Predictive Power Model of a Photovoltaic System under Partial Shading Conditions Using Neural Networks

Photovoltaic energy has emerged as an extremely attractive alternative for electricity generation, especially with advancements in control systems that facilitate its integration with electrical applications.However, despite careful planning in the placement and fixing of panels, shading is inevitable in various circumstances due to space limitations and obstacles such as clouds, buildings, trees, and snow.This article aims to anticipate the energy production of photovoltaic systems under partial shading conditions.Energy loss in such conditions is addressed using bypass diodes, and artificial neural networks are implemented to predict power output.Three key parameters are trained for this prediction, enabling a more accurate estimation of energy generation under shading conditions.