Solar energy has been receiving tremendous interest in this decade, due to their broad applications in areas of renewable energy sources. To identify the accurate modeling of current and voltage characteristics of solar cells will help to improve the performance of solar energy systems. However, the I-V characteristics are multimodal and complex non-linear equations making these optimization techniques to get trapped into local optima, low efficiency due to an initial solution far from the global optimal solution. Therefore, this paper proposes a modified JAYA (MJA) algorithm for estimating the solar cells parameters. In MJA, the modified solution updating phase by hybrid classical JAYA operator with differential evolution (DE/rand/1) operator, and self-adaptive population size are employed to improve global searching ability. In order to validate the feasibility and performance of the proposed MJA algorithm, experimental studies are carried out in identifying the parameters of a single diode, and double diode models and full comparison with different variants of JAYA, DE and other well-established algorithms in the last two years. Experimental results demonstrate that MJA can obtain better performance in terms of robustness and accuracy than all compared algorithms.
Read full abstract