Research on the Output Prediction Model of Low-Voltage Photovoltaic System

Meteorological factors will change rapidly and irregularly, and these changes are interwoven and influence each other, bringing many uncertainties to the output prediction of low-voltage photovoltaic system. Therefore, this paper focuses on the construction and application of the output prediction model of the low-voltage photovoltaic system. Through the in-depth analysis of the composition, working principle and output characteristics of the low-voltage photovoltaic system, the multi-layer feed-forward neural network method is selected, and the research starts from data collection, preprocessing, neural network structure design and practical application and verification. After the practical application of the low-voltage photovoltaic system in a small industrial park, the model shows a high prediction accuracy, which provides a reliable decision-making basis for the energy management and power grid dispatching of the low-voltage photovoltaic system.