As the integration of wind power into electrical energy network increasing, accurate forecast of wind speed becomes highly important in the case of large-scale wind power connected into the grid. In order to improve the accuracy of wind speed forecast and the generalization ability of the model, Extreme Gradient Boosting (XGBoost) as an improvement from gradient boosting decision tree (GBDT) is trained and deployed in the cheaper central processing unit (CPU) devices instead of graphics processing unit (GPU) devices, thus, a wind speed forecast model based on Extreme Gradient Boosting is proposed in this paper. Firstly, the historical data is taken as a part of the input vectors for the model. Moreover, considering the monthly change of wind speed characteristics, the dataset of wind power is divided into four parts by month so that the models are constructed in different complexity by month. Finally, compared with back propagation neural networks (BPNN) and linear regression (LR) models, the experimental results show that the improved XGBoost model can promote the forecast accuracy effectively. INDEX TERMS Short-term wind speed forecasting, XGBoost, time series, historical characteristics, power grid.
With the rapid development of social media, the number of online comments has exploded, and more and more people are willing to express their attitudes and feelings on the Internet. Under the influence of a series of major events all over the world, production order is facing a serious challenge, which severely impact on energy market. In 2020, a large number of investors' and consumers' comments related to social events began to appear on the Internet from China. However, the style and quality of online comments vary greatly, making it difficult to accurately extract users' views and tendencies. Based on the investors' and consumers' statements published on Chinese Internet, in this paper, we use statistical methods to classify the sentiment orientation of the Netizens firstly, and then use Bidirectional Encoder Representations from Transformer-Bidirectional Long Short Term Memory (BERT-BiLSTM) which is the combination forecasting method of Bidirectional Encoder Representations from Transformer (BERT) and Bidirectional Long Short Term Memory (BiLSTM), to model and forecast the sentiment orientation of users' statements, as well as being compared with its based models, BERT and BiLSTM. Among them, the accuracy and recall,which represent the predictive abilities on the overall samples and on the focus(the samples of Label 1) respectively, of BERT-BiLSTM model are 0.8620 and 0.7078 respectively, which are superior to those of BERT model (0.8559 and 0.5576) and LSTM model (0.7775 and 0.0747). The research results can accurately predict the sentiment orientation of Internet users during the social events so as to provide technical support for grasping the energy market trend. INDEX TERMS sentiment analysis, combination model, BERT-BiLSTM, energy market, investors and consumers.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.