Accurate parameters extraction of photovoltaic models using Lambert W-function collaborated with AI-based Puma optimization method

As the global demand for renewable energy sources increases, reliable and efficient photovoltaic (PV) systems are positioning themselves as one of the most promising candidates among sustainable energy resources. Accurate modelling of photovoltaic generators is essential to enhance their energy conversion efficiency, making this need even more paramount. Precise parameter extraction from PV generator models not only aids in optimizing performance but also plays a crucial role in fault diagnosis and maximum power point tracking. Metaheuristic optimization techniques are essential to address the challenges posed by the nonlinearity of PV cell models that are adversely affected and more complicated by shading, temperature and irradiance variations, and other environmental factors. This paper introduces a novel approach to extract key parameters of double diode model (DDM) of PV units using the recently developed Puma optimizer (PO), which has not been applied previously in this context. A new effort of using an improved version of PO when the Lambert W-function ( P O L a m ) is employed for truthful calculus of PV unit current is investigated. Lambert W-function is employed for the precise solving of the I-V curve in the DDM non-linear equations as an alternative to the iterative Newton–Raphson approach. Extensive simulations are conducted to validate the effectiveness of the PO and the P O L a m , comparing their performance against twelve well most cited and newest metaheuristic optimization techniques. The results obtained for wide-spread RTC france and aSi PV cells and the LSM 20 PV module demonstrate that P O L a m significantly reduces the Root Mean Square Error (RMSE) metric. It achieves the best RMSE of 7.218852E-04, 3.927434E-05 and 1.000744E-03 for RTC Cells, aSi Cells and LSM 20 Module, respectively. The results confirm the superior performance and accuracy of P O L a m compared with the existing approaches, revealing that it provides more reliable parameter estimates and helps to improve energy conversion efficiency in photovoltaic applications. • An innovative PO-based optimizer is proposed to estimate the PV units key parameters. • Lambert W function is included with PO to solve the DDM non-linear I-V characteristic. • Experimental validation for different technologies of PV cells and modules is implemented. • The performance of PO and P O L a m is compared with well-cited optimizers. • Statistical analysis conducted for all used algorithms proved the P O L a m performance.