Efficient Estimation of Nyquist Diagrams for Smart Health Monitoring of Electric Vehicle Batteries Using Sparse EIS Measurements

Electrochemical Impedance Spectroscopy (EIS) is a powerful diagnostic tool for assessing the internal state and degradation mechanisms of lithium-ion batteries. The Nyquist diagram extracted from EIS measurements provides key information on ohmic resistance, charge-transfer processes, SEI layer evolution, and ion diffusion. However, in automotive applications, performing full EIS acquisitions at every charge-discharge cycle is impractical due to the long measurement time and the significant impact on system availability. This work presents a novel and efficient methodology for estimating complete Nyquist curves in cycles where EIS measurements are not feasible, starting from a limited number of reference EIS acquisitions. The proposed approach extracts a set of meaningful features from the Nyquist spectra of the sparse measured cycles, including the width of the intermediate arc, high- and low-frequency slopes, SEI-related curvature, and flex points. In detail, the presented algorithm interpolates the temporal evolution of the estimated features to reconstruct the missing Nyquist diagrams. Moreover, the reconstruction combines empirical modeling of different frequency regions with Bezier-based curve connections, ensuring physical coherence across the spectrum. Experimental validations on six Li-Ion battery (LiB) cells over 400 aging cycles demonstrate high accuracy, with reduced deviations between estimated and measured EIS curves. The method has been proved to be robust across multiple cells and aging stages, enabling practical integration of EIS-based State-of-Health diagnostics into real life online applications by decreasing measurement time.