Data Enhancement and Aging Warning Within Cloud Battery Management System

The Cloud Battery Management systems (CBMS) can potentially harness battery big data to enhance the safety and efficiency of battery utilization. However, data loss during transmission between cloud and edge severely compromises the performance of CBMS. To address the challenge of data loss during transmission, a Self-Attention-based Imputation for Time Series (SAITS) model is introduced for the first time to restore missing voltage, current, and state of charge (SOC) data. Additionally, a residual threshold-based aging warning method is proposed, detecting battery degradation by analyzing discrepancies between model predictions and actual measurements. Experimental results show that the proposed method outperforms existing techniques, with RMSE values of 0.0432V, 0.1424A, and 0.02% for a 20% loss rate. The aging warning mechanism effectively detects early signs of battery aging, improving the robustness of CBMS against data loss and aging effects.