我国道路交通事故演化趋势分析和原因探究.doc

PAGE PAGE 26 我国道路交通事故演化趋势分析和原因探究 摘要：本文选取1990-2014年期间全国与各省市交通事故指标的时间序列数据，分析近些年来交通事故的变化趋势，探索引发交通事故的影响因素。本文先采取主成分分析法，对交通事故指标进行降维处理。接着，通过分析我国各省市经济发展状况、人口发展状况、道路发展状况与交通事故指标之间的定量关系，确定我国交通事故的大致情况。再构建道路交通安全评价指标体系，利用熵值法确定权重，从而评价各省市的道路交通安全情况，分析各省市之间的差异。最后结合交通法规改革，利用时间序列模型和设置虚拟变量，进行平稳性检验和单位根检验，挖掘道路交通事故的影响因素。根据上述结果，对道路交通安全变化状况和减少交通事故造成的人员、经济损失提供宝贵建议。 关键词：交通事故指标；主成分分析；时间序列分析；道理交通安全评价体系；熵值法 The Analysis of Evolutionary Trends in Road Traffic Accidents and Reason Finding in China Abstract: This paper selects national and provincial and municipal traffic accidents time series data during 1994-2014，in order to analyze Chinese road traffic accidents trend indicators in recent years so as to explore factors which affect these accidents. This article first makes use of the principal component analysis method for dimension reduction. Then, By analyzing the quantitative relationships between the provinces of economic development, population development, road development and traffic accidents, to determine the general situation of traffic accidents in our country. In addition, to establish evaluation index system of road traffic safety, using entropy method to determine the index weight to study road traffic safety situation in various provinces and cities were evaluated, analyzing the differences between the provinces. Finally, with various reforms of traffic laws, using time series model and set dummy variables, processing stationary test and unit root test to conclude the factors which affect the traffic accidents. On this basis, we can make recommendations to improve road traffic safety and reduce personnel accidents caused economic loss. Keywords: Accident indicators; Principal Components Analysis; Time Series Analysis; Road Safety Evaluation System; Entropy method 一、 引言 改革开放三十余年来，国民经济增长、现代化交通工具快速发展，人民交通需求日益增加，给予社会全新的面貌。在创造经济效益的同时，同时也带来了许多问题，如不断的交通拥堵、频繁的交通事故，这些对快速的现代化生活造成了一定的困扰。 各类重大道路交通事故的出现逐渐成为各界关注的热点事件，吸引了大众的眼球。据《国家统计局》官方数据表明，2014 年我国交通事故总共有19.68 万起，5.8 万人死亡，同比于2013年差距较小，与2012