Rayleigh copula for describing impedance data-with application to condition monitoring of proton exchange membrane fuel cells

Condition deterioration of PEM fuel cells affects their electrochemical impedance characteristic, thus a multitude of condition monitoring approaches are based on electrochemical impedance spectroscopy (EIS). By employing recent results of statistical characterisation of impedance components, it becomes possible to describe the complete impedance characteristic through a multivariate distribution that belongs to the exponential family of distributions. As a result, the complete process of condition monitoring of fuel cells can be performed by quantifying the changes of the multivariate distribution with respect to the nominal one. In the presented approach, the multivariate distribution is built through Rayleigh bivariate copula and D-Vine algorithm. It is shown that the selected copula is the optimal choice for modelling impedance characteristics of PEM fuel cells. The performance of the approach was evaluated and validated on an industrial grade 8.5 kilowatt PEM fuel cell system subjected to the flooding and drying faults.