Extraction of health indicators from electrochemical impedance spectroscopy for state of health estimation of lithium-ion batteries

Abstract

Accurate real-time assessment of the state of health (SOH) of lithium-ion batteries is critical for ensuring their safe operation. Owing to its non-destructive nature, rapid response, and abundant electrochemical information provided, electrochemical impedance spectroscopy (EIS) has become a well-established technique for SOH estimation. Hence, the core challenge is to extract potential health indicators (HIs) from EIS data in order to establish robust SOH mapping models. This review initially introduces SOH definitions and the fundamental principles of EIS; then, it comprehensively surveys the research progress made in EIS-based approaches for HIs extraction, including raw data, equivalent circuit model (ECM), distribution of relaxation times (DRT), and automatic unsupervised identification (AUI) analyses. Crucially, this work summarizes the technical routes connecting HIs extraction methods to SOH estimation and provides the first systematic comparison of AUI and conventional techniques. These approaches leverage advanced empirical models and artificial intelligence to effectively identify and quantify key HIs of performance degradation. Furthermore, the advantages and limitations of these approaches are introduced, analyzed, and compared. Finally, the outlook and challenges for enhancing the SOH estimation are discussed from three perspectives: mechanisms, measurements, and applications. Overall, this review provides a theoretical framework and a technical route for advancing EIS-based SOH estimation, while outlining a future roadmap for non-destructive evaluation technologies, measurement devices, and battery pack-level SOH monitoring.