A Study On The Occurrence And Severity Of Traffic Accidents On Secondary Highways In Mountainous Areas

Secondary highway in mountainous areas play an important role in promoting regional economic development,promoting the development of regional resources and decisive battle against poverty.However,with the continuous growth of residents’ travel demand,vehicle quantity and freight traffic along the route,the number of traffic accidents and the number of deaths on the secondary highway in mountainous areas has become an urgent problem to be solved.The function of secondary highway in mountainous area is fuzzy,the traffic composition is complex,the linear index is poor,the traffic facilities are not perfect,the traffic safety situation is not optimistic.In this paper,based on the original data of 877 traffic accidents occurred in the past 5 years in K15～K67 of yuanshuang secondary highway,the basic characteristics,spatial and temporal distribution characteristics of traffic accidents were statistically analyzed using the specific accident classification ideas of single vehicle,vehicle-vehicle and multi-vehicle,and the accident occurrence and severity model was constructed,which has important practical significance and theoretical value.In this paper,the traffic accidents of secondary highway in mountainous areas are mainly studied from the following two aspects:Traffic accident occurrence model.Firstly,the unit sections of the sample sections are divided by using the method of 500 m and 1km fixed length and the method of indefinite length based on the plane linear GQ/HZ points.Secondly,8 independent variables related to road alignment and driving environment are defined.Then,based on the negative binomial regression and zero accumulation negative binomial regression,the accident occurrence model of single vehicle,vehicle-vehicle and multi-vehicle accidents on the secondary highway in mountain area is constructed.Finally,the effects of different section division methods and different highway grades on the accidents are compared and analyzed.The results showed that the curve ratio was significantly negatively correlated with single vehicle accidents and positively correlated with vehicle-vehicle accidents.The proportion of villages and towns and the level crossing was significantly positively correlated with the accidents of single and vehicle-vehicle accidents.The model results of multiple variables related to road alignment are poor.The method of indefinite length with GQ/HZ as segment point is more accurate than the model of fixed length.Traffic accident are prone to occur on the straight section of secondary highway in mountainous and the dangerous section with poor alignment.Traffic accident severity model.Firstly,the dependent variable is divided based on whether there are casualties or not.Secondly,11 independent variables are defined based on the four dimensions of driver,collision type,road alignment and driving environment.Then,based on the binomial logistic regression model,the severity models of single vehicle,vehicle-vehicle and multi-vehicle accidents on secondary highway in mountainous areas were constructed.Finally,the simulation results and prediction accuracy are tested.The results show that there is no significant relationship between traffic volume,weather,time and accident severity.The probability of having a serious accident is significantly different among drivers around the age of 30.The highest rates of serious consequences were found in single-vehicle accident involving pedestrians and vehicle-vehicle accident involving motorcycles.