Analysis And Application Of Phase Change Characteristics Of Expressway Bottleneck Traffic Based On Vehicle Spatio-temporal Trajectory

As the most common bottleneck type of expressway,confluence area is the intersection area of main line traffic flow and ramp traffic flow,which is seriously affected by vehicle conflict,resulting in frequent traffic problems such as congestion and accidents.This research takes Yingtian Street Expressway Viaduct in Nanjing,Jiangsu Province,China as the research object,and uses UAV aerial photography and video image target detection technology to obtain high-precision vehicle space-time trajectory data.Combined with the bottleneck traffic congestion characteristics of Expressway on ramp confluence area,this paper analyzes the vehicle space-time trajectory characteristics,time-varying characteristics of three traffic flow parameters and their relationship from the macro and micro perspective Based on the characteristics of vehicle lane change,this paper identifies the capacity decline points and traffic phase change transition characteristic points,deeply understands the vehicle lane change behavior and its influence,constructs the approximate detection model of vehicle trajectory phase change points,and discusses the appropriate control strategy combined with the actual situation,and verifies the effect through SUMO simulation.This research has important practical significance and practical value for enriching domestic vehicle trajectory database,clarifying congestion mechanism,further using the potential capacity of transportation system,guiding traffic planning,control and management,and achieving congestion mitigation or even avoidance.Firstly,the typical bottleneck section of Expressway confluence area is selected,and the UAV aerial photography ideal video is used.The two-dimensional coordinate data movement sequence set of all vehicles in the video is extracted by video image target detection technology.The sequence set extracted from the video is processed by data association,splicing,denoising,smoothing and other technical means,and then the macro image is calculated according to the frame association characteristics And micro traffic parameters,so as to extract the high fidelity trajectory data of all vehicles in the video,establish the space-time trajectory map of all vehicles,and lay an accurate and reliable data foundation for subsequent research and discussion.Secondly,the paper analyzes the vehicle operation characteristics of the merging area,and based on the microscopic angle of traffic flow and vehicle trajectory,studies the impact of the vibration wave generated by lane changing on traffic flow,and analyzes the time-varying characteristics of vehicle speed,flow and density in each lane,as well as the relationship among the three parameters of traffic flow;Then,the relationship between the evolution process of capacity decline and vehicle trajectory is clarified.The adjustment cumulative curve of the continuous section of the main line lane in the bottleneck area is constructed,and the characteristic points of capacity decline of the continuous section are extracted.Identification and analysis of transition feature points of traffic phase transition by combining spatio-temporal feature map.Based on the track data,the effect of vehicle lane change on traffic state is discussed.Finally,the approximate detection model of phase change point of vehicle trajectory is constructed,and the common Expressway mainline management and control methods and on ramp management and control methods are discussed.Combined with the actual bottleneck section of Expressway confluence area,the most appropriate management and control strategy suggestions are obtained,and the alinea single point dynamic ramp feedback control algorithm is verified by using the open source traffic simulation software sumo It’s a good effect.