Research On Early Warning Of Traffic Safety At Urban Road Intersections

The intersection of urban roads is the node of urban traffic.It plays an important role in the traffic management of road networks.Its traffic safety has always been concerned.How to improve the security of intersections has always been a concern for road traffic managers.If the intersection traffic conditions can be detected in real time and an early warning can be implemented,and the intervention of road traffic safety is achieved through the control means,the traffic safety at the intersection will be greatly improved.Safety factors of the intersection should be analysed and appropriate indicators should be selected to implement early warning.The research object of the paper is the intersection of urban roads affected by large-scale traffic attraction points.The purpose of the research is to find an early warning method that can make predictions on the traffic safety status of urban road intersections and provide timely feedback.After the traffic safety warning information at the intersection was issued,the influencing factors were promptly fed back and the security status of the intersection was specifically improved through control and management measures.In this paper,starting from the research of domestic and foreign related literature,the direct indicators of traffic safety early warning are studied,and accidental accidents are eliminated.Through the analysis of the formation mechanism of traffic conflicts at intersections,the number of traffic conflicts is finally selected as a direct indicator of traffic safety warning for plane intersections.The influence factors of traffic safety at the intersections were analyzed by the aspects of people,vehicles and roads,and 11 warning system indicators were determined according to the data from field observations.The grey relational analysis and Pearson correlation analysis are used to correlate the indicators of the early warning system.The grey relational analysis is used to determine the order of the weights of the indicators.The Pearson correlation analysis is used to determine the indicators with strong correlations.The combination of grey GM(1,1)prediction model and BP neural network prediction model forms a grey GM(1,1)conflict prediction model based on grey incidence analysis and a BP neural network conflict prediction model based on Pearson correlation analysis.And the data collected from representative intersections in Beijing,Shenzhen,and Jinan were used to test the model,and the BP neural network prediction model based on Pearson correlation analysis was finally determined.After the prediction model gives the conflict prediction number,the safety of the plane intersection is expressed by the ratio of the conflict number and the mixed traffic volume,and the warning level is given according to the warning threshold.After the release of warning information,the traffic management department can determine the measures for improvement by weighting the early warning indicators at intersections.