PV Generation Prediction Based on EEMD-IWOA-LSTM Model

The power of photovoltaic (PV) is related to many factors, and power prediction is of vital significance to maintain the stable operation of the power system. A prediction model of PV power based on EEMD-IWOA-LSTM model is proposed herein. EEMD decomposition is used to decompose the historical power generation, and various factors such as wind speed and IMF component obtained by plate temperature decomposition are used as prediction dimensions to train the LSTM network. Meanwhile, the parameters are optimized by an improved whale optimization algorithm. We pick the best model. The model comparison results show that the prediction accuracy of power generation is improved by about 20%.