Analysis And Modeling Of Traffic Accident Risk Based On Driver’s Traffic Violation

Road traffic accident is a random event,with a certain chance,the cause is often complicated.The existing road traffic accident analysis technology is more concerned with the objective causes,quantify the impact of various factors,identify the road accident hotpot,accident risk estimation or forecast.As the subjective factors of the driver are difficult to quantify effectively,the traffic safety risk analysis technology based on the driver’s subjective factors is restricted.Traffic violations more accurately reflect the driver’s daily driving habits,while the driver’s history traffic illegal data collection of a variety of dangerous driving behavior.Therefore,the traffic offense as a starting point to study the relationship between dangerous driving behavior and traffic accidents can provide a new perspective for the analysis of traffic accident causes,more intuitive and in-depth understanding of the occurrence of traffic accidents.In order to explore the relationship between traffic accidents and traffic violations,the paper starts from two aspects according to the data of drivers’ historical traffic violations.:one studies is the relationship between the traffic accident interval and the traffic violation use a mature medical survival analysis method.Based on the Kaplan-Meier method,the relationship between the traffic accident interval and the traffic violation frequency is established.The risk function and the cumulative survival function of the traffic accident in the unit time are obtained for the driver with different traffic violations.Another is research the correspondence between traffic accident and traffic violation in type.The normalized and threshold method is used to select the traffic violation type which is statistically significant in the proportion distribution.The corresponding analysis algorithm is used to study the main and secondary related illegal types of different traffic accident types.and the correlation model is established.Through the case study of the actual data,the results of survival analysis show that:Different types of drivers have a significant difference in their survival curves.Survival curve and cumulative risk curve display that with the increase in the number of illegal history of the driver,the time interval of traffic accidents is greatly shortened.Median Survival Time shows that driver with a traffic offense greater than or equal to 5 times a year is less than 60,85 days for a driver who has 1-4 traffic violations and no offense.In the first three months,there was no significant difference between the drivers who not violations and less than or equal to four times violations.Correspondence analysis of traffic accident type and violation model results show that the type of traffic accident caused by driver has obvious correspondence with its historical illegal type.In the event of a "motor vehicle violation of the provisions of the use of special lane" of this type of traffic off the driver,for example,it is easier to take traffic,straight traffic accidents.Traffic accidents and traffic accidents within the various categories are also related to the illegal type of "not according to the provisions of the use of light"and "driving safety facilities are not complete motor vehicle",for example,these two types of traffic violations often lead to parking traffic accidents.The results of example analysis show that the proposed method can effectively capture the link between traffic accident and traffic violation,obtain the different types of driver accident interval,the risk rate estimation and the different types of accidents corresponding to the main type of accident,the research results can be urban traffic safety Management,prevention to provide reference.