Genetic algorithms (GAs) are used to optimize solutions to problems, particularly those that are analytically impossible to solve. As their name suggests, they are inspired by the biological concepts of genetics and evolution. Our work aims to study and model a silicon-based photovoltaic generator (PVG). Among the various models available is that of the diode. Modeling was used to approximate the PVG output (voltage, current) as a function of two inputs: temperature and irradiation. The parameters of our model were identified using a real coding algorithm, with the cumulative square error was used for selection. To test the effectiveness of our model, we carried out simulation tests on the power-voltage (P-V) and current-voltage (I-V) characteristics of a wide range of irradiation and temperature variations. This study demonstrates the effectiveness and accuracy of the proposed approach (GAs) and validates the parameters obtained and used in the single-diode electrical model. The results indicate that the GA technique is a better conventional parameter extraction strategy in terms of convergence. It provides globally optimal solutions.