Modeling of Photovoltaic Systems and Parameter Extraction Using the Ant Colony Optimization Algorithm
Abstract
Accurate extraction of electrical parameters from photovoltaic (PV) cell models is essential for the simulation and performance optimization of solar energy systems. This study introduces a hybrid algorithm based on Ant Colony Optimization (ACO), enhanced with a local search mechanism and a reduced search-space strategy, for parameter extraction in the single-diode (SDM) and double-diode (DDM) models. Using a publicly available experimental dataset from a commercial silicon PV cell (R.T.C. France, under 1000 W/m² irradiance and 33 °C), the proposed algorithm was implemented and validated. The method exhibited high accuracy, achieving Root Mean Square Error (RMSE) values of 1.930 x 10⁻³ for the SDM and 2.777 x 10⁻³ for the DDM. A comprehensive statistical analysis over 30 independent runs demonstrated that the proposed hybrid ACO algorithm provides strong consistency (standard deviation ≈ 6.0 x 10⁻⁶ for SDM and 9 x 10-6^ for DDM) and rapid convergence for both SDM and DDM models (5 iterations for SDM and 11 iterations for the DDM to achieve a RMSE = 5 x 10-3). Compared with well-established metaheuristics such as Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), the proposed approach proved to be a highly reliable and computationally efficient alternative for PV parameter estimation and system modeling.
Keywords
Optimization, modeling, PV system, parameters, Ant colony optimization