Internal Short Circuit Fault Diagnosis Method for Lithium-Ion Batteries Based on Characteristic Function and Fréchet Distance

Internal short-circuit (ISC) faults represent a significant safety hazard for lithium-ion batteries (LIBs), potentially leading to thermal runaway and catastrophic safety incidents. Conventional threshold-based detection methods often fail to identify ISC faults at an early stage due to their subtle electrical signatures. To tackle this issue, this work introduces a fault diagnosis approach that integrates a bilateral filter-based denoising strategy, a characteristic arcsine function (FAF) for fault feature amplification, and an improved Fréchet distance algorithm for anomaly detection and localization. The bilateral filter effectively suppresses random noise in voltage signals, improving the quality of data for subsequent processing. FAF enhances weak fault characteristics, enabling early-stage ISC detection. The improved Fréchet distance method enables automatic identification and localization of faulty cells based on pairwise similarity analysis of extracted features. Experimental validation using a multi-cell battery module under various ISC fault conditions demonstrates that the proposed method achieves high accuracy, robustness, and reliability in both early detection and localization of ISC faults. The results confirm the method's potential for real-time battery safety monitoring and early warning applications in electric vehicle power systems.