Automated Analysis of EIS curves for PEM Fuel Cells using Dynamic Time Warping
| Autor: | A. Picot, R. Stephan, J. Regnier, M. Scohy, C. Turpin, O. Crassous, O. Abassie, M. Durand, C. Andrieux |
|---|---|
| Přispěvatelé: | LAboratoire PLasma et Conversion d'Energie (LAPLACE), Université Toulouse III - Paul Sabatier (UT3), Université Fédérale Toulouse Midi-Pyrénées-Université Fédérale Toulouse Midi-Pyrénées-Centre National de la Recherche Scientifique (CNRS)-Institut National Polytechnique (Toulouse) (Toulouse INP), Université Fédérale Toulouse Midi-Pyrénées, Safran Power Units, Safran Helicopter Engines |
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
Aeronautical
Dynamic time warping Materials science Electrochemical Impedance spectroscopy [SPI.NRJ]Engineering Sciences [physics]/Electric power Proton exchange membrane fuel cell Automated Analysis 7. Clean energy Automotive engineering PEM Fuel Cell Start-up and shut-down [CHIM.OTHE]Chemical Sciences/Other Dynamic Time Warping Hydrogen |
| Zdroj: | 2021 IEEE 13th International Symposium on Diagnostics for Electrical Machines, Power Electronics and Drives (SDEMPED) 2021 IEEE 13th International Symposium on Diagnostics for Electrical Machines, Power Electronics and Drives (SDEMPED), Aug 2021, Dallas, TX, United States. pp.303-309, ⟨10.1109/SDEMPED51010.2021.9605563⟩ |
| Popis: | International audience; Fuel cell (FC) is a promising solution in order to tackle global warming problems. Though, efforts are needed for the development of reliable tools to monitor the FC state of health and to extract useful information in order to detect possible malfunctioning. The present paper propose an original method based on the Dynamic Time Warping (DTW) technique in order to process and analyze electrochemical impedance spectroscopy data. The proposed method extracts information on the similarities between 2 EIS curves. This method is evaluated on data from start-up and shutdown experimental campaign on a high temperature PEM-FC stack. Several hundreds of EIS curves are processed over 5 different conditions. The proposed method reaches 92% of correct unsupervised classifications. From the different classes identified, the ohmic resistance is extracted in order to study the impact of 2 different start-up and shutdown protocols on the FC performance. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|