A proton exchange membrane fuel cell (PEMFC) inevitably experiences performance degradation during long-term operation under various corrosive conditions. In particular, cell operation under cathode flooding conditions results in severe cell degradation through sequential processes that induce the degradation of each component in the membrane electrode assembly (MEA). Furthermore, an excessive amount of water remaining at the cathode can diffuse into the anode side (back diffusion of water), and repeated instances can cause anode flooding (leading to fuel starvation). Herein, we focus on the collateral anode degradation of a stationary PEMFC operating under cathode flooding conditions with an oversaturated cathode in terms of relative humidity (RH, 150% RH). We suggest low-humidity electrochemical impedance spectroscopy (EIS-RH30) as a fault identification method to detect the degradation of the MEA, including the anode, before severe cell deterioration. Specifically, the EIS-RH30 analysis enables the detection of the malfunctioning anode reaction by limiting the hydrogen oxidation reaction (HOR) and proton conductivity, through which the phase difference between transporting protons from the anode to the cathode can be reflected as the horizontal aspect of the Warburg impedance. A generalized finite-length Warburg is introduced to understand the variation in the Warburg impedance spectra, and its numerical analysis can act as a diagnosable signal to monitor the sequential degradation of an MEA. A feasibility study of EIS-RH30 on actual degradation behavior that occurs under cathode flooding conditions was performed along with a microstructural investigation inside the MEA.

Effects of Oversaturated Cathode Humidity Conditions on the Performance Degradation of PEMFCs and Diagnostic Signals of Warburg Impedance under Low Humidity Conditions

Quantitative diagnosis of PEMFC membrane humidity with a vector-distance based characteristic mapping approach

An improved differential evolution optimization controller for enhancing the performance of PEM fuel cell powered electric vehicle system

To settle the failure detection conundrum of the PEMFC power generation system, a failure detection approach of the PEMFC power generation system ground on local reservation projection (LLP) and learning vector quantization neural network (LVQNN) is devised. In this method, normalized preprocessing and LLP algorithms are used to shorten the dimensionality of the primary data, and the LVQNN is used to identify the pre-processed data. The results of 600 groups of fault data of the PEMFC water management subsystem reveal that the proposed approach can diagnose the faults of the PEMFC power generation system with 98.33% accuracy and can apply to the failure detection field of PEMFC power generation system.

Failure Detection of PEMFC Power Generation System Based on Locality Preserving Projection and Learning Vector Quantization Neural Network

This paper investigates effectiveness of various features in fault diagnosis of polymer electrolyte membrane fuel cell (PEMFC) system, including RMSF (root mean square frequency), ACSD (autocorrelation standard deviation) and kurtosis. Test data is collected from a PEMFC system with various conditions, such as normal operation, flooding and drying out scenarios. By extracting selected features from PEMFC voltage, the performance of various features in isolating PEMFC states is investigated using k-means clustering. Results demonstrate that the combination of RMSF and ACSD could provide reliable fault diagnostic performance. Moreover, kurtosis might be used as a fast diagnostic indicator for various PEMFC degradation mechanisms.

Investigation of Feature Effectiveness in Polymer Electrolyte Membrane Fuel Cell Fault Diagnosis

This paper investigates the empirical mode decomposition (EMD)-based feature selection technique in identifying polymer electrolyte membrane fuel cell water management issue In the study, EMD is applied to fuel cell voltage signals at various states, including fuel cell flooding and membrane drying out, from which intrinsic mode functions (IMFs) are obtained Sensitivity index is used to determine the IMFs more sensitive to fuel cell fault, which are then used for the fault diagnosis The effectiveness of the selected IMFs in discriminating flooding and drying out scenarios is validated using experimental data from a PEM fuel cell system The results show that different faults can be distinguished effectively From the findings, a generalized feature extraction and selection technique can be provided for fault diagnosis in practical PEM fuel cell applications

Fault Classification of Polymer Electrolyte Membrane Fuel Cell System Based on Empirical Mode Decomposition

This paper proposes an image processing-based approach for the fault diagnosis of polymer electrolyte membrane (PEM) fuel cells As more abundant information is contained in the image than that that in 1D signal, features representing PEM fuel cell faults could be better highlighted with the image Experimental data from a PEM fuel cell system at different states, including flooding and dehydration scenarios, is used to validate the proposed method By converting the PEM fuel cell voltage signal into a 2D grey image, several features are extracted from the image, their performance in discriminating different PEM fuel cell states is investigated, and two optimal features are determined for fault diagnosis Moreover, the diagnostic performance of optimal features from grey image is compared with features from PEM fuel cell voltage Results demonstrate that better diagnostic performance could be obtained with the proposed method

Yousif Yahia Ahmed Abuker

Papers

Investigation of Feature Effectiveness in Polymer Electrolyte Membrane Fuel Cell Fault Diagnosis

Fault Classification of Polymer Electrolyte Membrane Fuel Cell System Based on Empirical Mode Decomposition

A Novel Method of PEM Fuel Cell Fault Diagnosis Based on Signal-to-Image Conversion