Comprehensive Review of Data-Driven Degradation Diagnosis of Lithium-Ion Batteries through Electrochemical and Multi-scale Imaging Analyses

Abstract

Electrochemical degradation diagnoses for evaluating the state of health (SOH) in lithium-ion batteries (LIBs) have been extensively utilized for real-time assessments in electric vehicles (EVs) and for determining the reusability of spent batteries. However, the criteria for the accuracy of these diagnostic methods have not yet been established, highlighting the need to develop methods for validating or cross-checking ones. This review encompasses cutting-edge and innovative diagnostic approaches that incorporate machine learning (ML)-applied analyses to expedite big-data-based electrochemical analyses and enhance their accuracy. Moreover, it introduces emerging non-electrochemical analysis methods, particularly imaging-based degradation diagnosis, which can provide the atomic, particle and electrode level examinations, for assessing the SOH in LIBs. Lastly, this paper provides a comprehensive perspective on the future of rechargeable battery diagnostic fields through the integrated concepts of electrochemical and imaging diagnostics in conjunction with data-driven informatics analyses.