Regression of the Response Variable of a High Temperature PEMFC–PBI Membrane

Abstract

We measured the output voltage (response variable) of a high temperature polybenzimidazole (PBI)-based proton exchange membrane fuel cell (PEMFC) doped with phosphoric acid under the influence of four factors: the temperature, the pressures (of air and hydrogen), and the current (load). Then, we used a statistical approach to obtain a good regression model to unravel the impact of these factor variables on the response variable. The validation of the models was achieved by using various comparison statistical tests (measures) applied on the experimentally observed data. In most models, the results show a good agreement between hypotheses based on the early obtained experimental observations and this massive data approach. We conclude that there are a sufficiently large number of regression models that accurately predict the behavior of the cell within the range of variables tested and modeled. Our recommendations are based on the proofs of validity of a model and the best value of the coefficient of determination.