Cause Analysis Of Road Traffic Accidents Based On Fault Tree And Bayesian Network

With the continuous development of urbanization and the application of advanced science and technology in the field of transportation,the progress of transportation brings convenience to human life,economic benefits and social prosperity,but also leads to the increasingly serious traffic safety problems.Although the government departments have improved the level of traffic safety by strengthening legislation and supervision,there is still a huge gap compared with developed countries such as Europe and the United States.The main performance is that the number of traffic accidents,injuries,deaths and direct economic losses show a high growth trend.At present,the research on the cause analysis of road traffic accidents in our country is not deep enough,and even only attributes the cause factors to a single factor,ignoring that the traffic system is a complex dynamic system composed of people,vehicles,roads and environment.Therefore,in-depth analysis of the causes of road traffic accidents and effective accident prevention measures are of great value to ensure safe travelIn this paper,fault tree and bayesian network are used to analyze the causes of road traffic accidents.Firstly,the theory of accident cause,fault tree and bayesian network are introduced to lay a theoretical foundation for the analysis of traffic accident cause.Secondly,the traffic accident case data provided by the traffic police department is extracted,and the factors that may affect the occurrence of the accident are analyzed from the perspectives of people,vehicles,roads and environment,and the index system is constructed according to the influencing factors of the accident.The subjective probability of the index system is obtained by the analytic hierarchy process,and the objective probability is obtained by the normalization of the case data,and then the comprehensive probability of the two is obtained by using the principle of minimum identification information.Then,taking the urban road traffic accident as the top event,taking the personnel factor,vehicle factor,road factor and environmental factor as the intermediate event,and 38 detailed indicators as the basic event,the fault tree model of the accident is constructed,and the qualitative and quantitative analysis is carried out.Through the analysis,the main factors causing the accident can be found.However,the construction of fault tree model is relatively easy,it has some limitations,and can only carry out one-way reasoning.Bayesian network can realize two-way reasoning and more comprehensive analysis function,but its model construction is more complex.Therefore,the corresponding relationship between fault tree events and bayesian network nodes,fault tree logic gates and bayesian network connection strength can be used to transform the built fault tree model into bayesian network to solve this problem.With the help of Netica software,the Bayesian network model is subjected to posterior probability reasoning,sensitivity analysis,and maximun possible interpretation,and the analysis results of the two models are compared and analyzed.Among the four indicators,"brake failure" and "long straight line section" are obtained."," "Wet roads","Rain and snow weather","Large vehicles",and "No lighting at night" are the main causal factors;in the secondary indicators,the people,vehicles,roads,and environmental factors that caused the accident Their respective proportions were 76.4%,49.9%,62.6%,and 39.8%,indicating that the human factor in the secondary indicators is the main cause.To sum up,the study of this paper can provide support for decision-makers to improve the level of traffic safety,provide reference and guidance for managers to improve the traffic environment,and has certain reference significance and value for preventing or reducing the occurrence of traffic accidents.