车辆碰撞事故后乘员伤情影响因素研究

Study of Influencing Factorson Occupant’s Injury After Vehicle Collision

为了能够在车辆碰撞事故发生后预测乘员的伤情,便于车辆事故自动呼救系统向呼救中心提供更多伤员信息,对美国国家公路交通安全管理局事故案例调查数据库中的267例碰撞事故案例进行了统计,记录了事故样本中乘员伤情等级、碰撞方向、碰撞速度变化量和乘员位置等信息。基于Probit回归模型,分析了影响乘员伤情的显著性因素,研究发现,碰撞方向、碰撞速度变化量、乘员年龄、乘员位置和是否佩戴安全带与乘员伤情严重程度显著相关。最后,建立了乘员伤情与各个显著性影响因素的Probit回归模型,检验了模型的拟合优度和预测准确度。结果表明,建立的Probit回归模型在事故后乘员伤情的预测方面具有较高的准确性。车辆事故自动呼救系统可以使用该模型预测乘员伤情,给救援中心提供更多伤情信息。

In order to predict occupant’s injury after vehicle collision based andfacilitate vehicle automatic crash notification system to provide more information about driver for call center,267 vehicle collision cases from Special Crash Investigation of National Highway Traffic Safety Administration were analyzed.Then some important information(occupant’s injury scales,collision directions,change in velocity,occupant’s position,etc.)in the crash case reports was recorded.The factors that influence driver’s injury scales were analyzed based on Probit model and it was found that change in vehicle velocity,collision directions,occupant’s age,occupant’s position and use of safety belt are the main factors influencing occupant’s injury scales.At last,the Probit regression model between occupant’s injury and the main factors was constructed.Theproposed Probit regression was tested on the predictive and goodness-of-fit accuracy.This model is able to be used for automatic crash notification system to predict occupant’s injury and provide more information to medical rescue center.