改进模糊C均值聚类算法及锂电池配组应用  

Improved Fuzzy C-Means Clustering Algorithm and Application of Lithium Battery Grouping

作　　者：付媛[1] 全书海[1] 

机构地区：[1]武汉理工大学自动化学院,湖北武汉430070

出　　处：《河南科技大学学报（自然科学版）》2017年第4期43-48,共6页Journal of Henan University of Science And Technology:Natural Science

基　　金：国家"973"计划基金项目(2013CB632505);国家自然科学基金项目(51477125);湖北省科技支撑计划基金项目(2014BEC074)

摘　　要：在锂电池化成管理的智能配组过程中,当处理大规模数据或锂电池结构较复杂时,速度和准确度不高。因此,提出了一种基于遗传算法与密度加权的改进模糊C均值聚类算法。首先,由遗传算法优化得到初始聚类中心。然后,将样本对象的高斯密度函数作为其权值,并采用Xie-Beni有效性指标改进目标函数。将改进的算法通过标准测试数据集Iris和锂电池配组进行实验验证。验证结果表明:本文算法改善了聚类效果,与模糊C均值聚类算法相比,锂电池配组的正确率提高了0.8%,并且计算迭代次数从14次降低到8次。In the intelligent grouping process of lithium battery formation management,the speed and accuracy of clustering were reduced by the large-scale data or the complex structural relationship. Therefore,an improved fuzzy C-means clustering algorithm based on genetic algorithm and density weighted was proposed. Firstly,the initial cluster centers was optimized by genetic algorithm. Then the Gaussian density function of the sample object was used as weight,and Xie-Beni index was applied to improve the objective function. The algorithm was tested and confirmed through the standard data Iris and lithium battery grouping experiment. The results show that the algorithm in this paper improves the clustering effect. The grouping accuracy of the lithium battery is increased by 0. 8% and the number of iteration times are reduced from 14 to 8.

关 键 词：模糊C均值聚类算法 遗传算法 高斯密度加权 Xie-Beni指标 锂电池配组 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]