Research On Multi-Velocity Difference Car-Following Model With Rear-View Effect And Acceleration Information Of Front Vehicles Under Auto-Driving Environment

In order to face and solve the important problems of road congestion and traffic safty caused by the greatly increasing number of cars in nowadays society,the new development of the intelligent-network connected auto-driving vehicles will play more and more roles.Many countries have been researching and developing auto-driving cars accelerately.Under the auto-driving environment(ADE for short),to propose and analyze car-following models can provide a new theoretical basis for relieving road congestion and ensuring traffic safety exactly.In this dissertation,by considering the monitoring characteristics of front and rear traffic information of autonomous vehicles under the auto-driving environment,mainly,an improved car-following model suitable for describing the car-following characteristics of auto-driving car is proposed on the basis of the relevant traditional following models.The following resarch results are obtained by theoretical,numerical simulation and safety analysis:Firstly,based on the Two-velocity Difference(TVD)model,an improved car-following model(named as BL-ATVD for short)is proposed by considering the rear-looking effect and the acceleration information of the vehicle in front of the autonomous vehicles on the car-following behavior.The stability conditions of BL-ATVD model are obtained by linear stability method,and some cases of enlarged stability region are analyzed by computing the ratios of the stability region’s area with multiparameters.Burgers equation,Kd V equation and m Kd V equation of the BL-ATVD model in the stable,metastable and unstable regions are derived respectively by using the reduced perturbation method,and the corresponding solitary wave solutions as well as the kink and anti-kink solutions are given.Numerical simulation results by Matlab shows that the stability of BL-ATVD model is better than that of the Full Velocity Difference(FVD)model,and the stability is enhanced with the decrease of rear-looking effect’s weight or the increase of the acceleration sensitivity coefficient of the vehicle in front.Secondly,based on the BL-ATVD model and considering the influence of multi-vehicle information in front of the car-following vehicle further,a collaborative car-following model with rear-view effect and multi-vehicle information in front of the vehicle(BL-MIC for short)is established.The linear stability criterion of the BL-MIC model and the expansion of the stability region under different parameters is given,and the ratios of the enlarged stability region’s area with multiparameters.It is found that the stability of the traffic flow is enhanced more as the number of front vehicles is more.Burgers equation,Kd V equation and m Kd V equation in the stable,metastable and unstable regions of the BL-MIC model are derived respectively,and the corresponding solitary wave as well as the kink and anti-kink solutions are given.Numerical simulation results show that the BL-MIC model is more effective model in restraining traffic congestion and enhancing road capacity than the BL-ATVD.Finally,the safety of BL-MIC model is evaluated under the heterogeneous traffic flow environment.By designing simulation experiments and calculating the safety indexes of road,the obtained simulation results show that the safety of traffic flow become better as the proportion of autonomous vehicles on road increases.This implies justly that autodriving vehicles can improve effectively road safety and reduce traffic congestion.