Roadside Accident Forecasting And Risk Evaluation Method Of Highway Horizontal Curve Segment

In the traffic accidents of all grades of roads,the lethality rate of roadside accidents is extremely high.The safety situation of roadside accidents in China is not optimistic The improvement of roadside safety level requires accurate assessment of roadside accident risk in order to take targeted measures.It can also provide a theoretical basis for rational setting of roadside clear recovery zone.This paper began with highway alignment and pavement condition,and built the maximum safe speed model that vehicles can't run off roadway on the basis of vehicle driving theory of highway curve segment.The normal distribution function was adopted and the speed distribution forecasting model was established on highway.Then the probability that vehicles run off roadway can be obtained by corresponding the maximum safety speed to the speed distribution cumulative frequency curve on a percentile speed,which is the proportion that operation speed is higher than maximum safe speed in the traffic flow of highway curve segment.Thus multiple relationship model was established among operation speed,highway alignment indexes and adhesion coefficient.Then a case study was carried out.In order to devide the threshold that affects the occurrence of roadside accident of highway curve segment,this paper selected the roadside accident form as the indicator.Based on the roadside historical accident data and the accident simulation software,the roadside accident data were obtained.By setting different driving speed and roadside environment variables,the results of the accident can be analized.After the vehicle running into the roadside,it may occur these four consequences of accidents: none slightly off the ground,slightly off the ground or fall off,occurred at most twice rollovers and more than 2 times rollovers.Thus the roadside accident severity was divided into four different grades according to the accident pattern.Based on the thresholds division of the roadside accident causes of highway curve segment,the bayesian network of roadside accident risk assessment was constructed.By calculating the conditional probability of occurrence of roadside accidents created by roadside accidents causes,the probability of occurrence of roadside accidents in different severity can be obtained in the bayesian network topology.Based on the gray clustering theory,the risk of roadside obstacle was evaluated,and the thresholds division of roadside obstacle was given.The case analysis was carried out by selecting the roadside accident of the typical vehicle,and the results of the bayesian network were compared with the results of the roadside accident simulation test,which proves the accuracy of the proposed risk assessment method.The research results of this paper are expected to provide theoretical basis and decision reference for road traffic planning,road traffic engineering construction and maintenance project.It also has application value for improving the safety level of highway roadside in China.