Analysis And Prediction Of The Severity Of Freeway Traffic Accidents Under The Influence Of Roads And The Environment

The number of freeway traffic accidents and fatalities in my country has been increasing year by year.Reducing the frequency and severity of accidents has become a hot issue in freeway traffic safety research.This paper conducts a study on the severity of freeway traffic accidents.By exploring the influence of the characteristic factors of freeway traffic accidents on the severity of accidents,this paper proposes targeted and effective measures to reduce the severity of accidents and improve road safety.In the existing research,although many studies divide accident factors into multiple dimensions such as "person-vehicle-road-environment" and construct models to reveal the law of the impact of accident factors on the severity of the the research on the influence law of the severity of freeway traffic accidents is still lacking.Based on this,this study established a prediction model for the severity of freeway accidents affected by road and environmental factors,and explored the impact of road factors and environmental factors on the severity of accidents.The main research work of this paper is as follows:(1)Data processing and characteristic analysis of freeway traffic accidentsVariable screening of historical traffic accident data on expressways in Hebei Province is carried out.11 influencing factors are summarized from the perspective of road and environment and the variables are discretized.The characteristics of freeway traffic accidents are statistically analyzed and the distribution law of traffic accidents and the influence of various factors on the severity of accidents are analyzed from two aspects:road conditions and environmental conditions.(2)Analysis of factors affecting the severity of freeway traffic accidents based on the XGBoost modelEstablish the XGBoost model and introduce the SHAP value to interpret the XGBoost model and obtain the importance ranking of the impact of each feature on the target variable.Realize the visualization of the global impact of each feature on the target variable and qualitatively analyze the seriousness of each feature factor to the freeway traffic accident The degree of influence.(3)Construction of a prediction model for the severity of freeway traffic accidents based on Bayesian networksAccording to the selected variables and their values,a Bayesian network-based freeway traffic accident severity prediction model is constructed.Understand the basic theory of Bayesian network,select the K2 structure learning algorithm to search for the optimal Bayesian network structure;select the appropriate parameter learning algorithm and obtain the conditional probability table of each factor variable of the network structure.After obtaining the Bayesian network,the group tree propagation algorithm is used to establish a freeway traffic accident prediction model based on the Bayesian network.Finally,the learning accuracy and prediction accuracy of the model are verified,and the results show that the performance of the prediction model is better.(4)Forecast and analysis of the severity of freeway traffic accidentsCarry out case study based on freeway traffic accident data in Hebei Province and use the constructed prediction model to predict the severity of accidents on roads and environmental factors respectively.Then predict the severity of accidents under the interaction of different factors and analyze the severity of the accident according to the analysis results,corresponding safety management measures are put forward in a targeted manner to reduce the severity of the accident.