An Ensemble Neural Network Based on Variational Mode Decomposition and an Improved Sparrow Search Algorithm for Wind and Solar Power Forecasting
Zhiqiang Wu, Bo Wang
Abstract
Accurate forecasting methods for wind and solar power are important for power systems due to their potential to improve economic and environmental performance. For this purpose, an ensemble neural network framework composed of the LSTM, SVM, BP neural network and ELM is proposed for wind and solar power forecasting in China. Three common ways to improve prediction accuracy are adopted. First, unstable wind and solar power time series are decomposed into smooth sub-sequences by VMD, which reduces the undesirable effects caused by the volatility of the original series. Then, based on the decomposed sub-sequences, four basic models that are optimized based on the EOSSA algorithm are used for forecasting wind and solar power. Finally, the prediction results of the ENN are reconstructed by weighting the prediction results of the four models. The proposed ENN model is compared with 9 state-of-the-art prediction models for wind and solar power forecasting. The results show that the ENN model has the lowest MAPE, MAE, MSE, and RMSE in both wind and solar power forecasting. These comparison results also show that the ENN model not only has the best prediction accuracy but also the most reliable prediction performance.