Application of coRNA-GA based RBF-NN to model proton exchange membrane fuel cells

This paper proposes a co-evolution RNA genetic algorithm (coRNA-GA) based RBF-NN modeling approach of proton exchange membrane fuel cells (PEMFCs). Inspired by the biological RNA, coRNA-GA encodes the chromosomes with RNA nucleotide bases and adopts some RNA operations. Some genetic operators are adopted to maintain the population diversity. Two sub-populations are selected by the different evaluation functions, in which different evolutionary strategies are utilized to balance the exploration and the exploitation. The effectiveness of coRNA-GA is validated by the numerical experiments with some benchmark functions. Furthermore, the coRNA-GA based RBF-NN is applied to solve the nonlinear modeling problem of PEMFCs. The simulation results demonstrate that coRNA-GA based RBF-NN is capable of predicting the stack voltage under different operation conditions with better accuracy.