Font Size: a A A

Study On The Influencing Factors Of Single-vehicle Crashes On The Mountainous Highways Considering Drivers' Route Familiarity

Posted on:2022-09-26Degree:DoctorType:Dissertation
Country:ChinaCandidate:G XueFull Text:PDF
GTID:1482306464482394Subject:Transportation planning and management
Abstract/Summary:PDF Full Text Request
With the goal of developing China into a country with strong transportation network in an all-round way,highway traffic in mountainous areas has also been developed rapidly.Meanwhile,the complex driving environment of mountainous highways requires better vehicle performance,higher driving skill and attention level than non-mountainous highways,which results in particularly prominent traffic safety problems including more crashes and more injuries.Single-vehicle crash is one of the most important crash types,and the occurring mechanism and key influencing factors are different from the multi-vehicle crash.In addition,route familiarity is an important factor affecting driving behavior,and previous studies have shown that it has both positive and negative effects on driving behavior and traffic safety.However,there is no unified standard to distinguish familiar and unfamiliar drivers,which makes the research results not comparable.In addition,the factors related to drivers with different route familiarities involved in mountainous highway single-vehicle crashes,and the influencing factors of the injury severities of the corresponding involved drivers are still unknown.The mountainous highway safety improving measures have not yet taken drivers'different route familiarities into account,which is not conducive to the formulation and improvement of more precise measures.This research highlights the characteristics that mountainous highway single-vehicle crashes occurring frequently and resulting in high injury severities,and route familiarity having important effects on driver behavior and traffic safety.From the perspective of driver's route familiarity,this paper explores the factors related to drivers with different familiarities involved in the crashes,and reveals the influencing factors of injury severities suffered by drivers with different familiarities.This research aims to provide more accurate measures for drivers with different familiarities to reduce the frequencies and injury severities of the single-vehicle crashes they are involved.The main research contents include:(1)Study on the measuring methods of driver's route familiarity.Firstly,based on the lens model,the process of adjusting driving behavior of drivers with different familiarities and the differences are analyzed.Secondly,based on researches of measuring familiarity in the field of psychology,a new method for measuring route familiarity considering the total driving time and time interval of drivers is proposed.Then,by analyzing the research results of the measuring methods of route familiarity based on the driving frequency and the distance between the driver's passing location and the permanent address,the boundary value of significant difference in driving behavior between drivers with high and low familiarities was obtained.This unifies the measure of driver's familiarity.Finally,based on the spatial distance between the accident location and the permanent address of the driver involved,the drivers were divided into two categories,namely,high and low route familiarity,which laid a foundation for the classification of research objects in the influencing factor analysis models.(2)Analysis on the distribution characteristics of single-vehicle crashes and the drivers'perceiving accuracy of different mountainous highway risk factors.By collecting and analyzing the crash data on the mountainous highways in Yunnan province,the distribution characteristics of single-vehicle crashes in the field of driver,vehicle,road and environment are obtained.Then,the independent variables in the influencing factors analysis models are defined,including season,time,the types of highways,weather conditions,lighting conditions,the speed limit value,the types of crashes,the types of vehicles,vehicle age,drivers'age,gender,drunk driving,the use of seat belts,etc.It is proved that there is no multicollinearity relationship between independent variables,which provides the data basis for the influencing factor analysis model.Then,by comparing the subjective and objective risk values of 4 types of special sections,including sharp turns,continuous long downhills,dangerous roadside environment,and tunnels,the accuracy of drivers'perception of the dangerous degrees of the special sections is analyzed.The influences of drivers'age,gender,crash experience and severity of injuries on the accuracy of perception were analyzed.(3)Study on related factors of the drivers with different familiarities involved in mountainous highway single-vehicle crashes.Based on the data of single-vehicle crashes on the mountainous highways,the drivers'high and low route familiarity with the accident locations were taken as dependent variables.Based on the binary logistic regression model,the correlation factor analysis model for drivers with high and low route familiarities is established,and the similarities and differences between the model results of drivers with high and low familiarity were compared and analyzed.It is found that there are large differences among the research results.Finally,from the two aspects of driving behavior adjustment of drivers with high and low route familiarity,optimization of ordinary mountainous highways and tourist highways with a large number of drivers with low route familiarity,the precise measures of single-vehicle crashes prevention on the mountainous highways are proposed.(4)Study on influencing factors of injury severities of drivers with different route familiarities in single-vehicle crashes on the mountainous highways.Based on the single-vehicle crash data on the mountainous highways,the standard ordered probit model and random-effects generalized ordered probit(REGOP)model were used to construct the influencing factor analysis model of driver injury severity with high and low route familiarity respectively.The Pseudo-R~2 value is used to verify the applicability of the established REGOP model over the standard ordered probit model.The results show that there are significant differences in the influencing factors of injury severities between drivers with high route familiarity and drivers with low route familiarity.At the same time,the factors that significantly affected both the injury severities only included 00:00-06:59,rainy days and the type of rollover accidents.Other factors have different effects on the injury severities of the two types of drivers.Then,temporal stability of the influencing factors is tested.Finally,from the two aspects of driving behavior adjustment of drivers with high and low familiarity and optimization of ordinary highways and tourist highways in mountainous areas,precise measures are proposed to alleviate the drivers'injury severities in single-vehicle crashes on the mountainous highways.
Keywords/Search Tags:Mountainous highway, Single-vehicle crash, Route familiarity, Injury severity, Binary logistic regression, Ordered probit model, Random-effects generalized ordered probit model
PDF Full Text Request
Related items