Dynamic ant colony optimization algorithm for parameter estimation of PEM fuel cell

Abstract

Abstract Proton Exchange Membrane Fuel Cells (PEMFCs) provide a reliable, pollution-free, sustainable, and stable power generating alternative to non-renewable resources, and they do not self-discharge. Proton exchange membrane fuel cells (PEMFCs) necessitate correct parameter estimates for effective investigation, modelling and designing effective fuel cells, highlighting the importance of exact modelling for successful use in many industries. The present research aims to determine the approximate parameters estimation of PEMFC using a modified algorithm derived from the Ant Colony Optimization (ACO) meta-heuristic algorithm. In order to provide justification for the algorithm, it is initially benchmarked against 10 functions. The study compares the outcomes of PEMFC parameter estimation through the Dynamic Ant Colony Optimisation (DACO) algorithm including some additional metaheuristic algorithms such as Ant Colony Optimisation (ACO), Particle Swarm Optimisation (PSO), Artificial Bee Colony (ABC), Differential Evolution (DE) algorithm, and an algorithm known as Grey Wolf Optimisation - Cuckoo Search (GWOCS) which is hybrid in nature. The suggested algorithm’s performance evaluation is based on minimising the Square Error (SSE). The modified proposed optimization algorithm exhibits superior performance compared to other alternative meta-heuristic algorithms due to its minimal SSE value. The effectiveness and efficiency of the modified method based on the Ballard Mark V datasheet were evaluated using statistical error analysis and non-parametric testing. The convergence curves of DACO demonstrate a faster convergence compared to the other optimization algorithms.