at Google ScholarTitle data
Winkler, Luis ; Danzer, Michael A. ; Palm, Herbert:
Optimal Experimental Design for Fuel Cell Model Parameter Identification.
In: IEEE Access.
Vol. 14
(2026)
.
- pp. 59371-59389.
ISSN 2169-3536
DOI: https://doi.org/10.1109/ACCESS.2026.3677883
Abstract in another language
Model-based development provides a systematic approach for investigating fuel cell (FC) and fuel cell system (FCS) operation. It places particular emphasis on understanding reaction kinetics, transport phenomena, and aging mechanisms as functions of operating conditions. Characterizing these processes by physical models requires effective and efficient parameter identification, i.e., the derivation of precise parameter information with minimal experimental effort. Conventional strategies for parameter identification in FC models have so far suffered in particular from high time and resource requirements, leading to considerable bottlenecks in FC model development and system characterization. Therefore, this paper proposes a generic, systematic workflow for equally effective and efficient parameter identification in nonlinear, parametric models. The approach is grounded in structural and practical identifiability analysis based on the Fisher information matrix. It incorporates existing prior knowledge about parameter values, thereby enabling the calculation of optimal experimental designs. The methodology is universally applicable to differentiable parametric models. We demonstrate its practical applicability for a proton exchange membrane FC under realistic measurement conditions using a zero-dimensional model with six unknown parameters. Compared to conventional Latin Hypercube Sampling, the proposed approach achieves parameter identification with a 74 % reduction in uncertainty or, conversely, saves 84 % of the experimental effort required to achieve the same accuracy. The framework enables experimenters to conduct measurement campaigns that are both maximally effective and efficient. All methodological implementations, data, and demonstration results are made publicly available to facilitate reproducibility and community adoption.
Further data
| Item Type: | Article in a journal |
|---|---|
| Refereed: | Yes |
| Keywords: | Parameter estimation; Uncertainty; Noise measurement; Fuel cells; Calibration; Parametric statistics; Noise; Measurement uncertainty; Biological system modeling; Systematics |
| Institutions of the University: | Faculties > Faculty of Engineering Science > Chair Electrical Energy Systems > Chair Electrical Energy Systems - Univ.-Prof. Dr.-Ing. Michael Danzer Research Institutions > Central research institutes > Bayerisches Zentrum für Batterietechnik - BayBatt |
| Result of work at the UBT: | Yes |
| DDC Subjects: | 600 Technology, medicine, applied sciences > 620 Engineering |
| Date Deposited: | 18 May 2026 06:24 |
| Last Modified: | 18 May 2026 06:24 |
| URI: | https://eref.uni-bayreuth.de/id/eprint/97178 |