A Comparative Analysis of LSTM and BILSTM Network-Based Methods in PV Power Prediction

With the widely deployment of PV power installations, the intermittent operation property comes from PV power cause various impacts to electric grid. It is thus necessary to seek effective prediction technologies in order to well manage the PV power generation. Recently, many deep learning-based methods are introduced in renewable energy predictions due to their outstanding performance. To follow this technical trend, this paper gives a comparative analysis in using LSTM and BILSTM network-based methods for PV power generation prediction. Data collected from a 2.28 MW campus PV site are used for testing and validating of introduced prediction methods. Result shows that BILSTM presents the better performance than traditional LSTM in PV power generation prediction.