Research on photovoltaic power combination model prediction based on 5G+ technology and irradiance correction

Accurate prediction of photovoltaic power is crucial for the dispatch and operation of power systems. Aiming at the problem of poor photovoltaic power prediction due to the deviation of irradiance forecast and the low prediction accuracy of a single model, a combined model of extreme gradient boosting (XGBoost) and the Bidirectional Gated Recurrent Unit (BiGRU) based on 5G+ technology and irradiance correction is proposed for photovoltaic power prediction. Firstly, the global horizontal total irradiance is updated in real time by using the high-speed data transmission and low-latency characteristics of 5G+ technology, and a slope irradiance correction model is established by the slope conversion method to ensure the accuracy of the slope irradiance on the photovoltaic panel. Then, important features are selected as the input of the combined prediction model according to their contribution to the output power. Finally, a comparative experiment is carried out on a high-performance computing platform supported by 5G+ network using real operating data. The results show that compared with other methods, the proposed method reduces the MAE by at least 17.56% and the RMSE by at least 14.19%.