Trajectory Optimization Of Connected And Autonomous Vehicles On Roadway Segments In Fully Automated Driving Environment

The rapid development of connected vehicle and self-driving technologies makes the large-scale deployment of connected and autonomous vehicles(CAVs)possible in the future.The introduction of CAVs will bring new opportunities as well as challenges to urban traffic control and management.Controlling and managing urban traffic after the large-scale application of connected and autonomous vehicles is becoming an emerging topic for urban traffic management agencies.On the other hand,some scholars have proposed new traffic organization strategies at intersections,which can effectively increase intersection capacity under ideal situations.This paper intends to take the advantages of CAV technologies to support the realizations of new traffic organizations at intersections.Since directly controllable,CAV can be the executor of traffic control and management strategies.When all vehicles are fully autonomous,coordinated control of CAVs may positively impact the propagation of traffic flow.This paper will propose methods to optimize the longitudinal and lateral trajectories of CAVs to achieve different driving targets,so as to improve the traffic flow efficiency.The main research contents are as follows:(1)To achieve a proposed "stable traffic state",the relationships among the length of the road section,the signal timing plans at intersections and the travel speed on the road section were studied.Based on the geometric relationship among the road length,the signal timing plans and the speed limits of the road section in the time-space diagram,the minimum length of the road for vehicles to pass consecutive intersections without stopping is derived.Given road length,the upper and lower speed limits for vehicles to pass through consecutive intersections without stopping can also be determined.(2)A floating point algorithm was proposed to control the trajectories of CAVs.The algorithm determines the locations of CAVs in the traffic flow according to a series of prescribed rules,so that the entire trajectories for all CAVs from the very beginning that they enter the road to the end that they leave the road can be generated.The algorithm can be applied when the number of entering lanes and that of the leaving lanes are not equal at intersections.Results show that the floating point algorithm was able to calculate the trajectories of CAVs in limited time to achieve intended driving target.(3)Another rolling horizon method is developed to optimize the longitudinal and lateral trajectories of CAVs.A safety region,in form of an ellipse,is first proposed to ensure both longitudinal and lateral safety distances between two neighbor CAVs.CAV trajectories are optimized so that vehicles of different turning directions form tandem arrangement.The trajectory control procedure mainly includes two modules.A nonlinear program SEPARATION to separate CAVs with different turning directions along the driving direction,and another nonlinear program PLATOON to distribute CAVs with the same turning directions evenly across all lanes.Finally,the rolling horizon method is applied to solve SEPARATION and PLATOON models.Results show that the rolling horizon approach can generate tandem arrangement of CAVs faster than the floating point algorithm.