新能源汽车交通事故特征分析——以长沙市为例

An Analysis of Accident Characteristics of New Energy Vehicles:A Case Study in Changsha,China

针对日益突出的新能源汽车交通安全问题,为综合识别新能源汽车交通事故特征,基于长沙市309起新能源汽车交通事故数据,首先从事故严重程度、碰撞特性、车辆及驾驶员特性、环境特性等方面展开描述性分析。根据严重程度,新能源汽车事故被划分为了财产损失、伤人、死亡这3类。结果表明,从严重程度来看,新能源汽车事故中伤亡事故占比高达78.7%,其中死亡事故接近1/5;从碰撞类型来看,新能源汽车事故以碰撞运动车辆(65.0%)和碰撞行人事故(24.3%)为主,其中后者死亡事故比例高达46.7%。此外,新能源汽车事故多发生在城市道路路段,但并未呈现出显著的时间分布特征。在综合考虑人-车-路-环境-碰撞等各种因素后,利用有序逻辑回归模型对事故严重程度进行了建模分析,识别影响事故严重程度的关键因素。结果显示,碰撞类型和时间对事故严重程度有显著影响。与碰撞运动车辆相比,碰撞静止车辆/固定物事故后果更为轻微,而碰撞行人事故后果更为严重;与晚上相比,发生在白天的交通事故后果更为严重,发生在早晚高峰时段的交通事故中伤亡事故比例接近100%。其他因素,包括天气、事故地点、车辆种类和使用性质等,对事故严重程度均未呈现显著影响。

Aiming at the increasingly serious safety issues of new energy vehicles,a descriptive analysis is carried out first to identify the characteristics of the severity,collision type,driver&vehicle features,and environmental features based on 309 traffic accidents of new energy vehicles in Changsha.Here,these accidents are divided into three types by severity,i.e.,property damage only,injury,and fatal accidents.It is found that 78.7%of these accidents are injury-or-fatality ones,with the fatality proportion approaching 20%.In terms of collision types,these accidents are dominated by collisions with moving vehicles(65.0%)and collisions with pedestrians(24.3%),and the fatality proportion of the latter reaches up to 46.7%.Besides,these accidents mostly occur on urban roadway segments,but do not show obvious patterns in time.Then,an ordered Logit model is developed to identify the factors which significantly influence accident severity.The estimated results indicate that both collision type and the time of day show significant effects on accident severity.Compared with striking a moving vehicle,striking a stationary vehicle or object tends to be less severe,whereas pedestrian-vehicle crashes tend to be more severe.Besides,accidents involving new energy vehicles in daytime are found to be more severe than those at nighttime,and for those occurring at morning and evening peak hours,their injury-or-fatality proportion is close to 100%.The rest factors,including weather,location,vehicle type,etc.,do not show significant effects.