Research On Freeway Traffic Conflict Prediction Model Considering Traffic Behavior Characteristics

As the main part of the freeway operation system,drivers have different impacts on traffic safety in the traffic flow.Analyzing the impact of freeway traffic behavior characteristics on operational risks has great significance to improving operational safety.This paper took the freeway traffic behavior characteristics as the main line,analyzed the internal relationship between driver characteristics and traffic behavior characteristics,built a traffic conflict prediction model and an operation risk analysis model based on the traffic conflict theory,and proposed improvement measures from the perspective of traffic flow operation and traffic behavior characteristics.Analyzed the traffic behavior characteristics of the freeways from the macro and micro levels.At the micro level,a scientific and reasonable questionnaire was designed around the characteristics of driver’s traffic behavior,and the driver’s scale data of the research section was collected.The results of reliability test and validity test of the scale show that the scale has good structural validity and stability,and it was suitable for exploratory factor analysis.Dimensionality reduction was performed by principal component analysis to obtain three dimensions:aggressive driving tendency,negligent driving tendency and bad traffic behavior attitude.The inner relationship between driver’s gender,age,driving experience and traffic behavior characteristics were analyzed by analysis of variance and least significant difference method.At the macro level,the traffic flow data of eight research roadway sections was collected.Determined the processes of traffic conflicts caused by different traffic behaviors of vehicles on freeways through theoretical analysis methods,and traffic conflicts were mainly classified as rear-end conflict and lateral conflict according to the conflict angle.Compared the judgment indicators of rear-end conflict and lateral conflict,and determined the TTC indicator and extended TTC indicator as the judgment indicators of rear-end conflict and lateral conflict according to the characteristics of the collected data,then collected traffic conflict data of the research road sections.Based on the accident pyramid theory and the traffic conflict data of the research sections,the freeway traffic conflict prediction model was established.A non-parametric test method was utilized to conducted hypothesis testing on the traffic conflict data.The test results show that the traffic conflict mainly obeys the negative binomial distribution.12 modeling independent variables were initially selected based on the characteristics of the data and the results of the collinearity test.The random effects negative binomial model and the random forest algorithm were utilized to establish a freeway traffic conflict prediction model,and the models were compared through 10-fold crossvalidation.The traffic conflict was utilized as a reflection index of the roadway operation risk level,the roadway operation risk level was divided by the K-means clustering method,and the freeway operation risk analysis model was built by the multiLogistic model to study the influencing factors of roadway operation risk.Based on the research results of models,from the aspects of traffic flow operation characteristics and traffic behavior characteristics,measures to improve freeway operation safety such as adjustment of road speed limits and large-scale vehicle control were proposed.The research results of this paper reveal the internal connection between freeway traffic behavior characteristics and driver characteristics.The established traffic conflict prediction model and operation risk analysis model provide new methods or new ideas for freeway traffic safety analysis,and provide a certain reference for the arrangement of freeway operation control plan and safety guarantee measures.