Analysis of photovoltaic power station (PPS) modeling using artificial neural network and PVsyst software

Abstract

The possibility of using the method of artificial neural networks to analyze the modes of complex electric power systems with integrated large photovoltaic stations is considered. Based on the correlation analysis, the main factors influencing the energy parameters of photovoltaic power plants were selected and the boundary conditions for the Pearson coefficient were determined. The algorithm of the developed program for calculating the modes of electric power systems using neural networks is described, which makes it possible to more accurately predict generation, taking into account climatic conditions. On the example of calculations of the modes of the South-Western part of the energy system of Uzbekistan, taking into account the change in power flows as the generation of the Navoi photovoltaic plant with a capacity of 100 MW changes, a comparative analysis of the results obtained by calculation with real measurements was carried out.