山区双车道公路摩托车事故黑点及致因集成分析

Black Spot and Cause Integrated Analysis of Motorcycle Accidents on Two-lane Highways in Mountainous Areas

为治理山区双车道公路突出的摩托车事故问题,结合山区公路特性提出了摩托车事故黑点鉴别及致因分析集成方法.基于移动步长法比较事故密度选取最佳移动步长,采用累计频率法对黑点进行初选,再将研究路段按线形划分为直线路段、纵坡路段、平曲线路段和平纵组合四类,利用正态分布曲线模型计算各类线形路段黑点阈值进行黑点精确识别.利用Logistic模型对黑点进行事故致因分析,获取摩托车事故的主要诱因.以典型山区双车道公路云南省元双公路为例,进行实例分析.结果表明:元双公路摩托车事故初选黑点12段,共3.2km,精确识别黑点8段,共2.8km;曲线路段黑点的危险性整体高于直线路段黑点,有坡路段黑点危险性整体高于无坡路段黑点,直线路段、纵坡路段、平曲线路段和平纵组合路段事故黑点的判别阈值分别为60.8、68、62、69.3;驾驶员年龄、事故时段和事故形态与各类线形路段的摩托车黑点事故相关性较高,纵坡路段事故黑点的主要诱因为黑夜、单车参与事故;平曲线路段和平纵组合路段事故黑点的主要诱因为黑夜、单车参与事故、低龄驾驶员.

In order to deal with the frequent occurrence of motorcycle accidents on mountain two-lane highways,the method of black spot identification and cause analysis of motorcycle accident was proposed based on the characteristics of mountain highways.Firstly,based on the moving step method,the authors selected the optimal moving step by comparing the accident density,and used the cumulative frequency method to make a preliminary selection of the black spots.Secondly,the road sections were divided into four types:straight line sections,longitudinal slope sections,flat curve sections,and vertical and horizontal combinations.The black point thresholds of various linear sections were calculated by the normal distribution curve model,and the black points were accurately identified by the thresholds.Then,based on the Logistic model,the cause of the accident was analyzed for accurately identified accident black points to obtain the main accident-induced factors.Finally,Yunnan Yuanshuang Road,a typical mountain two-lane highway,was taken as an example to analyze the case.The results show that among the 12 motorcycle accident black spots in Yuanshuang Road with a total length of 3.2km,8 black spots with a total length of 2.8 km accurately were identified.The black spots on curved sections are generally more dangerous than the ones on straight sections,and the black spots on sloped sections are generally more dangerous than the ones on non-slope sections.Thresholds for discriminating black spots on straight line sections,longitudinal slope sections,flat curve sections,and combination sections are 60.8,68,62,and 69.3,respectively.Three variables—drivers’age,accidentperiods and accident patterns—are highly correlated with black spot accidents on various linear sections.The three variables of driver's age,accident time and accident form have a high correlation with black spot accidents of motorcycles in various linear sections.The main inducing factors of black spots in longitudinal section are night and single vehicle driving.The main inducing factors of black spots in flat curve section and horizontal vertical combination section are night,single vehicle and young drivers.