This study focuses on the development and application of new methods for optimal experimental design to efficiently characterize the aging behavior of lithium-ion batteries (LIBs). Motivated by the need for precise aging models for the design and operational optimization of battery storage systems, the work aims to obtain maximum information about model parameters with minimal experimental effort.

The questions on how to improve the precision of parameter estimation while simultaneously reducing experimental workload is addressed. To achieve this, the mathematical foundations of parametric models, parameter estimators, and experimental design methods are analyzed. Building on this, a new approach called parameter-individual optimal experimental design (pi-OED) is developed, enabling targeted optimization of experimental plans to enhance the estimation accuracy of selected parameters.

In a further step, pi-OED is integrated into a framework for multi-stage optimal experimental design (MS-OED). By combining space-filling sampling methods in early experimental phases with pi-OED in later stages, experiments can be efficiently designed even with limited prior knowledge. Simulation studies demonstrate the potential of MS-OED to reduce experimental duration while simultaneously improving model quality.

The practical applicability of the developed methods is demonstrated through an experimental aging study on commercial LIB cells. In a two-stage experimental plan, pi-OED is compared with conventional designs. The results show that pi-OED significantly reduces the uncertainty of critical parameters in calendar aging. For the more complex cyclic aging, limitations are identified that provide starting points for future work.

Overall, this work makes an important contribution to the advancement of statistical experimental design in battery research. The developed methods enable a more efficient characterization of aging behavior and thus support the development of long-lasting and reliable battery storage systems.