基于改进K-means算法的退役动力电池一致性分析

Consistency Analysis of Retired Power Battery Based on Improved K-means Algorithm

针对现有退役动力电池聚类筛选方法计算量大、效率低等问题,本文提出了一种基于轮廓系数的K-means聚类算法并对退役动力电池进行聚类筛选,并以电压标准差作为评价指标对电池成组后的一致性进行评价。通过对电池进行HPPC实验,先提取电池容量、欧姆内阻、电压作为电池特征参数,再将特征参数归一化处理,同时在电池筛选成组过程中通过计算轮廓系数来评价聚类效果以确定k值。结果表明,该方法降低了电池筛选过程计算量,并且能够快速有效地将退役动力电池进行聚类重组,同时重组后的电池组一致性较好,进一步提高了退役动力电池测试装置的工作效率。

Aiming at the problems of large computation and low efficiency of existing clustering screening methods for retired power batteries,this paper proposes a K-means clustering algorithm based on silhouette coefficient to cluster and screen retired power batteries,and uses the standard deviation of voltage as an evaluation index to evaluate the consistency of grouped batteries.Through HPPC experiment,the battery capacity,ohmic resistance and voltage are extracted as the character-istic parameters of the battery,and then the characteristic parameters are normalized.The k value is determined by calculating the silhouette coefficient to evaluate the clustering effect in the process of battery screening and grouping.The experimental re-sults show that this method reduces the calculation amount in the battery screening process.It can quickly and effectively clus-ter and reorganize the retired power batteries,and the consistency of the reorganized battery packs is good,further improving the working efficiency of the retired power battery testing device.