The Capacity Estimation Of Turn Lane For Noncooperative Automated Environment

Traffic stream cross each other at intersections,which causes traffic conflicts.Left-turn traffic is likely to conflict with other traffic flows,which constrains the capacity of intersections and ultimately the urban road network.On the other hand,it is difficult to balance the supply and demand of traffic,with traffic congestion the consequence.The emergence of automated vehicle(AVs)provides a new way to manage traffic congestion.Due to the both communication technological immaturity and security policy decisions,it will take time to realize the fully-connected V2 X automated-driving environment.On the other hand,non-cooperative AVs are therefore the mode of AV operation that are being tested most intensely and appear likely to commercialize most rapidly.This paper researches the capacity of protected movements within left-turn lanes for non-cooperative automated vehicle(vehicles per hour per lane).To the best of my knowledge,it is the first study to quantify the capacity of left-turn lane under the automated-driving traffic environment,which provides unique support for traffic management and road design in the presence of automated vehicles.We analyzed the unique trajectory of the turning maneuver(as compared to driving along tangent road segments),employed best-available parameters likely to control the kinematics of automated vehicles,and then formulated a model of car-following for automated vehicles.Input turning radius,speed and acceleration to simulate the vehicle's continuous turning behavior at the intersection;input parameters headway,leading vehicle's braking angle,we simulate the continuous braking process of the two vehicles..Based on the output of the turning model,we set two strategies for choosing minimum headway,and analyzed the capacity of the left-turn lane under these two principles.The first strategy is "defensive",in which a vehicle always keeps an ACDA(assured clear distance ahead)behind the vehicle ahead of it in the traffic stream and thus is “guaranteed” not to strike that vehicle from behind.We simulated the process of emergency braking for the two vehicles in which the leading vehicle could unpredictably initiate braking at any time,and determined the minimum “safe” headway which carried out no-collision results of the model.The kinematics of both ‘wheels-locked' and ‘wheels-not-locked' braking maneuvers were considered and based on the minimum safe headway,the saturated flow rate is calculated.The following conclusions are drawn:(1)Under different parameters,the capacity of left-turn lane is increased,by 0.2%(under the most conservative circumstances)to 43%;(2)The influence of turning radius on traffic capacity is determined by the influence of turning radius on vehicle speed.The larger the turning radius is,the greater the vehicle turning speed is,and the greater the capacity of a left-turn lane is;and(3)When braking rate is treated as stochastic rather than deterministic,a change in crash-risk per turning maneuver(as characterized by the probability distribution of deceleration)from one in ten thousand to one in one hundred million leads to a decrease in the capacity of left-turn lane by 4%.The second strategy is based on minimizing total cost,considering that costs associated with possible crashes decrease with increased spacing,while costs associated with travel time increase with increased spacing.Empirical turning vehicle trajectories at a field intersection were extracted as the input parameter of the turning model,and the crash cost under different vehicle spacing is obtained.Considering the lowest total cost,the optimal headway is obtained.By calculating the saturated flow rate of left-turn lane using this calculated optimal headway,we draw the following conclusions:(1)The capacity of left-turn lane is increased by 59.5% in the presence of non-cooperative AVs(relative to human drivers),and(2)When spacing is small,the effect of crash cost is dominant.When spacing increases,the effect of time cost becomes larger,and the choice of the optimal headway is determined how the two categories of cost interact with each other.