Modeling a PEMFC by a support vector machine

This paper reports a modeling study of proton exchange membrane fuel cell (PEMFC) performance by using a support vector machine (SVM). A PEMFC is a nonlinear, multi-variable system that is hard to model by conventional methods. As regards the SVM, it has a superior capability for generalization, and this capability is independent on the dimensionality of the input data. These two merits combine to make it a powerful tool to predict how a PEMFC will behave under different operating conditions. Herein a SVM model of a PEMFC system is built, optimized and tested. First, the model is determined with selected experimental data, and then it is used to predict PEMFC performance. It is shown that the model can make the prediction in 10 ms with the squared correlation coefficient as high as 99.7%. Therefore, the proposed black-box SVM PEMFC model applies to the simulation, real-time control and monitoring of a fuel cell's performance.