Navigation-data-based Risk Evaluation Method at Intersection Entrance

Abstract:

The research adopted the traffic big data such as driving behavior，maps，traffic conditions and user volume collected by the navigation software to develop a traffic order index and evaluate the road safety risks.By analyzing the traffic safety features of road alignment，traffic control devices，traffic signals at intersection entrance，a structural equation model of intersection was build，the causes of risk at intersection entrance were analyzed in depth.The results show that the number of lanes in all directions，signal phase，the number of lanes related to the entrance，the number of U-turn lanes，the type of cross section，the type of median strip and the number of illegal behavior monitoring equipment have significant impact on traffic safety of the intersection.Congestion factors have the greatest impact on traffic safety of the intersection，then traffic signal factors and road alignment factors，and illegal behavior monitoring equipment have the least impact on traffic safety of the intersection.On improving the safety of intersection entrance， people are encourged to travel at off-peak hours to avoid the congestion，and all-red signals are proposed.In road designing， lanes should be increased in the entrance direction and median strip should be set，and the U-turn lanes should be set if the space allows.A safety evaluation method of intersection roads，facilities and signals based on driving behavior data collected by navigation software is proposed，which provides a reference for safety evaluation and cause analysis of other road types and transportation network.

Key words:intersection;traffic safety assessment;driving behavior;structural equation model（SEM）;navigation data