Research on Photovoltaic Prediction and Load Voltage Regulation Based on Transformer

In recent years, with the widespread application of renewable energy, solar photovoltaic power generation has gradually been widely integrated into medium and low voltage distribution networks. Photovoltaic power generation poses greater risks to the transmission system, and new power systems require photovoltaic output prediction and voltage regulation. This article first uses Pearson correlation analysis to screen the main factors that have a significant impact on photovoltaic power, and then establishes a neural network method based on Transformer structure as input for photovoltaic power prediction. The power flow calculation method uses predicted photovoltaic output and some line parameters to calculate active and reactive power, ensuring stable operation of distribution lines. All experimental data in this article were collected in real-time from the Tongwei area of Gansu Province, and were divided into training and testing sets. This method reduces the relative error of photovoltaic output prediction to 13.32% and can effectively regulate voltage.