Modeling and control of a PEM fuel cell performance using Artificial Neural Networks to maximize the real time efficiency

In recent years, the proton exchange membrane (PEM) fuel cell is regarded as the best choice in the next generation automobile power source owing to its high fuel conversion efficiency, low noise, almost zero emissions, and low operating temperature. The working condition of PEM fuel cell depends upon several environmental parameters including the flow rate of fuel and oxidant, cell temperature, catalyst activity, and cell fittings. Mostly the data driven techniques are used to predict the voltage and power losses from a fuel cell in particular time. So instead of using a whole analytical model of fuel cell it is better to use Artificial Neural Network (ANN) model due to some of the parameters are very difficult to measure with respect to time. In this present work, it is investigated to develop a PEM fuel cell model using ANN technique. The experimental test on a real time fuel cell has been carried out to validate the ANN model. The different set of operating data is investigated with changing the environmental parameter. The ANN model is applied to emulate real operating conditions such as temperature, hydrogen consumption. After analysis the results it can be concluded that this presented model have good accuracy. Moreover, ANN learning methodology can be implemented to improve the PEM fuel cell stack efficiency. The model is implemented to determine the I-V performance of a single cell PEM fuel cell at different operating settings. The model could obtain the optimized values for the input variables corresponding to the value of objective function. Results showed a consistency between experimental data and the data made by the model. Therefore, it is indicated that the developed model is an effective method, which can predict the performance of fuel cell with high accuracy.