The Application of Deep Neural Network Algorithm Considering multi-dimensional Meteorological Feature Extraction in Short-Term Wind Speed Prediction

Abstract

Wind speed prediction is very important for grid-connected dispatch of wind power generation. In order to improve the accuracy of short-term wind speed prediction, a CNN+LSTM depth neural network model based on multi-dimensional meteorological feature extraction is established in this paper, aiming at the situation that there are many meteorological features and it is difficult to extract historical wind speed data. Firstly, multi-dimensional historical meteorological data are collected through multiple channels and divided into dry season and rainy season. Secondly, the multi-layer perceptron (MLP) is used to extract the optimal meteorological features from the multi-dimensional meteorological features, and the optimal meteorological characteristics and historical wind speed are used as input of the prediction model. Then, the CNN+LSTM depth neural network is used to establish short-term meteorological conditions. Wind speed prediction model. The results of wind speed prediction for a wind field in Gansu Province show that the short-term wind speed prediction model considering multi-dimensional meteorological feature extraction in this paper improves the calculation speed and accuracy, enhances the generalization ability of the model, and has a certain practical value.