Identification Of Black Spot Intersection Based On Empirical Bayes Method

With the acceleration of China’s urbanization,the occurrence of traffic accidents is increasing.In recent years,traffic safety has gradually become an important research direction in the field of domestic traffic engineering.The identification of black spots in urban intersection accidents is particularly important as an important part of traffic safety research.Accurate identification of black spots in urban intersections is of great significance to improve the traffic safety of the city.In this research,the basic theoretical research of accident blackspot identification is firstly explained,and the existing accident blackspot identification methods are analyzed.Based on the basic theory of accident blackspot identification,the urban road blackspot identification methods are divided into two types: numerical evaluation method and indirect evaluation method according to different evaluation indexes.The advantages and disadvantages of the two types of methods are analyzed.The traffic accident data in Shenzhen from 2014 to 2016 is analyzed,then the target intersections for the study are selected.Through the research on the basic information and attributes of traffic accidents in Shenzhen,the law of different attributes such as time,collision type,and location of accidents in the whole city of Shenzhen has been obtained.With reference to the method of dividing the study area of intersections in the "Road Traffic Safety Handbook",several accident-prone intersections are selected as the objects of follow-up research.Subsequently,an improved empirical Bayes method is proposed to identify accident black spots at intersections.In order to take into account the advantages of the urban road black spot identification method and overcome its disadvantages,the empirical Bayes posterior estimation method combined with the quality control method is used as the evaluation method to study the black spot of urban road intersection accidents.In order to reduce the error value predicted by this method and reduce the subjectivity of the evaluation index,the environmental factor influence correction and the evaluation index correction are carried out,and an improved empirical Bayes posterior estimation model is proposed.Using an improved empirical Bayesian posterior estimation model,the basic information of traffic accident data in Shenzhen from 2014 to 2016 and identifies black spots at intersections where accidents were frequent are studied.Using the intersection performance function as a priori estimate and the environmental impact factor correction coefficient to correct the predicted value of the prior estimate,the quality control method of the equivalent accident rate method is combined to identify the accident black spot.Finally six accident black spots are recognized as traffic black spots in Shenzhen.Based on the empirical Bayesian method,the model modified in terms of environmental factor coefficients and evaluation indicators is used to identify black spots in Shenzhen.Based on the examples of this research,a research method system for the identification of accident blackspots for specific intersections is proposed,and it is extended to the accident blackspot identification methods applied to intersections at different levels,which provides pertinence for the relevant departments of traffic management.