Electrochemical Impedance Spectroscopy (EIS) is a widly used method for the characterization of electrochemical systems. For batteries, EIS measurement results show a strong correlation concerning the State of Charge (SOC), State of Health (SOH), degree of wetting in the production, temperature, and many more characteristics. Assuming that the Kramers-Kronig relation validates the EIS results, thus being linear and equilibria, different fitting methods can be used to validate and calculate the correlation.
With the above assumptions, the EIS curves, i.e. the impedance represented as magnitude and phase in a Bode diagram or real and imaginary parts in a Nyquist diagram, can be fitted using various approaches. Starting with linear curve fitting, the complexity increases through Laplace transform to radial basis function (RBF) interpolation. The goal is not only to fit a single measurement but also to provide a reasonable interpolation over temperature, SOC, SOH or other external or internal parameters.
This poster gives an overview of possible approaches and names their advantages and disadvantages, thus providing a guide for different applications and adaptation scenarios. In particular, the amount of available data has a significant impact on the applicable methods. Artificial intelligence (AI) used to model EIS data shows promising results but requires a significant amount of data on the one hand and subsequent reliability testing on the other. In online applications, such as in Battery Management System (BMS), other aspects besides reliability play an essential role.