To enhance the operational safety of energy storage batteries under complex conditions, this study proposes a multimodal data-driven abnormal condition identification method based on EAA-Informer.First, an analysis of thermal experimental data under various abnormal conditions reveals that temperature changes during the early stages are often subtle. As a result, relying on a single temperature parameter is insufficient to accurately reflect the battery’s state. This necessitates the integration of multiple parameters into a joint modeling framework to improve identification accuracy.Subsequently, a multimodal temporal feature extraction network based on LSTM-MFM is constructed. This network enables deep modeling of multi-source data—such as voltage, current, temperature, and state of charge (SOC)—and facilitates cross-modal correlation learning.To further enhance the model’s ability to capture both critical time steps and long-range temporal dependencies, an Enhanced Attention Allocation (EAA) mechanism is introduced for feature optimization. This is followed by an improved Informer network to perform temporal modeling and abnormal condition identification.Experimental results demonstrate that the proposed method achieves high identification accuracy and low false positive/negative rates across various types of abnormal operating conditions, confirming its potential for application in the safety monitoring of energy storage systems. • A novel multimodal fusion method integrates voltage, current, temperature, and SOC data to capture subtle battery behaviors. • LSTM-MFM with Enhanced Additive Attention optimizes temporal feature extraction, improving anomaly detection accuracy. • An improved Informer with modality-aware encoding achieves robust, efficient abnormal condition identification with low false positive/negative rates.
Sun et al. (Sun,) studied this question.