Driver's Accident Interval Distribution And Risk Factors

The increase in the number of vehicles and drivers worldwide has brought convenience to people's life,as well as more traffic accidents.Driver has a crucial influence on traffic accidents.The driver's accident interval is the interval between two consecutive accidents,and the shorter accident interval of the driver means that he has a higher risk of accidents.Therefore,the identification and management of dangerous drivers can be realized by analyzing the distribution of the driver's accident interval.The driver's accident interval is affected by personal attributes of driver,vehicle attributes,road characteristics,weather conditions,and many other factors.The data comes from the driver accident records and violation records of the capital of G Province.Using data cleansing and data fusion to process the data,the driver's accident data is obtained.This paper used the survival analysis theory to model the driver's accident interval,and analyze the driver's accident interval distribution and key influencing factors.The research results of this paper can provide reference for dangerous driver's warning and driver safety management.The main work is as follows:(1)The overall distribution of driver's accident interval were explored.Based on the historical data of driver's accidents and violations,Using data cleansing and data fusion to process the data,the driver's accident data is obtained.The non-parametric method of survival analysis was used to estimate the overall distribution of driver's accident interval.Based on the non-parametric method of survival analysis,the distribution of driver's accident interval under different conditions of personal attributes,vehicle attributes,and violation attributes were compared.The results show that there are significant differences in driver's accident interval in different ordinal number of accidents and different vehicle attributions.(2)Analysing of driver accident interval based on parametric model.Based on the driver's accident interval data,this paper used the accelerated failure model with six commonly distribution types such as Exponential,Weibull,Lognormal,and Generalized Gamma to analysis driver's accident Interval.The optimal distribution of parametric models with and without considering covariates was discussed.The goodness of fit for each parametric model was tested.The results show that the generalized Gamma parametric model works best.The age of the driver,the location of the vehicle,the number of violations,the existence of specific violations,serious first accident or not,and the ordinal number of accidents are the main factors affecting the driver's accident interval.(3)Influencing factors of driver's accident interval were analyzed by recurrent event model.The study of recurrent events in survival analysis and the treatment of heterogeneity in multivariate time data were presented.The PWP model and AG model of the driver's accident interval were proposed.The results show that the PWP model is better than the AG model.The location of the vehicle,the number of violations in the previous year,whether there was a specific violation in the previous year,whether first accident was serious have significant influence on driver's accident interval.