Series-connected lithium-ion battery pack health modeling with cell inconsistency evaluation

The lithium-ion battery has become the prevalent technology to store and serve electric power. The state of health (SOH) for a battery cell directly influences the working safety and reliability of the host system. Moreover, since the battery cells are series connected for higher terminal voltage, the cell inconsistency will also impact the performance of the whole battery pack. In order to quantitatively evaluate the inconsistency of lithium-ion cells and represent the battery health state, this paper conducted a numerical study on inconsistency analysis and proposed a novel degradation feature for lithium-ion battery pack health state modeling. The charge cut-off voltage of each lithium-ion battery cell is utilized for battery cell inconsistency analysis according to the practical operating conditions. The normality test results prove that the inconsistency among lithium-ion battery cells is normal distributed. Furthermore, the correlation between the normal distribution parameters and battery pack performance degradation features are then quantitatively analyzed. The Spearman's rank correlation coefficient is -0.88, which means the scale parameter of the normal distribution is negatively related with the battery pack performance degradation. Moreover, the experimental results also prove that with the battery pack performance degrades, the inconsistency among the battery cells is getting larger. The scale parameter in normal distribution is not only capable for cell inconsistency representation but also promising for series connected lithium-ion battery pack degradation evaluation.