基于主成分和聚类分析的区域道路交通安全综合评价  

Comprehensive Evaluation of Regional Road Traffic Safety Based on Principal Component Analysis and Cluster Analysis

作　　者：马社强[1,2] 邵春福[1] 左忠义[1] 马壮林[1] 

机构地区：[1]北京交通大学交通运输学院,北京100044 [2]中国人民公安大学交通管理工程系,北京100038

出　　处：《武汉理工大学学报（交通科学与工程版）》2010年第6期1090-1094,共5页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国家科技支撑计划项目资助(批准号:2007BAK35B06)

摘　　要：对我国省级区域的道路交通安全水平进行综合评价和分类,基于道路交通事故死亡与机动化水平等社会经济因素的相关性,构建了8个评价指标,利用主成分分析法将它们综合为3个主成分变量并评价了我国2006年31个省级区域道路交通安全水平;在此基础上,采用K-均值聚类分析法将31个省级区域的道路交通安全水平分为了3类.结果表明,我国省级区域的道路交通安全水平差异明显,社会经济发达地区的道路交通安全水平高于社会经济欠发达的地区.In order to scientifically evaluate and classify the road-traffic safety-level of the 31 provinces in China,8 feasible evaluation-indexes were put forward based on the correlation between road traffic accident fatalities and regional social-economic factors.3 principal components abstracted from the 8 indexes based on principal-component-analysis were used to evaluate the road-traffic safety-level of the 31 provinces in China.According to the road-traffic safety-level,the 31 provinces were classified into 3 clusters by K-means cluster method.The results show that the differences of road-traffic safety-level between provinces are comparatively obvious,generally the road-traffic safety-level in the social-economic developed provinces are better than in the social-economic less-developed provinces in China.

关 键 词：区域道路交通 交通安全 综合评价 主成分分析 聚类分析 

分 类 号：U491.31[交通运输工程—交通运输规划与管理]