Title:Accuracy and robust early detection of short-circuit faults in single-cell lithium battery

Abstract:Effective early-stage detection of internal short circuit in lithium-ion batteries is crucial to preventing thermal runaway. This report proposes an effective approach to address this challenging issue, in which the current change, state of charge and resistance are considered simultaneously to depict the voltage differential envelope curve. The envelope naturally utilizes the inherent physical information of the battery and accounts for error interference, providing a high-precision range for battery voltage fluctuations under any operating conditions. This study validates the algorithm using data from 10 fault intervals under dynamic operating condition. The results demonstrate that the algorithm achieves 100% accuracy and responds rapidly, enabling timely detection of early-stage internal short circuit faults in batteries. Compared to signal processing-based and neural network methods, the proposed approach offers significant advantages in both accuracy and practicality, making it highly relevant for the safe application and widespread adoption of lithium-ion batteries.