| In recent years,with the rapid growth of the national economy,the social demand for road transportation of “two passengers and one dangerous” vehicles has continued to increase,but at the same time,traffic accidents in this field have occurred frequently.These accidents have brought serious threats to people ’s lives and property safety and social and economic development.Therefore,it is of great significance to deeply explore the occurrence law,influencing factors and risk assessment of road transportation accidents of “two passengers and one dangerous” vehicles to ensure the safe transportation of “two passengers and one dangerous”vehicle.Firstly,based on the induction and summary of the relevant research status at home and abroad,this paper collects the relevant data information of 175 operating passenger cars and386 dangerous goods road transport vehicle accidents,and makes a statistical analysis of its distribution characteristics,and obtains the basic law of “two passengers and one dangerous”vehicle accidents.In addition,through the analysis of the cause statistics and key elements of the “two passengers and one dangerous” vehicle road transportation accident,the influencing factors of the accident risk research model are preliminarily determined from multiple perspectives.Secondly,when the Apriori association rule algorithm is used to mine the “two passengers and one dangerous” accident data,a large number of invalid and redundant data rules will be generated.Through the combination of grey correlation analysis and Apriori association rule algorithm,the factors affecting the “two passengers and one dangerous” vehicle road transportation accident are first screened to reduce the generation of redundant rules,and then the association rules between the factors are mined.Based on association rules and expert knowledge,the interpretative structural model of road transport accidents of “two passengers and one dangerous” vehicles is established.Finally,10 significant factors affecting road transport accidents of operating buses and 13 significant factors affecting road transport accidents of dangerous goods vehicles are obtained from personnel attributes,vehicle attributes,road attributes and environmental attributes.Furthermore,based on the statistical analysis of the “two passengers and one dangerous”vehicle accident data,the interpretative structural model and the K2 learning algorithm,the accident risk assessment models of operating passenger cars and dangerous goods road transport vehicles based on Bayesian network structure are established respectively.The model parameters are learned by Netica software to obtain the conditional probability of Bayesian network,and then the stability and accuracy of Bayesian network are verified by K-fold crossvalidation method.The comparison shows that the prediction accuracy of the Bayesian network model for the accident risk assessment of commercial buses and dangerous goods road transport vehicles is high,and the model is accurate and effective.Lastly,based on the Bayesian network model,the risk assessment of the “two passengers and one dangerous” vehicle road transport accident is carried out,and the severity and accident attitude of the accident of the operating passenger car and the dangerous goods road transport vehicle are predicted.The cause of the accident is diagnosed,and the sensitivity of the two nodes of the accident severity and accident attitude of the Bayesian network model of the operating passenger car and the dangerous goods road transport vehicle is analyzed.On the basis of the full text research,in order to take the initiative to prevent the occurrence of “two passengers and one dangerous” road transportation accidents,the corresponding prevention suggestions and improvement measures are put forward for the traffic management department,“two passengers and one dangerous” operating enterprises and drivers. |