Parameter estimation of proton exchange membrane fuel cells using eagle strategy based on JAYA algorithm and Nelder-Mead simplex method. (15th April 2019)

• Record Type: Journal Article
• Authors: Xu, Shuhui; Wang, Yong; Wang, Zhi
• Is Part Of: Energy. Volume 173(2019)
• Page Start: 457
• Page End: 467
• Publication Date: 2019-04-15
• Subjects: Proton exchange membrane fuel cell -- Parameter estimation -- Eagle strategy -- JAYA algorithm -- Nelder–Mead simplex search algorithm; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2019.02.106
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Building an accurate mathematical model is vital for the simulation, control, evaluation, management, and optimization of proton exchange membrane fuel cells (PEMFCs). Usually this work is processed through building a mathematical model based on empirical or semi-empirical equations firstly and then estimating the unknown model parameters using optimization technologies. In this study, a simple two stage eagle strategy based on JAYA algorithm and Nelder-Mead simplex algorithm is proposed for effectively estimating the unknown model parameters of PEMFCs. In the proposed strategy, JAYA algorithm is employed for the coarse global exploration, and Nelder–Mead simplex search algorithm is employed for the intensive local search. The effectiveness of the proposed strategy is verified through estimation experiments with 7 and 9 unknown parameters. Compared with the basic JAYA algorithm and four other newly reported excellent meta-heuristic algorithms including Grey Wolf Optimizer, Grasshopper Optimization Algorithm, Salp Swarm Optimizer, and Multi-Verse Optimizer, the proposed strategy possesses better performance in terms of accuracy, convergence speed, and stability, thus it is a promising approach for parameter estimation of PEMFC. Highlights: A two stage eagle strategy is proposed for estimating the optimal parameters of PEMFC. This strategy is based on JAYA algorithm and Nelder–Mead simplex search algorithm. This strategy is examined using estimation experiments with 7 and 9 unknown parameters. This strategy is compared with JAYA algorithm and four other recently-reported algorithms. The obtained results indicate that this strategy is superior to the others.