基于主成分分析的蓄电池健康状态识别  

State of Health Identification of Battery Based on Principal Component Analysis

作　　者：周芝远 李富强 汪保[2] ZHOU Zhiyuan;LI Fuqiang;WANG Bao(Ningbo Power Supply Company of National Network,Ningbo,Zhejiang,315012,China;Ningbo University of Technology,Ningbo,Zhejiang,315211,China)

机构地区：[1]国网宁波供电公司,浙江宁波315012 [2]宁波工程学院,浙江宁波315211

出　　处：《宁波工程学院学报》2018年第3期21-25,共5页Journal of Ningbo University of Technology

摘　　要：为识别蓄电池的健康状态,采用主成分分析的方法对615个蓄电池进行综合评价。通过Lilliefors检验,可知该组蓄电池综合得分以95%的置信度服从均值为-0.0524、标准差为0.6988的正态分布;通过与已知健康状态的电池对比发现,分数在[-2.026,1.3722]为健康电池,其他为非健康电池。In order to identify the health status of battery, the method of principal component analysis（PCA） is used to calculate the comprehensive evaluation score of 615 batteries. Lilliefors test showed that the comprehensive score obeys normal distribution with mean of-0.0524 and standard deviation of 0.6988 at 95% confidence level. By comparing with batteries of known health status, it is found that the scores are [-2.026,1.3722] for healthy batteries and others for non-healthy batteries.

关 键 词：蓄电池故障识别 主成分分析 正态分布 Lilliefors检验 

分 类 号：TM912.1[电气工程—电力电子与电力传动]