SGMD-TSNE-TCNre long-term prediction of wind power based on continuous learning
YANG Xiaohua
DAI Shengguo
LI Jiahao
LI Jia
Abstract：For wind power prediction,a time convolutional network(TCNre)algorithm based on continuous learning is proposed,and combined with symplectic geometric mode decomposition(SGMD),t-distribution and random neighbor embedding(TSNE)data processing methods,the prediction model SGMD-TSNE-TCNre is constructed.In order to verify the effectiveness of the continuous learning method,a continuous learning model(TCNre)based on parameter freezing is built and compared with the TCN model.On this basis,considering that the power of wind turbines is affected by many complex factors,SGMD model is introduced to reduce the non-stationarity caused by environmental factors,and TSNE is used to reduce the input dimension of the model.One year's measured data of a wind farm is used to verify the results,and compared with other common prediction models.The results show that the SGMD-TSNE-TCNre model is effective and has higher accuracy.
Keywords：wind power forecastTCNresequential convolutional neural networkmodal decompositionfeature dimension reduction TSNE
Publication Date：2025-01-31
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：5( 21-25 )