Research on CNN-LSTM photovoltaic power output prediction method considering gray correlation analysis

With the promotion of the "double carbon" goal, the new energy industry, particularly photovoltaic (PV) power generation technology, has made significant advancements. However, the fluctuation and unpredictability of PV power generation present challenges to grid security. Therefore, precise prediction of PV output is crucial for ensuring grid security and advancing the new energy industry. To enhance the accuracy of PV output prediction, this study utilizes gray correlation analysis to identify the key factors influencing PV output within the existing prediction model. By integrating convolutional neural network and long short-term memory neural network, a CNN-LSTM prediction model is developed to enhance prediction accuracy. The model is validated using PV output data from a PV station, demonstrating a substantial improvement in prediction accuracy and practical value.