基于逻辑回归的二轮车-汽车碰撞事故深度分析

In-depth Analysis on Cycle-Vehicle Crash Accident Based on Logistic Regression

以中国交通事故深入研究（CIDAS）数据库的二轮车-汽车碰撞事故为样本案例,运用PC-Crash软件进行事故重建,获得碰撞信息;对事故案例进行深度分析,得出质量比、保险杠高度、保险杠宽度、碰撞速度、碰撞类型与骑乘人员损伤程度等事故特征参数的统计分布;利用逻辑回归分析方法,研究二轮车骑乘人员损伤程度与质量比、保险杠高度、保险杠宽度、碰撞速度和碰撞类型之间的相关性。结果表明：在城市道路环境的碰撞事故中,质量比大于20;汽车的碰撞速度大于40km/h;二轮车速度为20~30km/h;保险杠高度为0.4~0.5m;前端碰撞情况下造成骑乘人员受致命伤的概率最大。此结果可为制定有利于二轮车骑乘人员损伤保护的策略提供参考。

By taking the cycle-vehicle crash accidents from China In-depth Accident Study（ CIDAS） as sample cases and using software PC-Crash to conduct accident reconstruction with the crash information obtained.Then an in-depth analysis is performed on accident cases to get the statistical distribution of accident feature parameters. Finally by using logistic regression analysis,the effects of mass ratio,bumper height and width,collision types and speed on cyclist injury severity are studied. The results show that for cycle-vehicle collision accidents in urban roads,the conditions with mass ratio more than 20,vehicle speed more than 40 km / h,a cycle speed of 20 ~30km / h and a bumper height of 0. 4 ~ 0. 5m,have the maximum probability of fatal injury for cyclists in front-end collisions. The outcome provides references for formulating strategy of injury protection in favor of cyclists.