Regional Photovoltaic Plant Power Prediction Based on Improved Spatio-temporal Graph Convolutional Network

With the large number of distributed PV connected to the grid, the uncertainty of PV power generation has brought great challenges to the safe and stable operation of the power system, and accurate and reliable regional PV power prediction is of great significance to the improvement of the regional coordinated dispatch capability of the power system. In this paper, we propose a power prediction model for regional photovoltaic power plants based on the Improved Time Series Dense Encoder and Graph Convolutional Network (ITDE-GCN) due to spatio-temporal correlation between regional photovoltaic power plants. The model extracts the temporal correlation of regional PV plants through an Improved Temporal Dense Encoder network, and utilizes a Graph Convolutional Network to extract the spatial correlation among regional PV plants. Finally, the validity of the model is verified by testing it with real PV datasets, which provides higher prediction accuracy in power prediction of regional PV compared to other novel prediction models.