| In recent years,the continuous development of intelligent network and Autonomous driving technology has gradually realized the commercialization of Connected and Autonomous Vehicle(CAV).However,it will take a long time for Human-Driven Vehicle(HDV)to be completely replaced by CAVs,so the mixed traffic flow scenario composed of CAVs and HDVs will exist for a long time in the future.Modelling the behavior of CAVs and HDVs in mixed traffic flow has been the focus of research in the field of intelligent transportation.In the low-visibility environment such as rain,snow and fog,not only the road wetness increases,but also the driver’s reaction time increases with the reduction of visibility.Statistics show that the death rate of traffic accidents on low-visibility conditions is about 40%.Therefore,it is of great significance to use scientific models to objectively describe mixed traffic prevalence in low visibility and to seek effective platoon control methods to improve driving safety and mixed traffic efficiency.This paper focuses on the safety and efficiency improvement of mixed traffic formation operation in low visibility,and studies the modeling of mixed traffic flow and the control of vehicle formation under lane change and non-lane change.The main work of this paper is as follows.(1)Studying on car-following behavior modelling in mixed traffic flow in low visibility.The mixed traffic flow consists of HDVs and CAVs.For HDV,the effects of low visibility on drivers’ visible distance and reaction time were analyzed comprehensively based on the literature,and the calculation formulas between visibility and safe distance and between visibility and limited speed were proposed respectively.On this basis,the formulas were introduced into the velocity recurrence formula of following vehicles,and the optimal speed calculation model under low visibility was constructed.For CAV,the rule-based following behavior characteristics of the current autonomous vehicle were analyzed.Considering the differences in following decision when the leading vehicle was HDV and CAV respectively,the following model based on Adaptive Cruise Control(ACC)and Cooperative Adaptive Cruise Control(CACC)was constructed based on the speed difference and following distance between the front and rear vehicles.The stability of the car-following model is proved by numerical simulation.(2)Study on control method of mixed platoon without lane change based on Model Predictive Control in low visibility.Firstly,to solve the problem that it is difficult to realize rapid and efficient centralized control of large-scale platoon in short time,the concept of mixed sub-platoon is proposed based on the car-following behavior characteristics of vehicles,which divides large-scale platoon into multiple sub-platoon s.Secondly,based on the deviation between the current speed and the ideal speed and the following distance of individual vehicles in the sub-platoon,the state-space equation of the following error of the mixed platoon was established.The goal of formation control was to minimize the overall following error of the platoon.The value ranges of speed,acceleration and following distance were defined from the perspective of vehicle safety and driving comfort and taken as the constraint conditions of the model.Based on the model predictive control,CAV acceleration was controlled by iterative calculation,so that the speed and following distance of individual vehicles in platoon within a period of time could be optimized.Finally,a mixed platoon optimal control method based on model predictive control is developed.The mixed platoon control method proposed in this paper and the platoon control method based on Follower Stopper are compared under different visibility.The results show that the control method proposed in this paper reduces the time required for the formation to reach the stationary state by 28.6%.The results of random vehicle interference experiments under different CAV permeability show that the time required for the platoon to recover to the stable state is gradually shortened with the increase of CAV permeability.(3)Study on control method of mixed platoon formation with lane-changing in low visibility based on quintic polynomial.In this paper,lane-changing of CAV is divided into three operating conditions: accelerated lane-changing,decelerated lanechanging or uniform lane-changing.Based on the influence factors of lane-changing of vehicles in different working conditions and the differences in the space requirements of lane-changing position,the relationship between lane-changing trajectory of vehicles and lane-changing time and speed was constructed by quintic polynomial.With lateral acceleration and change rate of lateral acceleration as the comfort index of lane change,the time of vehicles completing lane change as the efficiency index of lane change,and the longitudinal running distance of vehicles after lane change as the influence index of lane change on the surrounding traffic,the optimal lane change trajectory calculation formula was constructed.From the perspective of vehicle dynamics,minimizing vehicle trajectory tracking error was taken as the target of vehicle trajectory tracking control,and the value space of front wheel angle and increment was defined from the perspective of vehicle lane-changing stability as the constraint condition of the model.Based on the idea of model predictive control,the lateral acceleration of CAV was calculated iteratively.The mixed vehicle platoon control method based on lateral optimal trajectory tracking control is formed,and finally the vehicles can quickly and accurately complete the lane-changing and join the new platoon.In the simulation experiment,three lane-changing conditions with visibility of 50 meters were set to verify the effect of the proposed lane change formation control model,which realized that vehicles could accurately and quickly complete the lane-changing and then form platoon with vehicles in adjacent lanes. |
PDF Full Text Request