基于聚类分析的交通事故成因机理分析

Analysis of traffic accident mechanism based on clustering method

交通事故频发不仅造成重大伤亡和经济损失,而且对社会安定造成一定影响,因此,降低交通事故发生频率对我国经济社会发展具有一定积极作用。以目前我国群死群伤特大交通事故数据为基础,深入分析影响交通事故频发的关键性因素,鉴于交通事故发生的偶然性和随机性,采用统计分析方法,研究我国群死群伤特大交通事故的时间分布、空间分布、事故形态分布的规律。利用聚类分析方法对基础数据进行分类,对同一类的数据样本从人-车-路各个环节进行分析,提取出不同时空分布对交通事故的影响度,并结合相关历史经验分别从主动性和被动性角度提出预防对策。

Frequent traffic accidents can not only cause significant casualties and economic losses, but also a certain influence on social stability. Therefore, reduce the traffic accident frequency has a certain positive role to China＇s economic and social development. Based on numerous casualties traffic accidents data in China, key influence factors of frequent traffic accidents was deeply analyzed. In view of the contingency and randomness of the traffic accident, statistical analysis method was used to study numerous casualties traffic accidents data in China. The laws of time distribution, space distribution and modality distribution accident were analyzed Basic data was classified by clustering method, and data sample of a kind was analyzed from man-vehicle-road link to find out the different effect to traffic accidents by different time and space distribution. At last, some initiative and passive preventive countermeasures were put forward respectively.