Multiscale Fusion for Abnormality Detection and Localization of Distributed Parameter Systems

Early internal abnormalities in the distributed parameter systems (DPSs) may develop into uncontrollable thermal failures, causing serious safety incidents. However, traditional first-principle methods heavily depend on governing equations, and existing data-based single-scale methods have insufficient performance under dynamically changing conditions. Based on these considerations, the multiscale information fusion is proposed for internal abnormality detection and localization of DPSs under different scenarios. We introduce the dissimilarity statistic to identify abnormalities for lumped variables, whereas the spatial and temporal statistics are presented for abnormality detection for distributed variables. Through appropriate parameter optimization, these statistic functions are integrated into the comprehensive multiscale detection index, which outperforms traditional single-scale detection methods. The proposed multiscale statistic has good physical interpretability from the system disorder degree. Experiments on the internal short circuit (ISC) of a battery system have demonstrated that our proposed method can swiftly identify ISC abnormalities within 20 s and accurately pinpoint problematic battery cells under different testing currents and fault types. © 2024 IEEE.