Fault diagnosis can help extend the lifetime of fuel cells and has been widely explored for externally humidified fuel cells. However, fault detection in self-humidified fuel cells is scarce, despite its importance. This paper explores various wavelet-based descriptors for identifying hydration levels in self-humidified stacks. Wavelet-based techniques are nonintrusive and provide concurrent examinations in time and frequency. Thus, they encapsulate health related data that can be used as health monitoring features. The specific wavelet-based techniques used here are (i) impedance obtained through the wavelet-based fast electrochemical impedance spectroscopy (EIS) and (ii) changes in wavelet energy through wavelet and wavelet packet decomposition. Results are compared based on their trends under dehydration con-ditions and the robustness of those trends across different currents and sampling rates. They are then assessed based on their predictive ability using RReliefF (Regression-ReliefF)-a regression based feature ranking tool. Based on the qualitative and quantitative analysis, descriptors are assessed in the context of creating a short-term health monitoring system for self-humidified fuel cells. While a modified version of the wavelet energies provides the best results for differentiating faults, fast EIS opens the possibility for analysis of more complex fault modes and is shown to be more robust.

Hydrogen fuel cells, particularly proton exchange membrane fuel cells (PEMFC), are promising, robust, clean energy sources. However, their high cost and short lifespan under dynamic loads impedes their widespread usage. Accurate and real-time prognostics, especially remaining useful time (RUL) estimation, can help ameliorate the commercial viability of PEMFCs. Data-driven methods are increasingly considered for RUL estimation. This paper looks at two such methods – Gaussian Process Regression (GPR) and Long-Short Term Memory (LSTM) Networks and assesses them in terms of accuracy and suitability for real-time applications when tested against the IEEE PHM Challenge 2014 data set. Gaussian Process Regression is a non-parametric kernel method. LSTM, on the other hand, is a recurrent neural network based architecture that is effective at detecting both long term and short term trends in time series predictions. For the cases investigated here, the results derived using LSTM are more accurate, especially since they effectively capture long term trends. However, GPR assigns a probability to its prediction - a desirable aspect in a real-time setting so that corrective action can be applied appropriately. The paper then proposes the use of a variant of these methods - Gaussian Process-Long Short Term Memory Network (GP-LSTM) as an alternative that combines the higher accuracies of the LSTM method and the probabilistic output from GPR. The results attained using GP-LSTM are close in accuracy to the LSTM results and have a probability associated with them, making them suitable for real-time applications. The effectiveness of GP-LSTM is further proven using a dynamic data set and strategies are suggested to appropriately apply GP-LSTM to real-world scenarios.