, such as Recursive Neural Network (RNN), Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) are dealing with this complex There are difficulties in the time series, and the inability to perform parallel operations makes the model training time too long

Images