Hybrid Optimization Algorithms for Accurate Extraction of InGaN Solar Cell Parameters

"Accurate extraction of model parameters for Indium Gallium Nitride (InGaN) solar cells is essential for enhancing device efficiency and enabling reliable simulation under diverse operational conditions. This paper presents a comparative study of hybrid optimization techniques that integrate the Levenberg-Marquardt Algorithm (LMA) with four popular metaheuristic algorithms: Bat Algorithm (BA), Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and Differential Evolution (DE). The one-diode model is employed to characterize the solar cell behaviour, involving five critical parameters: photocurrent density (Jph), reverse saturation current density (Js), series resistance (Rs), shunt resistance (Rsh), and the diode ideality factor (A). Parameter extraction is performed using experimental current-voltage (J-V) data under Air Mass Zero (AM0) illumination, which simulates the extraterrestrial solar spectrum and is relevant for space-based photovoltaic applications. Each hybrid method begins with a global search performed by a metaheuristic algorithm, followed by local refinement using LMA to enhance accuracy and convergence. The experimental data serves as the benchmark to evaluate the performance of these methods in terms of accuracy and computational efficiency. Results indicate that the hybrid algorithms outperform their standalone counterparts by achieving lower root mean square error (RMSE) values. Among the compared approaches, the BA-LMA hybrid exhibited the best overall performance, achieving superior convergence reliability and computational speed. This makes it a promising candidate for real-time solar cell modelling and parameter estimation tasks, particularly in high-precision applications such as aerospace and satellite systems. This study underscores the effectiveness of hybrid optimization frameworks in extracting solar cell parameters with high accuracy and efficiency. The integration of deterministic and stochastic techniques offers a robust solution for nonlinear optimization problems commonly encountered in photovoltaic modelling."