Cell-to-cell capacity inconsistency evaluation considering temperature effect of battery pack with cloud data
Limei Wang,Lei Wang,Ying Zhang,Dong Lu,Chaofeng Pan,Zhigang He,Yang Li,Jiejie Sun,Dan Shao,Liangyong Hu,Aihua Wu
DOI: https://doi.org/10.1177/09544070241284065
2024-10-12
Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering
Abstract:Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, Ahead of Print. Due to the initial and dynamic differences of battery cells, cell-to-cell capacity inconsistency exists in a battery pack. Considering the difference between the laboratory data and the operation data, this paper studies the battery pack capacity inconsistency of an electric vehicle based on cloud data. Firstly, the characteristic of different charge modes is analyzed, and the charge segment suitable for Incremental Capacity (IC) calculation is screened. Secondly, the Probability Density Frequency (PDF) method is introduced to obtain a relatively smooth IC curve considering voltage accuracy and sampling frequency. Meanwhile, the accuracy of the IC curve calculated by the low-precision cloud data is verified by high-precision laboratory test data. Based on the characteristic peak height of IC curve, an in-depth analysis of the correlation between capacity inconsistency and temperature difference is carried out. Results show that capacity consistency is affected by temperature difference. Fortunately, the capacity inconsistency affected by the temperature difference is recoverable. Subsequently, the normal distribution of battery capacity is detected, and the results show that the distribution of battery cell capacity is also subjected to the temperature. Finally, a battery cell capacity classification method considering temperature effect is proposed to realize low-capacity battery detection. The proposed method can achieve the detection of low capacity state batteries.
engineering, mechanical,transportation science & technology