Net Power Prediction for High Permeability Distributed Photovoltaic Integration System

Abstract

The large-scale grid-connected access to distributed PV power generation has posed a great challenge to the new power system. Distributed PV power output and power load have strong uncertainty and volatility, which increases the difficulty of distribution network net power prediction to a certain extent. To improve the prediction accuracy of distribution grid net power, the paper proposes a combined distribution grid net power prediction method based on XGBoost and RBF neural networks. The combination of the two neural network algorithms into the power prediction model makes up for the lack of learning ability of the single neural network model for the input features of net load prediction, and can greatly improve the generalization ability and prediction accuracy of the model. The experimental results show that the described method improves the net power prediction accuracy of the distribution network and outperforms the comparison model.