The Performance of ACO in the Parameter Identification of a PEM Fuel Cell Model in Comparison to GA and PSO

Renewable energy sources are attracting a lot of attention from researchers and governments around the world. The utilization of renewable and green energy sources in fully or partially supplying our electric power demand, will significantly reduce our dependence on fossil fuels, which would therefore mitigate the harmful effects of those fossil fuels on the environment and the living. Proton Exchange Membrane (PEM) fuel cells are one of those newly erupting green technologies that are attracting a lot of research interest. The parameter recognition of the fuel cell's mathematical model is a crucial part for any further research to be performed on the system, such as analysis, simulation and fault prognosis/diagnosis studies. The analytical approach to recognize the parameters of the model is made challenging because of the complexity of the system. However, metaheuristic techniques such as Ant Colony Optimization (ACO) allows us to identify those parameters with minimal mathematical effort.