As the number of series connections of battery cells increases, individual cells are operating in different temperature profiles, and the aging patterns of the cells become dissimilar from each other. Thenceforth, individual state-cell-characteristics should be tracked online for higher safety. Although Kalman-filter-based battery state estimation is one of the most popular methods, it is sensitive to the accuracy of the battery model parameters and difficult to be applied to every cell. This work proposes an online cell-by-cell state-of-charge (SOC)/state-of-health (SOH) estimation method to mitigate this limitation. The aging patterns of the individual cells are predicted by introducing a combination of a switch-matrix flying capacitor and electrochemical impedance spectroscopy (EIS) model parameter scanning techniques. Accordingly, the accuracy of the SOC estimation for individual cells is enhanced. The proposed method is verified by a real-time simulation platform, where the SOC and SOH levels of the cells are individually estimated within a 1.24% error.

随着电池单元串联连接数量的增加，各个电池单元在不同的温度曲线下运行，并且电池单元的老化模式变得彼此不同。此后，应在线跟踪单个状态单元的特征以提高安全性。虽然基于卡尔曼滤波器的电池状态估计是最流行的方法之一，但它对电池模型参数的准确性敏感，难以应用于每个电池。这项工作提出了一种在线逐个电池充电状态 (SOC)/健康状态 (SOH) 估计方法来缓解这一限制。通过引入开关矩阵飞跨电容器和电化学阻抗谱 (EIS) 模型参数扫描技术的组合来预测单个电池的老化模式。因此，提高了单个电池的 SOC 估计的准确性。所提出的方法通过实时仿真平台验证，其中电池的 SOC 和 SOH 水平单独估计在 1.24% 的误差内。