An inverse-problem approach: Estimating multi-physiochemical properties of porous electrodes of single cell of a hydrogen-fueled solid oxide fuel cell by applying the pattern-search optimization to a pseudo-2D numerical model

This paper presents a systematic inverse approach to estimate the electrodes properties of a high-temperature anode supported solid oxide fuel cell (SOFC) by applying the pattern search optimization algorithm to a pseudo-2D model. The paper uses a previously developed simplified pseudo-2D model of the SOFC to solve the forward problem and the root mean square value of the percentage errors in polarization current as its fitness function. Values of anodic reference exchange current density, cathodic reference exchange current density, anode effective porosity, and cathode effective porosity are estimated both separately and simultaneously. A sensitivity analysis of the effects of design variables on the polarization curve of the cell is presented, and the cell polarization curve is shown have low sensitivity to the anodic reference exchange current density. It is demonstrated that the inverse approach can be utilized to effectively estimate the electrodes microstructural properties with either no error, in the case of single-parameter estimation, or within an acceptable error band, for the case of multi-parameter estimation. The results for estimating multiple parameters indicate that the initial guess has significantly more effects on the estimation error of parameters to which the polarization curve has less sensitivity.