Aging mechanism analysis and capacity estimation of lithium - ion battery pack based on electric vehicle charging data.

Due to the incompleteness of charging data, the voltage step caused by fast charging conditions and sampling accuracy of the battery management system, the conventional mechanism model is not applicable to the aging mechanism analysis and capacity estimation of electric vehicle batteries. Therefore, this study applies support vector regression to achieve the actual charging condition equivalence based on the variable operating conditions charging data of electric vehicles. The aging parameters and open circuit voltage reconstruction based on the dual-tank model are applied to obtaining the aging state and the capacity of cells. The capacity of the battery pack is calculated by the pack formation theory. The maximum error of the aging parameters obtained by the multiple stage constant current is 5.572% compared with the 1/20 C (C is the charge/discharge current rate unit) constant current charging of the experimental battery. As to the maximum relative error of cell capacity estimation based on vehicle data is 0.99%, and battery pack capacity estimation is 0.86%. The method proposed in this paper is not only able to quantitatively analyze the dominant factors of battery capacity decay, but also achieves high accuracy capacity estimation of the vehicle battery pack and its individual cells. • Support vector regression model is used to process operating voltage data of vehicle battery. • The effect of current switching on voltage is eliminated in the multi-stage constant current charging process of electric vehicles. • The dominant decay mode of the vehicle battery is analyzed based on the mechanism model. • The dual-tank model is successfully applied to the aging analysis of vehicle batteries. • The capacity of the vehicle battery pack and its cells are accurately estimated. [ABSTRACT FROM AUTHOR]