Ensemble of different improvements in differential evolution for parameter extraction of PEMFC model

In order to improve the design of the proton exchange membrane fuel cell (PEMFC) model, in this paper, a modified differential evolution (MDE) method is employed for extracting the unknown parameters of PEMFC model. In MDE, an ensemble of three improvements presented in the differential evolution (DE) literature is implemented. These improvements are: i) opposition-based population initialisation; ii) tournament-based base vector selection; iii) single population structure of DE. To verify the performance of MDE, it is used to solve the parameter extraction problems of PEMFC model. Experimental results indicate that the simulated data of the EPMFC model with the extracted parameters well agrees with the experimental data. In addition, compared with artificial bee colony, the original DE algorithm, and the comprehensive learning particle swarm optimisation, the superiority of MDE is demonstrated.