Research On Vehicle Trajectory Optimization At Intersection In Intelligent Connected Vehicle Mixed Traffic Environment

With the gradual improvement of 5G,Internet of vehicles,big data and other technologies,intelligent connected vehicles will gradually become one of the main travel modes in the future.The traditional manual driving vehicles may be gradually replaced by intelligent connected vehicles.In this process,it is inevitable that intelligent connected vehicles and manual driving vehicles will be mixed.The intersection gathers traffic flows in multiple directions,and its traffic situation is complex.Therefore,according to the operation characteristics of intelligent connected vehicle mixed traffic flow,reasonably organizing intelligent connected vehicle mixed traffic flow through the intersection is of great significance to alleviate intersection delay and ensure traffic safety.This paper discusses the operation characteristics of intelligent connected vehicles and manual driving vehicles,analyzes the traffic flow characteristics in the case of mixed traffic,and combined with the cellular automata model,establishes the cellular automata model for road sections and the vehicle trajectory optimization model for intersections in the mixed traffic environment of intelligent connected vehicles.After that,MATLAB software is used for programming,and the simulation experiments are carried out on the two models to analyze the traffic flow characteristics of intelligent connected vehicles and manual driving vehicles in different scenarios.In the process of building the cellular automata model of road section in the mixed traffic environment,firstly,combined with the idea of safe distance,this paper analyzes the respective safe distance of intelligent connected vehicle and manual driving vehicle when following.At the same time,according to their different operating characteristics and physical characteristics,their respective evolution rules are constructed: the manual driving vehicle adopts the full speed difference model to evolve,and the randomization process of the vehicle is considered;The intelligent connected vehicle adopts the adaptive cruise control vehicle following model in the path Laboratory of the University of California,Berkeley to evolve.In addition,when constructing the vehicle trajectory optimization model in mixed traffic environment combined with the intersection,the manual driving vehicle continues to adopt the full speed difference model,and the intelligent connected vehicle adopts the Q-learning model in reinforcement learning to realize the vehicle trajectory optimization at the intersection.This paper uses MATLAB to simulate and analyze the cellular automata model for road sections and the vehicle trajectory optimization model for intersections under the mixed traffic environment of intelligent connected vehicles.The results show that in the road section,under the same traffic flow density,the higher the penetration rate of intelligent connected vehicles,the smoother the running track of vehicles,the higher the average speed and traffic flow,and the less traffic congestion.In the simulation experiment for the case intersection,the intelligent connected vehicle controlled by Qlearning algorithm can effectively optimize its own driving trajectory and guide the manual driving vehicle.With the gradual increase of the penetration rate of intelligent connected vehicles,the queue length of vehicles at intersections and the average fuel consumption of vehicles are gradually reduced.Compared with the traditional traffic environment,when the penetration of intelligent connected vehicles rises to 0.2,the average queue length of vehicles in peak and off-peak hours is reduced by 34 meters and8 meters respectively,and the average fuel consumption of vehicles is reduced by about10% and 5% respectively;When the penetration rate reaches 1,the phenomenon of vehicle queuing disappears in peak and off-peak hours,and the average fuel consumption of vehicles decreases by about 40% and 20% respectively.