Study Of Anti-collision Car-following Model In V2V Intelligent Networked Environment

With the growing number of car ownership,road driving safety issues are becoming increasingly prominent.With the growth of car interconnection technology,intelligence and networking will become the main characteristics of transportation in the future.In the V2 V environment,networked vehicles can realize to share their traffic information.This helps the drivers to predict their driving state of the vehicle ahead based on the currently obtained information,so as to drive safely and reduce the occurrence of vehicles' collision.Therefore,it is of great significance to study the vehicles' car-following behavior of anti-collision in V2 V environment and carve out its corresponding traffic flow evolution pattern.In this dissertation,two improved multi-front vehicle velocity difference models are proposed in the V2V-networked environment for cars.Theoretical analysis,parameter calibration,numerical simulation and safety analysis are given.The study is as follows:Firstly,based on the OV model,a multi-front vehicle velocity difference model(MCF-CT)is proposed by considering the V2V-network environment.The linear stability criterion of the MCF-CT model is obtained,and its stability region is larger than the OV one.It shows by computation that the ratio of the area of stable region of the MCF-CT model is higher obviously.Through the reduced perturbation method,the density wave equations,Burgers and m Kd V equations,are derived corresponding to the traffic area,and the kink-anti-kink wave solution of the m Kd V equation,The solitary wave Solution of Burgers is obtained.Futhermore,under the condition of periodic boundary conditions and by setting the normal vehicle's spacing as 8 m,the models MCF-CT and FVD is analyzed by numerical simulation in case of emergency as the vehicle's spacing is reduced by 5m.It shows a better anti-collision effect for the MCF-CT.Secondly,based on FVD model,a new car-following model(V2V-MFVD)is proposed by considering the influence of multi-front vehicles' spacings on traffic flow in V2 V network environment.The stable region of the V2V-MFVD model is analyzed under different parameter values.By the reduced perturbation method,the solitary wave solution of Burgers and the kink anti kink wave solution of m Kd V equation are analyzed and given,which can be used to describe the blocking phenomenon of traffic flow.Using the "US 101 data set" in NGSIM data set,the parameters of the V2V-MFVD model are calibrated by genetic algorithm,and the optimal parameters under the measured data are obtained.Using the calibrated parameters,two numerical simulation tests are designed for V2V-MFVD model by considering the continuous acceleration and deceleration of the model's leading vehicle.The results show that the V2V-MFVD model with calibrated parameters is more beneficial to the stable operation of the traffic streaming than the FVD model,and reduces the collision risk in car following behavior and increases the safety of traffic flow.Finally,the stability and safety are analyed for the traffic flow of V2V-MFVD model with calibrated parameters in a relatively dangerous simulation environment.A heterogeneous traffic flow scenario with normal and abnormal connected vehicles is designed.The V2V-MFVD and FVD models are simulated numerically under the same simulation conditions.The safety indexes such as collision time and standard deviation of vehicle velocity are calculated by using the numerical results.The test results show that the V2V-MFVD model has better anti-collision effect,and that the stability and safety of traffic flow are continuously improved with the increase of the proportion of normal network-connected vehicles in the total simulation vehicles.