Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering

First Advisor

Hassan Noura

Second Advisor

Hussain Shareef

Third Advisor

M. Hashem Nehrir


The world's awareness towards the amount of destruction that our extensive and ignorant lifestyles in the past few decades have imposed on the environment is growing day after day. This resulted in an increased governmental and research interest towards the development and use of green technology. Fuel Cells are one of the green technologies that received a major share of research interest in the past decade.

However, despite their promising features, Fuel Cell systems still lack a solid fault diagnosis and predictive maintenance study. There are numerous faults that have to be detected and diagnosed on a fuel cell power generator system, ranging from chemical faults, to electrical and power electronics faults such as: reactant leakages inside the Fuel Cell, Fuel Cell flooding and membrane drying out, membrane humidification and reactive gas feeding, the accumulation of nitrogen and/or water in the anode compartment, etc.

The aim of this dissertation work is to develop and implement a model based fault diagnostic scheme for a commercial Proton Exchange Membrane Fuel Cell (PEMFC) system in order to improve its safety and reliability; despite the lack of important system information. To achieve this aim, a diagnosis-oriented model of a fuel cell power generator is developed and validated using actual experimental data. Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) were both utilized in the parameter identification of two commercial PEMFC systems' models. GA and PSO were used to extract the correct parameter values of the fuel cell system to minimize differences between experimental and simulated results. Furthermore, PSO was found to outperform GA in the identification process. The effect of severe environmental conditions of hot climate countries such as the UAE on a commercial PEMFC system is then studied and analyzed in simulation using MATLAB/SIMULINK through the developed dynamic system model. The thermal analysis results suggested that the fuel cell system under study would fail to start properly at ambient temperatures of 40°C and higher.

Moreover, the faults that may affect the PEM fuel cell system are listed and their severity is analyzed. In a next step, a more comprehensive system model is developed and validated in LMS AMESim software using actual experimental data, and an appropriate fault diagnostic technique using residual generation is developed, tested and validated in LMS AMESim in order to detect and identify five potential abrupt faults, namely: drying of the membrane, flooding of the membrane, air leakage, hydrogen leakage in the supply manifold and cooling system failure.

The use of LMS AMESim in the proposed modeling and fault diagnosis approach of this dissertation makes it possible to develop a fault diagnosis oriented model for any commercial PEMFC system despite the lack of crucial system information that are usually considered essential for modeling PEMFCs.

Included in

Engineering Commons