Data restoration and aging classification warning within cloud-edge battery management system

The cloud-edge battery management system (CEBMS) can potentially enhance the safety of battery utilization by exploiting the value of battery big data. However, the data loss during transmission deteriorates the performance of CEBMS notably. Motivated by this, this paper proposes an innovative method combining the data restoration and aging warning within a CEBMS framework. In particular, a self-attention-based imputation for time series (SAITS) model is exploited, for the first time, to restore the lost voltage, current, and state of charge (SOC) simultaneously. Furthermore, a residual threshold-based aging warning approach is proposed by analyzing the mismatch between the model estimation and practical measurements. A CEBMS framework integrating the proposed methods are explored to improve the robustness of management to the low data quality and battery aging. Experimental results validate the superiority of the proposed method over state of the art in terms of data restoration and battery aging warning. The proposed approaches facilitate the deployment of battery big data platforms for enhanced safety management in the future.