Battery modeling using linear and nonli̇near electrochemi̇cal impedance spectroscopy

With the growing adoption of electric vehicles, state estimation (SOC and SOH) in lithium-ion batteries has become a critical research topic. Reliable assessment of battery health is essential for accurate range prediction, operational safety, energy efficiency, and extending battery lifetime. Although various approaches have been proposed for SOC/SOH estimation, the widely used linear electrochemical impedance spectroscopy (EIS) method reveals only the linear components of battery behavior and therefore provides limited information. Since lithium-ion batteries inherently exhibit nonlinear electrochemical characteristics, nonlinear impedance analysis techniques are required to overcome this limitation. In this study, linear (EIS) and nonlinear (NLEIS) impedance methods were comparatively investigated for SOC and SOH estimation of lithium-ion batteries. Within the scope of NLEIS, the sensitivity of second-harmonic components to mechanisms such as SEI layer evolution, nonlinear charge-transfer processes, and diffusion limitations was examined. Aspilsan, Power-xtra, and LG 18650 NMC cells were aged for 1000 cycles under 1 A, 2 A, and 3 A current conditions, and EIS/NLEIS measurements were performed every 100 cycles at various SOC levels ranging from OCV to 4.2 V. SOC and SOH estimation models were developed using first-harmonic parameters and second-harmonic |Z₂| values. The analysis demonstrated that EIS parameters exhibit characteristic responses to both SOC variation and aging. While the error in first-harmonic models increased as aging progressed, the second-harmonic models provided more stable results with lower error across all current levels and were significantly more sensitive to aging effects. Overall, the findings show that the NLEIS approach offers a more accurate, robust, and aging-responsive estimation framework compared to the conventional EIS method.