Parameter optimization of polymer electrolyte membrane fuel cell using moment-based uncertainty evaluation technique

Finding the optimal operational parameters of polymer electrolyte membrane (PEM) fuel cells is very critical for improving its efficiency. The incorporation of uncertainties in the operational parameters during the optimization process is very important to ensure that the fuel cell can operate reliably in real-life deployment. Unfortunately, no attention has been given thus far on studying the effects of parameter uncertainties on the fuel cell performance. In this paper, the output power and hydrogen flow rate are being optimized under uncertain conditions of the stack current, stack temperature, oxygen excess ratio, hydrogen excess ratio and inlet air humidity of the PEM fuel cell. The effective application of a new analytical moment-based uncertainty evaluation technique proposed in combination with response surface methodology provides an accurate and numerically efficient evaluation of the fuel cell performance. The paper has demonstrated that the optimal mean output power under uncertainty is 1329.56 W and its corresponding operating point is different from that of the optimal output power 1545.25 W calculated assuming all parameters are fixed. The results show that, for the selected PEM fuel cell model, meeting probabilistic constraints makes the fuel cell less susceptible to input variations, but this can only be achieved at the expense of optimal output power. The useful framework, analyses and discussions presented in the paper can be adapted into any fuel cell performance evaluation and design optimization.