Comparison between Different Activation Overvoltage Descriptions for Semiempirical Proton-Exchange Membrane Fuel Cell...

Area of Science:

• Electrochemistry
• Energy Conversion
• Materials Science

Background:

• Proton-exchange membrane fuel cells (PEMFCs) require accurate and simple models for performance prediction.
• Semiempirical models are widely used but can lack detailed physical insights.
• Parameter estimation is crucial for refining fuel cell model accuracy.

Purpose of the Study:

• To compare a standard semiempirical model with an agglomerate model for PEMFCs using parameter estimation.
• To evaluate the physical interpretability and fitting accuracy of each model.
• To assess the models' robustness in representing phenomena like reactant depletion.

Main Methods:

• Nonlinear regression was employed for parameter estimation.
• A parallel genetic algorithm was used to solve the optimization problem.
• Experimental polarization data was used to fit and validate the models.

Main Results:

• The agglomerate model demonstrated superior fits to experimental polarization data compared to the common semiempirical model.
• Parameters from the agglomerate model offered more straightforward physical interpretations.
• The agglomerate model showed greater robustness in modeling reactant depletion effects at high current densities.

Conclusions:

• The agglomerate model, combined with parameter estimation, is a highly effective strategy for developing accurate PEMFC models.
• This approach is particularly beneficial when dealing with limited experimental data.
• The enhanced physical interpretability and robustness make the agglomerate model a recommended tool for fuel cell research.