Short term wind power forecasting model based on secondary decomposition and hybrid deep neural network
HE Ningjing1, ZHANG Cheng1,2
1. School of Electronic, Electrical Engineering and Physics, Fujian University of Technology, Fuzhou 350118; 2. Fujian Provincal University Engineering Research Center for Simulation Analysis and Integrated Control of Smart Grid, Fujian University of Technology, Fuzhou 350118
Abstract:Given the volatility and randomness of wind power, a model for short term wind power forecasting which utilizes secondary mode decomposition and an secretary bird optimization algorithm (SBOA)-optimized temporal convolutional netwaork (TCN)-bidirectional gate recurrent unit (BiGRU)- Attention mechanism to enhance prediction accuracy. Firstly, complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and variational mode decomposition (VMD) algorithms are applied for secondary mode decomposition of the wind power time series. Secondly, the decomposed sub-series are fed into the SBOA-TCN-BiGRU-Attention network for combined prediction, with the SBOA optimizing the neural network's hyperparameters to avoid local optima. Finally, the predicted values of the sub-series are aggregated to derive the final result. The simulation findings indicate the proposed combined forecasting method predicts short term wind power with high accuracy.
何宁静, 张程. 基于二次分解和混合深度神经网络的短期风电功率预测[J]. 电气技术, 2025, 26(9): 34-44.
HE Ningjing, ZHANG Cheng. Short term wind power forecasting model based on secondary decomposition and hybrid deep neural network. Electrical Engineering, 2025, 26(9): 34-44.