Research On Traffic Vibration Phenomenon At Intersections In Autonomous Driving Environment

The problem of intersection congestion is an important factor that affects the traffic capacity of urban roads.When the signal controls the red light at the intersection,the traffic flow will be interrupted,which will cause traffic shock.How to effectively alleviate this problem and improve the traffic capacity of signalized intersections is an area that deserves the attention of researchers.In the context of the widespread application of big data and artificial intelligence,autonomous driving is an inevitable trend in future transportation development.Accurate regulation of vehicles based on autonomous driving technology will become an effective method to solve the problem of traffic turbulence.Aiming at the above-mentioned problems,this paper constructs a signal optimization and vehicle speed control model based on automatic driving technology,which provides a reference for the planning,construction and operation management of automatic driving intersections.First,from the perspective of intersection classification,conflict points,and signal control methods,the problem of intersection control is explained,and the concept and characteristics of traffic shocks are introduced.Analyze the defects of traditional signalized intersections,select a cross-shaped intersection in Xi’an to carry out the test,and further analyze the traffic shock based on the driver’s car following data.Secondly,an automatic driving signalized intersection control mode and system structure have been established.Based on vehicle-to-road communication technology,intelligent vehicle-mounted equipment and intelligent road test equipment have been built.In an ideal fully automated driving environment,construct a signal optimization model and a vehicle speed control model;improve and optimize the model based on the transitional environment of a mixture of autonomous and manual vehicles.Use the SUMO simulation platform and Gurobi optimization model solver to build a test platform to simulate and verify the model.The results show that compared with the traditional signalized intersection,the intersection control model constructed in this paper can effectively reduce the average time of vehicles and the number of stops,which verifies the feasibility of the model.In summary,this research can provide scientific basis and theoretical guidance for improving traffic turbulence at signalized intersections,and has practical and practical significance.