Intelligent Prediction of Photovoltaic Power Generation Considering Uncertainty Measurement

Abstract

Background As a renewable energy, solar energy has the advantages of being non-polluting, clean, and renewable. With the variability of solar radiation, the complexity of meteorological factors, and other uncertain changes, how to measure its comprehensive uncertainty is very important. Objective The inherent regularity of the uncertainty behavior of photovoltaic power generation is revealed by the change process of photovoltaic power generation. Methods Using the empirical wavelet transform (EWT), the uncertainty measurement was studied from the perspective of social physics, and an optimized intelligent prediction model (PSOBOA-LSTM) was proposed based on the uncertainty. Results The intelligent prediction model (PSOBOA-LSTM) has a better prediction effect and higher accuracy than other models, and reveals the internal mechanism of the photovoltaic power generation system from the perspective of physics and sociology. Conclusion Using the PSOBOA-LSTM model can better facilitate the power dispatching department to reasonably arrange conventional power generation, coordinate operations and make maintenance arrangements based on the predicted photovoltaic power generation, and solve problems arising from the connection between grid dispatching and photovoltaic power generation forecasting.