Battery pack SOC estimation by Noise Matrix Self Adjustment-Extended Kalman Filter algorithm based on cloud data

Due to the limited computing capacity of the Battery Management System (BMS), the closed-loop estimation method represented by the Kalman filter algorithm is not really widely used, resulting in estimated errors in State of Charge (SOC). Given the strong computing power and large storage capacity of the cloud platform, a SOC estimation algorithm based on cloud data is proposed in the paper. Firstly, the impact of model parameter timevariability on SOC estimation is researched. Subsequently, the adaptability of the Noise Matrix Self AdjustmentExtended Kalman Filter (NMSA-EKF) algorithm to the long data transmission period and low data transmission accuracy is discussed. Then, the adaptability of the model parameter identification algorithm is also analyzed. Furthermore, the model parameters are calculated by the combination of the direct method and Variable Forgetting Factor Recursive Least Square (VFFRLS) algorithm. Finally, the NMSA-EKF algorithm is used to estimate the SOC of the cloud-based discharging fragments. The results show that SOC estimation based on the NMSA-EKF algorithm has a high accuracy, and the overall relative error is within 3 %. The results also further validate the accuracy of the proposed parameter identification method.