Cyclic semidynamic aging dataset for PEM fuel cell prognostic training and validation

• Grant number: 25/02847-8
• Support Opportunities: Scholarships abroad - Research Internship - Master's degree
• Start date: May 27, 2025
• End date: November 14, 2025
• Field of knowledge: Engineering - Mechanical Engineering - Thermal Engineering
• Principal Investigator: José Alexandre Matelli
• Grantee: João Pedro Ribeiro Rosa
• Supervisor: Chetan Shrikant Kulkarni
• Host Institution: Faculdade de Engenharia (FEG). Universidade Estadual Paulista (UNESP). Campus de Guaratinguetá. Guaratinguetá , SP, Brazil
• Institution abroad: NASA Ames Research Center, United States
• Associated to the scholarship: 24/00161-9 - Degradation thermodynamic model of PEMFC for prognostics and health state evaluation, BP.MS

Abstract

In the context of increased energy demand and increased problems due to carbon emissions, fuel cells emerge as a promising solution. However, its low durability and high cost still prevent its commercialization on a large scale. Degradation of fuel cell performance can be caused by several reasons, such as: thermal and mechanical stress, humidity conditions, variable load profiles and start/stop conditions. Still, this phenomenon is not fully understood. To extend fuel cell lifetime, prognostics and health management (PHM) are gaining attention. The prognosis aims to estimate the remaining useful life of the system (RUL) and degradation trends to assist in performing predictive maintenance. There are two main approaches to prognosis: data-driven methods and model-based methods. While data-driven methods may be more accurate in estimating RUL and are easier to implement, they suffer from overfitting issues and require specialized knowledge to fit state-of-the-art models. Previously, an entropy generation-based framework that showed initially great capability to model fuel cell degradation data was proposed. However, further analysis are still needed. In addition, it was noticed that there are not many datasets available from aging experiments, especially from experiments operating with semidynamic loads and with start/stop conditions. Thus, this research project proposes an aging experiment operating cyclic at different semidynamic loads for model validation and characterization tests for new degradation indicators to be analyzed. The project is expected to contribute to the development of fuel cells and thus promote the use of clean energy technologies.