Research On Consistency Sorting Method Of Retired Power Batteries

After the electric vehicle power battery is retired,the remaining capacity can reach60%-80%.When it is used in groups in the echelon,the performance difference between individual batteries will reduce the service performance and cycle life of the module,and even cause safety problems.This paper takes the retired 18650 lithium battery as the research object,and studies the consistency sorting method of retired batteries,aiming at improving the consistency of retired batteries after grouping.By analyzing the mechanism of lithium battery inconsistency and the measures to improve it,it is found that capacity,voltage,and internal resistance are important factors affecting the consistency of retired batteries.(1)A consistent sorting method for retired lithium batteries considering the dynamic characteristics of the curve was proposed.In view of the problem that a single voltage or energy curve cannot fully describe the inconsistent characteristics of the battery,100 retired lithium batteries were taken as the research object for internal resistance and voltage tests,as well as charge and discharge experiments.The energy difference was used to express the polarization difference of the battery,and six indicators of capacity,energy difference,charging voltage,discharge voltage,charging internal resistance and discharge internal resistance were selected for multi-parameter presorting.Considering voltage deviation and capacity deviation comprehensively,dynamic feature sorting is carried out.The results show that the separation method can effectively improve the inconsistency of batteries.(2)A consistent sorting method for retired lithium batteries considering the characteristics of curve changes was proposed.In order to extract the effective information of battery voltage curve and energy curve data,improve the accuracy of the similarity measurement of retired batteries and the consistency after recombination,the overall distribution characteristics of the European distance measurement curve are used,the quintile method is used to extract the curve morphological characteristics,and the improved Longest Common Subsequence(LCS)algorithm is used to measure the curve morphological differences.The entropy weight method is used to match the two types of characteristics of the voltage and energy curves,the variance weighting method is used to weight the voltage and energy curves,the Kmeans++clustering algorithm is used to cluster the curves of 96 retired batteries.The results show that the sorting method can effectively distinguish the changing characteristics of voltage and energy curves.