Proton exchange membrane fuel cell model parameters identification using chaotically based-bonobo optimizer

Proton exchange membrane fuel cells (PEMFCs) have been considered the focus of study for energy conversion in various fields including the automobile sector. Nevertheless, PEMFCs face significant dynamic behavior that causes their properties to vary. Therefore, a precise parameter estimation is required to model the PEMFCs adequately. However, due to the complex and nonlinear nature of PEMFC, its parameter estimation is extremely difficult. This paper introduces a robust and efficient approach named Chaotically based-bonobo optimizer (CBO) for determining the unknown variables of the PEMFC model. The proposed method is an enhanced version of the basic bonobo optimizer (BO) where the chaos maps have been used to tune the BO parameters for boosting the optimizer accuracy and consistency. The CBO is examined with several datasets of different PEMFC (250 W and 500 W stacks) at various pressure and temperature levels. The proposed CBO has been evaluated statistically using Friedman, Wilcoxon signed-rank, and multiple comparison non-parametric tests versus recent state-of-the-art and basic BO. The analyses, fitting the datasets, and convergence curves affirm the significant enhancement that has been achieved via adaptive tuning of BO parameters as the algorithm achieved the highest consistency and accuracy with the fastest convergence speed. The standard deviation (STD) by CBO is in the range of [10-16, 10-18]; meanwhile, the basic BO has STD of [10-3, 10-7]. Moreover, CBO converges to the highest quality solution in less than 200 iterations. The non-parametric test has given a shred of evidence on existing significant difference between the proposed CBO, the BO, and the other state-of-the-arts.