A Forecasting Method of Photovoltaic Power Generation Based on NeuralProphet and BiLSTM

Aiming at the problem that the traditional photovoltaic power generation prediction model cannot fully mine the effective information contained in photovoltaic data, which leads to the low accuracy of photovoltaic power generation prediction, an short-term forecasting method based on NeuralProphet (NP) and bidirectional long short-term memory network (BiLSTM) is proposed. The NP is used to mine the time series characteristics of photovoltaic data and acquire the potential trend part and seasonal part hidden in the data. Then the results are merged with the original data and the three-dimensional data is constructed in a sliding window manner as the input of the BiLSTM model. Finally, the BiLSTM is used to realize the short-term prediction of photovoltaic power generation in the next six hours. The experimental results show that the proposed prediction model can effectively improve the prediction accuracy of short-term photovoltaic power generation compared with traditional machine learning methods.