Traffic Flow Optimization Method For Connected Autonomous Vehicles

The rapid development of self-driving vehicles and intelligenttransportation system technology is gradually unveiling the prelude to the arrival of the era of autonomous driving.However,with the continuous increase in car ownership,the problem of urban traffic congestion has become increasingly prominent.The traffic flow optimization of networked autonomous driving is of great research value.In connected automatic driving,vehicles perceive information interaction with the outside world,and a centralized system helps vehicles make decisions and control.Through the interaction with network information,vehicles can obtain real-time road condition information with higher accuracy,more detailed,larger field of view,and no dead ends compared with automatic driving.Diversified real-time information can achieve more refined traffic flow optimization scheduling,reduce travel time delay,alleviate traffic congestion,improve transportation efficiency,and enhance users’ travel experience.Traffic flow optimization can be divided into two aspects:intersection control planning and regional traffic flow assignment.This paper carries out research from the above two aspects.In the scenario of connected automatic driving,vehicles are guided and controlled by the system to implement intelligent intersection control without traffic lights through vehicle-to-road communication.The control of intersections without traffic lights relies heavily on information,which makes the safety of vehicles challenged in the case of poor channel conditions or network congestion.To cope with the above-mentioned challenges,this paper proposes a robust intelligent intersection control method,which can prevent the collision of vehicles under any network conditions under the premise of ensuring traffic efficiency.We use the vehicle driving interference graph to solve the vehicle traffic plan and consider the driving time,current lane traffic volume and user expected speed and other factors to ensure the fairness and efficiency of vehicle passage.To further optimize the performance,this paper proposes a particle swarm optimization algorithm and establishes a simulation model for system performance evaluation.Based on the single intersection control system,a regional traffic flow coordination strategy needs to be adopted to coordinate the correlation of traffic flow between adjacent intersections and balance the traffic load of each road in the region.Taking into account the travel experience of each user and their different demands,it is necessary to ensure the fairness of the distribution plan.In this paper,the incentive-based traffic flow distribution mechanism is used to obtain an incentive value acceptable to both parties by solving the Nash equilibrium of the two-person bargaining game between the vehicle and the system.Combining individualized travel needs such as tight vehicle time and maximum comprehensive income,this paper solves the multi-objective optimization of incentive expenditure and traffic efficiency.The simulation results show that the allocation mechanism proposed in this paper reduces the system expenditure as much as possible while improving traffic efficiency and user travel experience.