Study On Accident Characteristics Of Dangerous Goods In Road Transportation

With the growth of national economy and demand of dangerous goods,accidents of dangerous goods are emerging commonly.Accidents of road transportation of dangerous goods(RTDG),going by combustion,explosion and corrosion frequently,cause huge safety risks to people and society.By studying the accident characteristics of dangerous goods in road transport ation furtherly,the accident rate and accident hazard can be reduced.Based on the national and foreign present research of road transportation of dangerous goods accidents,the factors of 2013-2017 statistic dates of RTDG accident were analyzed comprehensively such as accident time,areas,goods category,forming,reason,severity level,weather conditions.Then the characteristics of RTDG-occurring time,morphological characteristics,accident reason,can be calculated by using qualitative and quantitative ways.Finally,the regional distribution characteristics of road transport accidents of dangerous goods are studied for the first time in combination with the concept of urban agglomeration.What’s more,the characteristics of the deaths number and reasons of RTDG’s accidents will be discussed furthely in the paper.In the aspects of characteristics of the deaths number of RTDG,the death toll of RTDG and the corresponding cumulative probability were gained in the paper.By using nonlinear fitting of number of deaths and cumulative probability in logarithmic coordinates,improving traditional F-N curve,comparing two F-N curve fitting effect of normal distribution function and Edgeworth-Cramer,the Edgeworth-Cramer F-N curve is more better that can be represented the situation of fatal accidents caused by RTDG and got more accurately deaths characteristics of RTDG’s accidents.In the aspects of characteristics of RTDG’s reasons,the paper analyses the causes of road traffic accidents.The accident model of dangerous goods in road transportation is established by Bayesian network knowledge.The structure and parameters of Bayesian network are studied for the cause of the accident,and the posterior probability of the cause of the accident is obtained.In the case of known accidents,the model can infer the possibility of accidents caused by various accident causes.The results show that the driver factor has the greatest probability of accidents of dangerous goods in road transportation,and the posterior probability is 0.504,which is the main factor leading to accidents.Secondly,the possibility of accidents caused by external factors is the greatest,and the posterior probability is 0.270.Facilities and management factors are relatively minor factors leading to accidents,with posterior probabilities of 0.147 and 0.027,respectively.The results of this paper can provide decision support for decision makers.