An efficient teaching-learning-based optimization algorithm for parameters identification of photovoltaic models: Analysis and validations

Accurate and efficient parameter estimation of the photovoltaic (PV) models is considered a dispensable process to simulate the PV systems. Therefore, many meta-heuristic algorithms have been recently proposed, but the parameters obtained are not as accurate and reliable as is desired, particularly when the PV models have a significant number of unknown parameters. Therefore, in this paper, a modified teaching–learning based optimization (MTLBO) approach is suggested to accurately and reliably extract the unknown parameters of PV models. Our modification to TLBO divides each of the teaching and learning phases into three levels: low, medium, and high according to the scoring level of each learner. The scoring level of each one is measured based on comparison between the fitness of the updated learner and the current leaner; if the fitness of the updated is better, the scoring level is reset to 0, and otherwise, it is increased by 1. Finally, to observe the efficacy of MTLBO, it is investigated on five PV cells and modules: single diode model and double diode model in case of RTC France, Photowatt-PWP201 module, STM6-40/36 module, and STP6-120/36 module. For those PV cells and modules, our proposed could respectively come true the following average outcomes: 0.0009860219, 0.0009825026, 0.0024250749, 0.0017298137, and 0.0166006031. To check the efficacy of MTLBO, it is compared with a number of recent and well-known algorithms. The experimental results show the superiority of the proposed algorithm, especially on double diode model.