Answers are grounded in this paper and cite their sources.
Demonstrates enhanced accuracy in short-term wind power prediction, highlighting model effectiveness in real-time power monitoring.
A complete ensemble empirical mode decomposition with adaptive noise, bidirectional gated recurrent unit and transformer (CEEMDAN-BGT) based cross-attention mechanism short-term wind power prediction model has been proposed in this paper to improve the accuracy of short-term wind power prediction and better monitor the real-time dynamic changes in the power system. Firstly, Pearson's analysis was employed to analyze the input parameters affecting wind power generation. Secondly, CEEMDAN was used to decompose the original data sequence into multiple intrinsic mode functions (IMFs) reduce complexity. Thereafter, the long-term dependence feature of the sequence was derived using the Transformer encoder layer based on the multi-attention mechanism as auxiliary input information. Meanwhile, the global attention mechanism (GA) optimized bidirectional gated recurrent unit (BiGRU) network is introduced to extract the globally relevant timing features, to obtain point predictions wind power uncertainty, which the traditional GRU cannot provide accurately when the data undergoes sudden changes. Finally, the output modal component data was fused and superimposed via the cross-attention mechanism. The feature enhancement fusion was performed by calculating the attention weights to achieve high-precision point prediction. The model performance was evaluated, by analyzing the short-term wind power data acquired from a wind farm in Nanning City, Guangxi Zhuang Autonomous Region, and comparing the proposed method with the existing prediction methods. The experimental results revealed that the proposed model combined the advantages of Transformer, GRU and other models, and its R-Square is improved by 0.93%, 8.41%, 7.94% and 3.86% compared with the VMD-BGT, Transformer, TCN, and LSTM models, thus confirming the high prediction performance of the proposed model.
No takes yet. Share an insight, caveat, or question.
Qin et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: