Research on Power Prediction of Distributed Photovoltaic Power Stations Based on TCN-GCN

In recent years, with the rapid growth of distributed photovoltaic installed capacity, the safety and economy of distribution networks are facing increasing challenges. The existing methods for predicting photovoltaic power generation typically rely on fine physical modeling or high-quality meteorological data, but this is often difficult to achieve in distributed photovoltaic systems. Therefore, this article proposes a ultra short term photovoltaic power generation prediction method based on TCN-GCN hybrid neural network, which can effectively extract spatiotemporal features from historical power generation data. This article uses the publicly available dataset provided by the National Renewable Energy Laboratory (NREL) to train the model, and compares the predicted results with four mainstream neural network models. The experimental results show that the TCN-GCN network performs excellently in prediction accuracy, with significantly reduced absolute mean error and square root error, proving that this method has important application value in the scheduling and management of distribution networks with distributed photovoltaic access.