Photovoltaic Prediction Algorithm Based on Hybrid Artificial Intelligence

With the development of science and technology, the proportion of photovoltaic power generation has increased significantly. The nonlinearity and instability of photovoltaic power generation make it difficult to integrate into the power grid on a large scale. Accurate photovoltaic power generation pre-can make photovoltaic(PV) power generation stably integrated into the grid system. This paper proposes a new Random Forest-Convolutional Neural Network-Long Short Term Memory (RF-CNN-LSTM) hybrid model. This model combines the advantages of CNN model which can effectively extract features and LSTM model which can deal with time series problems well. It makes it possible to make a good prediction for photovoltaic power generation. At the same time, this paper uses mean square error(mse) and mean absolute error(mae) as the criterion of model performance. Experimental results show that the performance of the hybrid model is more accurate.