Mechanistic calendar aging models provide physically interpretable insights into battery degradation but require extensive experimental effort and often suffer from poor parameter identifiability. This work combines model reduction with optimal experimental design (OED) to enable efficient and robust parametrization.

A structured workflow is applied, including identifiability analysis, targeted model reduction, initial parameter estimation, OED-based experiment selection, and validation. Model reduction improves numerical conditioning by removing weakly identifiable parameters, enabling stable OED application.

Using D-optimal design, the experimental space is reduced to a small number of highly informative conditions. Parameterization of degradation modes can be achieved with only 2–3 targeted experiments, and initial parameter estimation can be replaced by parameter transfer.

Validation on two lithium-ion cells shows that measurement effort is reduced by up to 81.1% while maintaining comparable model quality compared to full calibration on 22 conditions.