Battery condition monitoring using spectral centroid of active acoustic emission signals

Condition monitoring of State of Charge (SOC) and State of Health (SOH) is critical for ensuring the safety, reliability, and performance of lithium-ion batteries (LiBs) across a wide range of applications. Acoustic sensing has emerged as a promising technique for characterizing internal battery states in a non-invasive manner. This study introduces a novel acoustic methodology for real-time monitoring of SOC in pouch-type LiBs during charge–discharge cycles, as well as SOH following calendar aging, using spectral centroid (SC) analysis of frequency-swept active acoustic signals. Key contribution includes the development of a hybrid modeling framework integrating finite element (FE) simulations with experimental validation. SOH was evaluated by repeating on the same batteries after 5-month storage under controlled laboratory conditions. The results provide clear evidence of mechano-electrochemical coupling effects resulting from lithium-ion intercalation/deintercalation, solid electrolyte interphase (SEI) growth, and loss of active material (LAM) during both cycling and aging processes. The correlations observed between acoustic features and SOH offer a possible cost-effective and scalable enhancement to existing non-destructive evaluation (NDE) techniques. This research establishes an acoustic-guided wave-based diagnostic framework capable of tracking dynamic changes in material properties, thereby facilitating early failure mode detection and supporting the development of predictive maintenance strategies. • A novel framework with spectral centroid analysis with swept active AE signals. • FE and validation tests identifying material changes linked to LLI and LAM. • A hybrid method bridging insights with aging mechanism and SOC and SOH evaluation. • Resonance peak shifts reveal mechano-electrochemical effects for material tracking.