A Data-Model-Driven Predictive Method for PV Output Power

With the development of global energy structure transformation and the popularization of new energy sources, it has become an inevitable trend for new energy represented by solar energy to gradually replace fossil energy. However, with the integration of a large number of PV (Photovoltaic, PV) distributed power sources into the power system, the stochastic, fluctuating and intermittent of PV power sources pose serious challenges to the safe and stable operation of the power system. Therefore, accurate PV output power prediction is important for the economic and stable operation of power systems. The article proposes a model-data-driven PV power prediction method. Firstly, the desired output power of PV power supply under ideal meteorological conditions is quantitatively analyzed through the PV conversion principle of PV cells. Second, the obtained ideal output power is combined with CNN-LSTM for deep learning algorithm. During the training process, the detected environmental factors such as temperature, humidity, wind speed, and atmospheric pressure are used as the input parameters of the neural network, and the predicted power is used as the model output parameter. This allows fitting the effect of the coupling between various meteorological conditions on the output power of the PV power supply, thus providing a deterministic prediction of the PV power supply under different meteorological conditions at any location at any time. In addition, this article is based on the PV operation dataset of the Energy & Metrology Research Service Platform for simulation, verifying the effectiveness and reliability of the proposed strategy.