Progress in estimating the state of health using transfer learning–based electrochemical impedance spectroscopy of lithium-ion batteries

Abstract

With the widespread application of energy storage systems, health monitoring of lithium-ion batteries (LIBs) has become important. Transfer learning (TL) provides new ideas and methods for battery health management and life prediction in the field of battery life prediction. This article spotlights the application of TL in enhancing electrochemical impedance spectroscopy (EIS) for the state of health (SOH) estimation of LIBs. It delineates the pivotal role of TL in addressing data scarcity and domain discrepancies to refine prediction accuracy. The review synthesizes recent advancements in utilizing TL with EIS data, detailing the methodology from experimental data sourcing to feature extraction, accuracy metrics, and performance analysis. It concludes by forecasting potential research directions in leveraging TL for more precise health diagnostics of LIBs and life cycle prediction.