Photovoltaic Power Prediction Model Based on TDE-SO-AWM-GRU

Accurate prediction of photovoltaic (PV) power is of great significance for the safe and economic operation of power grids, but PV power generation is characterized by time-varying, intermittent, fluctuating, and high nonlinearity characteristics due to the combined effects of multiple meteorological factors, which makes it difficult to deeply explore the implicit information of the data. To solve such problems, a PV power prediction model based on time-varying data enhancement (TDE), snake optimizer (SO), adaptive weight module (AWM), and gated recurrent unit (GRU) was proposed. The expression of data features was improved by TDE with strong correlation, and a new input matrix was constructed. Then, the enhanced input matrix was automatically weighted by AWM and entered into GRU for prediction. At the same time, by considering the difficulty of hyperparameter selection of the combined model, SO was introduced to find the optimal threshold of the model to maximize the performance of the model. Finally, the model was validated using actual data from a PV power station, and the results show that the proposed model can effectively improve the prediction accuracy of PV power.