Advancing EIS-based characterization of PEMFCs through model-based DoE and parameter sensitivity analysisVoglar, Tit (Avtor) Kravos, Andraž (Avtor) Katrašnik, Tomaž (Avtor) PEM fuel cellselectrochemical impedance spectroscopymodel-based design of experimentsparameter sensitivity analysisSoX diagnosticsEfficient parameterization of fuel cell (FC) models is essential for reducing experimental cost and time. Although several design of experiments (DoE) methodologies exist for time-domain models, frequency-domain DoE using electrochemical impedance spectroscopy (EIS) remains largely unexplored. To address this gap, we advance EIS-based PEMFC characterization by combining a physicochemically consistent frequency-domain model with model-based DoE and parameter sensitivity analysis. This integration enables systematic assessment of the identifiability of calibration parameters, including intrinsic current densities, double-layer capacitances, and proton conductivities, as well as the construction of a reduced experimental operating space based on parameter sensitivities. Using the DoE methodology, 20 optimally selected points achieve up to a 28-fold increase in calibration-parameter information compared to the full experimental dataset, thereby reducing the number of required measurement points while retaining or improving parameter identifiability. Moreover, frequency analysis reveals that the dominant sensitivity of model parameters lies above 1 Hz, enabling further measurement time reduction by up to 100-fold without an increase in parameter variance. The resulting enhancement in parameter identifiability is expected to support internal-state observability and future use of the model, together with the proposed model-based DoE methodology, in virtual-sensor-based intra-cell State-of-X diagnostics.20262026-07-15 12:09:18Članek v reviji184816UDK: 621.352.6:543.428ISSN pri članku: 0378-7753DOI: 10.1016/j.jpowsour.2026.240788COBISS_ID: 284897795sl