Prediction of photovoltaic power output for microgrids using back-propagation neural network

The main purpose of this study is to predict photovoltaic (PV) power output using back-propagation neural network (BPNN). The mean relative error is used to evaluate the prediction accuracy. This study used the historical data of temperature, humidity, and PV power output to predict the day-ahead hourly power output. Eventually, the day-ahead hourly prediction results are applied to the energy management system of the microgrids (MGs) of the remote island of Taiwan. Results demonstrate that the proposed BPNN algorithm for prediction is accurate and effective. Outcomes are also beneficial for the day-ahead unit commitment of MGs.