Remaining Useful Life Estimation of Proton Exchange Membrane Fuel Cells Using Artificial Neural Network

With the recent investigation into the possible replacement of fossil fuel commodities due to their harmful effects on the environment, fuel cells today are being recommended by the research community as the suitable option, particularly in the automotive industry. Proton Exchange Membrane Fuel Cells (PEMFC) are energy converting devices that transform chemical energy into electrical energy with the byproducts of water and heat. With relatively high efficiencies and virtually no greenhouse gas emissions, fuel cells are often ideal for the automotive applications. Many studies and investigations have been carried out to explore ways to improve the overall performance of the fuel cell as well as to minimize the fuel consumption. Despite the progress made by the research community in developing various predictive models in order to address the performance issues, the accuracy of these developed models has lately become active research direction. Due to the complex and nonlinear behavior of the multiple fuel cell variables, the research has focused on modeling using artificial neural network (ANN), which can account for a system of multiple variables with nonlinear and complex behavior to fit a model and estimate the performance at low error and high convergence rate. The current study explored the accuracy of artificial neural network with varying neurons and learning algorithms in predicting the remaining useful life of a PEM fuel cell. Comparing the mean square error (RMSE) and coefficient of determination (R^2) of the models, the developed models yielded the relatively low error values (<0.25) and high R^2 values (>0.9) indicating a higher accuracy. The ANN based model shows promising results in the prediction of the PEMFC‘s dynamic performance and at different operating variables. In addition, the ANN model can be used to estimate the for automotive applications to effectively power efficiency and remaining useful life.