Study On The Following Interaction Behavior Of Expressway Entrance Confluence Area Based On Trajectory

While expressways have the important mission of providing fast access to long distances within cities and relieving urban surface traffic congestion,the traffic congestion in the merging zone of expressway entrances has been a widespread concern of society.The traffic oscillation phenomenon(also known as "stop-and-go" traffic phenomenon)in this area has been the focus of scholars’ attention.Asymmetric following behavior is an important feature of human driving behavior and a microscopic manifestation of the traffic oscillation phenomenon.In-depth study of asymmetric following behavior during traffic oscillations is important for understanding drivers’ behavior patterns during traffic oscillations and reducing traffic congestion and vehicle linkage control.The main contents of the full paper are as follows:First,a complete vehicle trajectory data acquisition method is built.Using deep learning algorithm,the whole process of vehicle trajectory data acquisition method is built based on the traffic flow research demand and vehicle trajectory data characteristics.It mainly includes four steps: vehicle identification,vehicle tracking,coordinate transformation and data noise reduction,and introduces the working logic and comparative advantages of each step in detail respectively,which provides technical support for the future traffic flow data extraction technology improvement.Secondly,the microscopic driving behavior characteristics of the whole process of vehicle encountering traffic oscillations are analyzed based on vehicle trajectory data.A quantifiable data screening principle is created to reasonably screen the measured vehicle trajectory data.The driving behaviors of vehicles in traffic oscillations(before,during and after the oscillation)in the merging zone at the entrance of the expressway are studied statistically,classified,split and combined in two dimensions: distance-time plane and speed-spacing plane.The microscopic asymmetric following behaviors of the vehicles in the merging zone at the entrance of the expressway are explored,and the correlation between the microscopic following behaviors of the vehicles is quantified and analyzed.It is important to understand the impact of traffic oscillations on drivers’ driving behaviors from the perspective of drivers’ microscopic driving behaviors in the merging zone at the entrance of expressways,which is important for understanding traffic oscillations and establishing a following model that is closer to the real vehicle microscopic behaviors,as well as providing a theoretical basis for the improvement of subsequent asymmetric driving behavior models.Finally,the chain asymmetric following behavior model is validated,improved and simulated.This paper calibrates and modifies the parameters of the chain-type asymmetric driving behavior model describing the following behavior in the merging zone at the entrance of the expressway.The key influence factor of vehicle spacing is introduced to improve the original chain asymmetric driving behavior model.Simulations were conducted to verify the validity of the model,and the simulations proved that the improved chain-type asymmetric driving behavior model is more suitable for the real vehicle following situation in the merging zone of the expressway entrance than the original model,and improved the efficiency of the model.The above research results propose a control strategy that allows microscopic response hysteresis and can effectively improve the queueing efficiency.