Neural Network Algorithm for Short-term Photovoltaic Power Prediction Accuracy Improvement

One of the methods to improve the accuracy of short-term Photovoltaic(PV) power prediction is the improvement of prediction algorithms. To sort out the application of neural network algorithms in short-term PV power prediction, four algorithms, Back Propagation Neural Network(BPNN), Recurrent Neural Network (RNN), Long Short Term Memory (LSTM), and Bi-directional Long Short-Term Memory (Bi-LSTM), are selected for analysis and validated by practical arithmetic cases in this paper. The results show that the neural network algorithms are well adapted to short-term PV power prediction and have high prediction accuracy. The accuracy of short-term PV power prediction can be improved by improving the neural network prediction algorithm.