Research On Safety Risk Generation Mechanism Of Steep Slope And Sharp Bends Based On VAR Model

The influence factors of road safety are complex and diverse,which has been widely concerned by peers.However,under the dynamic coupling effect of multiple factors,the research on road safety risk mechanism needs to be in-depth,the basic theory of road safety design consistency is not mature,and the idea based on active safety design is rarely embodied in modern road linear safety design.This thesis will take slope-curve section of mountain highway as an example,starting from the dynamic coupling system of human-vehicle-road and environmental factors,to conduct in-depth research on the mechanism of its safety risk.Firstly,the driving simulation platform of slope-curve section was built and driving simulation test was carried out.Based on the data of drivers’ psychophysiology and vehicle operation characteristics obtained in the experiment,the evaluation index set affecting driving safety was established by comprehensively considering the variation characteristics of human-vehicle-road and environmental state.Secondly,driving simulation experiment and theoretical analysis are combined to study the characteristics of driver psychology,driving behavior,vehicle driving state and accident rate under slope-curve section,which based on the in-depth analysis of information index variation rules of slope-curve sections,subjective survey results of drivers and accident data.The results show that with the increase of slope gradient and the decrease of curve radius,drivers’ psychology,driving behavior and vehicle running state change more significantly,resulting in higher accident rate.Then,according to the information transfer theory of quasi-reflection arc,the feature index set that significantly affects driving safety is divided into three levels:stimulus information layer,perception behavior layer and external performance layer.The VAR model was constructed to analyze the object,degree and mode of interaction between factors,and to study the structure of information transfer layer by layer.Furthermore,The characteristics of driver psychology,driving behavior,vehicle driving state and accident rate in slope-curve section were combined with the information transfer rule between layers of VAR model.The mechanism of safety risk was analyzed,and the danger points of the whole slope-curve section were ranked.The results show that lateral instability,psychological stress and improper driving behavior caused by complex alignment are the fundamental causes of accidents in slope-curve section.Finally,based on the mechanism of safety risks in slope-curve section,reasonable indicators and improvement suggestions are proposed in the aspects of speed limit,concentration,driver and vehicle management,road alignment and facility construction.It provides a basis for active safety awareness to improve the driving safety of slope-curve road and to design the alignment in line with the safe driving of intelligent vehicles.