Investigation on SOH estimation of EVs high-capacity lithium-ion batteries from EIS health-related frequencies

This study conducted experiments utilizing a high-capacity LiFePO4 lithium-ion battery cell to investigate the relationship between significant state-of-health (SOH) battery impedances and frequencies in electrochemical impedance spectroscopy. The main objective was to identify characteristic frequency points using the impedance spectrum deconvolute method namely Distribution of Relaxation Times (DRT). This approach allows for rapid on-board SOH estimation by focusing on specific frequencies rather than sweeping through entire frequency range. The results revealed characteristic frequencies associated with ohmic resistance, surface-electrolyte interface resistance, charge transfer resistances, and diffusion processes. The SOH estimation model, trained using absolute impedance at characteristic frequencies over cycles, demonstrated a relative error of estimation less than 1% compared to the actual capacity fade, also highlighting the potential of low characteristic frequencies for on-board State of Charge (SOC) estimation.