Short-Term PV Generation System Forecasting Model without Solar Radiation Based on Improved Wavelet Neural Network

Abstract

The PV generation system is an uncontrollable source and has an affect on the grid by randomness of output power. So we need to strengthen the study of PV output power forecast. In this paper, according to the related information of recent day with the same weather type, an improved wavelet neural network forecasting model without solar radiation was proposed. Furthermore, with the measured data came from a PV power plant, comparison experiments were made as opposed to improved wavelet neural network forecasting model and the wavelet neural network forecasting model with traditional learning algorithm. The experimental results indicate that the improved wavelet neural network forecasting model can significantly improve the precision of PV output power prediction. The comparison experiments considering solar radiation were also given, which also show the high precision and high efficiency of proposed model and algorithm.