Hybrid optimization based on the analytical approach and the particle swarm optimization algorithm (Ana-PSO) for the extraction of single and double diode models parameters

• Published in: Energy (Oxford) 2023-11, Vol.283, p.129043, Article 129043
• Main Authors: Choulli, Imade, Elyaqouti, Mustapha, Arjdal, El hanafi, Ben hmamou, Dris, Saadaoui, Driss, Lidaighbi, Souad, Elhammoudy, Abdelfattah, Abazine, Ismail
• Format: Article
• Language: English
• Subjects: Analytical method; Meta-heuristic method; Optimization; Photovoltaic; Single and double diode models; Solar cell/module
• ISSN: 0360-5442
• DOI: 10.1016/j.energy.2023.129043

Mathematical Modelling is one of the approved strategies for simulating the performances of photovoltaic devices. Each proposed model contains several parameters; and the accurate determination of these parameters remains a significant challenge for photovoltaic researchers. To this end, in this article, we propose a hybrid analytical/metaheuristic technique for extracting the parameters of single and double diode models. The proposed method estimates some parameters analytically. The other parameters are determined using the particle swarm optimization (PSO) algorithm. This proposed approach is validated by applying it to different types of PV cells and modules. The quality of the obtained solutions is evaluated by comparison with the results proposed by other optimization algorithms. The minimum RMSE values found (7.8093×10−4 for R.T.C France, 3.1491×10−4 for PVM 752, 2.1547×10−3 for Photo-watt-PWP 201, 1.7920×10−3 for STM6-40/36 and 1.5330×10−2 for STP6-120/36) demonstrate the high reliability of the approach used compared to other approaches and validate the performance of our approach. Furthermore, the analysis of convergence, speed, and robustness reveals the superiority of our proposed algorithm over the standard PSO algorithm. •A new hybrid method combining analytical and meta-heuristic approaches is proposed for parameter estimation of PV models.•The proposed method reduces the dimensions of the meta-heuristic algorithm's search space using an analytical procedure.•The approach utilizes an optimization process that generates more consistent and realistic solutions.•The method has been shown to outperform a set of meta-heuristic approaches regarding accuracy, speed and stability.