Simulation And Testing For Connected-Automated Vehicle Following In Heterogeneous Traffic Flow

The realization of a completely high-level connected aotomated vehicles will inevitably experience a long-term heterogeneous traffic scenario in which Autonomous Vehicles(AV)and traditional human-driven vehicles(HV)coexist,and the research on the following model of connected aotomated vehicles under heterogeneous traffic flow is of great significance.In view of the fact that most of the existing research focuses on the following behavior between the same type of vehicles,and does not consider the coupling relationship between different types of vehicles and other traffic subjects,the main research content of the paper is as follows:(1)The existing research on following behavior in heterogeneous traffic scenarios did not deeply explore the intrinsic relationship between surrounding vehicles and the main vehicle.In order to solve this problem,this paper proposes a Connected-Automated Vehicles(CAV)follow-up model that considers the influence of single rear vehicle and multiple front vehicles.Firstly,on the basis of the intelligent driver model,the influencing factors such as the nose spacing,speed difference and acceleration of single rear vehicles and multiple front vehicles are considered.Secondly,the molecular dynamics theory is introduced to quantitatively analyze the potential energy affecting the vehicle,and the potential energy is superimposed according to the strength of the influence of the motion state.Finally,the CAV following model was calibrated,evaluated and compared with the CAV model by using the real vehicle test data of PATH laboratory.The results show that the CAV following model proposed in this paper considering the influence of single rear vehicle and multiple front vehicles more accurately reflects the following behavior of connected aotomated vehicles under actual roads compared with the collaborative adaptive cruise control following model proposed by PATH Lab.(2)Most of the test verification of the following model is based on virtual simulation software,and there are problems such as idealization of prediction results.In this paper,a test and verification method for virtual and real twin scenarios of following queue based on digital twin technology is proposed for CAV follow-up model considering the influence of single rear vehicle and multiple front vehicle.Relying on the "Closed Test Base for Autonomous Driving" of Chang’an University,infrastructure road data and traffic scene parameters are collected,and dynamic following traffic flow data is generated by simulation software,and the traffic scene of the queue is virtualized.Finally,the real test vehicle,traffic flow simulation and static traffic scene are combined to inject into the virtual following queue scene to realize real-time dynamic interaction between the real vehicle under test and the virtual test scene and improve the test efficiency.(3)Based on the virtual and real twin scenario of the following queue,the following model proposed in this paper is tested and verified.By modifying the vehicle type in the same location,two interference signals of acceleration and braking are introduced,and the disturbance of the following queue and the speed and acceleration changes of the measured vehicle are compared and analyzed.The experimental results show that during the acceleration and deceleration of the following queue,the CAV model has a certain effect on the blocking and slowing of the disturbance signal,which increases the anti-interference ability of the following queue.In addition,the influence of CAV model on the traffic behavior of the entire queue is verified from the perspective of different mixing rates of CAV.Under the condition that the total number of vehicles in the following queue remains unchanged,with the increase of CAV mixing rate,the operating efficiency and traffic capacity of the following queue are significantly optimized.