Advanced parameter extraction optimization technique for the four-diode model approach

Abstract

• An advanced dynamic inertia particle swarm optimization with velocity clamping (ADIPSO-VC) method is proposed for the Four-diode model (FDM) parameter extraction. • The integration of dynamic inertia, and velocity clamping to traditional PSO fine-tune the search spaces and significantly enhances the computational efficiency of the proposed ADIPSO-VC. • To demonstrate the efficacy of the proposed method it is tested against the other algorithms simulated annealing, and conventional PSO. • Root mean square difference (RMSD), mean absolute error (MAbE), and relative error (RE) have also been employed to validate the comparative analysis. • The proposed ADIPSO-VC approach can be employed for the application of optimization of design parameters for extremely complex and Non-linear parameter spaces. Accurate performance estimation of photovoltaic devices is important in optimizing the efficiency and cost of the photovoltaic systems. The one-diode and two-diode models are used because they concisely represent the current-voltage (I-V) variations. However, these models mainly focus on the fundamental mechanism of diffusion and Shockley-Read-Hall recombination. Four-diode model (FDM) is developed from the standard three-diode model used to enhance the precision of PV system performance estimations. However, the better-detailed FDM offers modeling of other recombination and leakage currents that occur in rather complex solar cells or advanced cells like heterojunction, multijunction, or perovskite ones. This model helps to get a more accurate picture of the PV cell operation as several diodes are added to model recombination processes and defects. This work makes use of sophisticated forms of parameter extraction aimed at promoting the optimization of algorithms such as the one known as Advanced Dynamic Inertia-Particle Swarm Optimization with Velocity Clamping or ADIPSO-VC. For comparison with FDM, a three-diode model (THDM) is utilized, and the outcome of the former is then analyzed against the latter. In addition, as a confirmation of the reliability and repeatability of the results obtained by applying the developed algorithm for parameter extraction, FDM is compared with classical methods. To demonstrate the efficacy of the proposed method it is tested against the other algorithms Simulated annealing, and conventional PSO. Based on the comparison, it is evident that ADIPSO-VC surpasses the other methods by demonstrating lower error rates and shorter computational time.

Topics & Concepts