Research on photovoltaic power generation power prediction based on LSTM

The photovoltaic (PV) power generation system, characterized by uniform peak distribution and high response rate, has been extensive used in various fields and has become one of the important power generation equipment. Firstly, this paper takes LSTM neural network as a prototype and combines with Adam algorithm to construct a PV power prediction model. Secondly, $\mathbf{9 0 \%}$ of the data is extracted from the data set as the training set to train the model, and the remaining $\mathbf{1 0 \%}$ is used as the test set. The results show that the mean absolute error (MAE) of the training set and the test set is 79.01 and 17.41, and the root mean square error (RMSE) is $\mathbf{1 3 1. 0 7}$ and 32.56. Finally, the model proposed in this paper is used to predict the future photovoltaic power generation, and the validation concludes that the prediction results are in accordance with the law, which indicates that the prediction results of the model proposed in this paper are real, and it has certain guiding significance in optimizing the resource allocation of the electric power system and planning the proportion of different generation types.