Car-following Behavior Modeling And Longitudinal Coordinated Control Of Connected Mixed Traffic Flow With Multi-vehicle Response

With the rapid development of intelligent networking technology and autonomous driving technology,the intelligent networked driving system has ushered in the change of multi-dimensional information fusion,showing the coexistence of Connected and Autonomous Vehicles(CAV)and Human-driven Vehicles(HV),and emerging new road service functions and traffic behavior characteristics.The car-following behavior modeling of connected mixed traffic flow can help to understand the car-following characteristics.To study the car-following characteristics of connected mixed traffic flow and improve the stability of mixed traffic flow,the car-following behavior modeling of mixed traffic flow has become a hot topic in the field of microscopic traffic flow research.However,the current traditional research cannot accurately describe the traffic behavior characteristics of the connected mixed traffic flow.Therefore,it is of great theoretical value and application significance to carry out in-depth research on car-following behavior modeling and longitudinal coordinated control of connected mixed traffic flow.Firstly,based on the actual car-following trajectory data of CAV,Automated Vehicles(AV),and HV,the car-following behavior in mixed traffic flow is analyzed.In the mixed traffic flow composed of CAV and HV,the different types of vehicles before and after CAV will lead to different car-following behavior decisions.By analyzing the real vehicle data of the PATH laboratory,it can be seen that CAV can follow the front vehicle to achieve acceleration,uniform speed,or deceleration with low delay.And AV will appear to delay and speed fluctuation phenomenon.The NGSIM data is filtered and denoised,and the car-following behavior of HV is analyzed.It can be seen that the change in carfollowing acceleration is related to car-following speed,car-following distance and,carfollowing speed difference.Secondly,considering the optimal speed and optimal speed memory of the headway between the front and rear vehicles,the speed difference and acceleration difference of multiple front vehicles,a connected mixed traffic flow car-following model(Multiple Front and Rear Optimal Velocity Changes with Memory,MFROVCM model)suitable for CAV and HV interactive penetration is constructed.According to the different types of main vehicles and adjacent vehicles,the MFROVCM model expression of CAV and HV is sorted out.Using the real data,the parameters of the MFROVCM model are calibrated,the stability boundary conditions of the model are derived,the stability of the model is analyzed,and the numerical simulation experiment is designed to verify the stability of the model.It is proved that the constructed model has good stability,and the macroscopic fundamental diagram model of the connected mixed traffic flow is derived.Then,combined with the modeling idea of the car-following model,a trajectory prediction model based on CNN-Bi LSTM-Attention is constructed.The vehicle trajectory data conforming to the car-following characteristics after screening and noise reduction preprocessing is applied to the training and verification of the model.The optimal model structure is determined through experiments,and comparative experiments are designed.The established model is compared with the LSTM model,the GRU model,and the CNN-Bi LSTM model.The results show that the trajectory prediction model based on CNN-Bi LSTM-Attention has higher prediction accuracy.The trajectory prediction of HV can be realized in traffic congestion.Finally,a connected mixed traffic scene is built.According to the vehicle running state information data obtained by the sensors,the downstream CAV realizes the trajectory prediction of the tail car HV based on the CNN-Bi LSTM-Attention trajectory prediction model.Based on the predicted trajectory,the CAV is induced to perform a long-term small-amplitude acceleration and deceleration operation,so that the CAV can smoothly merge into the downstream traffic flow.The effectiveness of the speed guidance strategy is verified by numerical simulation.The influence of CAV permeability and bottleneck position on suppressing traffic oscillation is compared and analyzed.It is proved that the speed coordination strategy of connected mixed traffic flow based on predicted trajectory can effectively suppress the propagation of traffic oscillation and improve the stability and operation efficiency of traffic flow.