Prediction of Solar Photovoltaic Power Generation Based on MLP and LSTM neural networks

In recent years, with the increasing of photovoltaic (PV) penetration, it is very important to improve the fore-casting accuracy of the PV generation for the real-time dispatching operation of distribution network. In this paper, the multilayer perceptron (MLP) and long short-term memory (LSTM) neural networks are used to predict and analyze the electricity consumption data in Shanghai, Jiangsu, and Zhejiang. In the mean while the accuracy of the prediction result is improved by the enhanced algorithm during control experiments. It has been proved that the algorithm proposed in this paper is an effective and reliable method for PV power generation prediction. The comparison of prediction result shows that the root mean square error (RMSE) of LSTM model for PV power generation is slightly more accurate than that of MLP model.