Short-Term Offshore Wind Power Prediction based on VMD-SE-BP Neural Network Model
Hui Qin, Lijuan Huang, Kui Li, Gaihong Cheng · 2024
The strong volatility and randomness of offshore wind power output pose significant challenges to power grid dispatching. Accurate prediction of offshore wind power is of crucial importance for power supply-demand balance and ensuring the stable operation of the power grid. This paper proposes a combined offshore wind power prediction model based on variational mode decomposition (VMD), sample entropy (SE) and back propagation (BP) neural network. Firstly, the dataset is preprocessed to identify and handle outliers, followed by data normalization to mitigate differences among various data. Subsequently, wind power sequences are decomposed using VMD to obtain limited components. Furthermore, sample entropy of each component is calculated, and based on their similarity, components are categorized into random, trend, and oscillation components. Finally, the predictions of all components are aggregated to obtain the final offshore wind power prediction results. To validate the effectiveness and advancement of the proposed model, traditional BP neural network, extreme learning machine (ELM), and the proposed VMD-SE-BP model are applied to predict the actual power data from a certain offshore wind farm in China. Test results demonstrate that the proposed VMD-SE-BP combination model can effectively enhance the accuracy of wind power prediction.