Nonlinear Hammerstein model identification of SOFC using improved GEO algorithm

Solid oxide fuel cell (SOFC) is a nonlinear, multi-input and multi-output system that is hard to model by traditional methodologies. For the purpose of dynamic simulation and control, this paper reports a Hammerstein model identification of the SOFC using improved generalized extremal optimization (GEO) algorithm. During the identification, the static nonlinearity of the Hammerstein model is modeled by a two-layer radial basis function neural network (RBFNN) and the linear part is modeled by an autoregressive with exogenous input (ARX) model. GEO is a global search meta-heuristic, as the Genetic Algorithm (GA) and the simulated annealing (SA), but with the a priori advantage of having only one free parameter to adjust. In this article the improved GEO algorithm is adopted to optimize the hidden centers, the radial basis function widths and the weights of the RBFNN, and the structure of the ARX model at the same time. After the ARX model structure is determined, the least squares (LS) algorithm is used to estimate the parameters of the ARX model. Simulation results have illustrated the applicability of the proposed Hammerstein model in modeling the nonlinear dynamic properties of the SOFC. At the same time, the simulation result comparisons between the Hammerstein model and RBFNN model demonstrate that the Hammerstein model is superior to the RBFNN in predicting the nonlinear dynamic properties of the output voltage for the SOFC. Furthermore, based on this Hammerstein model, valid control strategy studies such as predictive control, robust control can be developed.