Photovoltaic model parameters identification using Northern Goshawk Optimization algorithm

The massive integration of photovoltaic (PV) systems into electric power grids creates a slew of new issues in today's power systems. Accurate modeling of photovoltaic modules is critical in strengthening the characteristics of its systems in simulation assessments. Modeling such PV systems is represented by a nonlinear current-voltage characteristic curve behavior with numerous unknown parameters due to insufficient data in the cells' datasheet. This manuscript presents an application of a recently introduced optimization algorithm called Northern Goshawk Optimization (NGO) for parameter identification of the triple diode model of the PV module. Three commercial PV modules are utilized in this study to accomplish this task. These models are multi-crystalline structures like Photowatt-PWP201 and Kyocera KC200GT and mono-crystalline like Canadian Solar CS6K-280 M. The simulation results show the ability of the NGO to extract the model parameters accurately. Experimental validation of the estimated parameters using the NGO optimizer is accomplished and compared with the simulation results under various environmental conditions. The simulation results show the superiority of the NGO over competitive optimization algorithms in terms of convergence speed and accuracy. The NGO can reduce the cost function to 1.35E-05, 9.42E-05, and 0.000195 for the PWP-201, KC200GT, and Canadian Solar CS6K-M modules. Moreover, the robustness of the NGO is evaluated by the statistical analysis and the Wilcoxon rank test.