The Longitudinal Trajectory Optimization Of Autonomous Vehicles Formation Based On Mixed Traffic Environment And Improved Car Following Model

With the development of autonomous driving and connected car technology,the status of autonomous driving vehicles in the future road traffic system continues to rise.The road traffic environment of fully-automated vehicles is difficult to achieve in a short time.In the foreseeable future,the mixed driving environment is more realistic.At present,there is little research on the system performance mining of the mixed driving environment of artificial driving vehicles and autonomous driving vehicles and the formation of small-scale autonomous driving vehicles.The thesis reviews the current status of domestic and foreign research on vehicle following models and vehicle trajectory optimization algorithms,summarizes the existing research results,and on this basis determines the research goals,research contents and technical roadmap of this paper.In the mixed driving environment of artificial driving and autonomous driving vehicles,for artificial driving vehicles,we build an improved following model based on the Newell simplified following model for signalized intersections to predict the trajectory of artificial driving vehicles.For the formation of autonomous vehicles,the corresponding traffic strategy is constructed at the signalized intersection,and an appropriate trajectory optimization algorithm is introduced to optimize the longitudinal trajectory of the formation of autonomous vehicles,so that the mobility and average delay of the intersection reach the optimal goal.A single-lane single-point and signalized intersection was constructed in the Vissim simulation software,and the accurately calculated data was input into it in the form of "expected speed" to obtain a vehicle trajectory that more closely fits the actual road driving conditions,and output multiple effective Performance measurement index,simulation evaluation of the entire system.The paper builds a software platform including Matlab,Excel and Vissim,making it a relatively mature system integrating control,storage and simulation evaluation.We focus on the optimization of the longitudinal trajectory of the vehicle.This problem is the basis of many automatic driving technology applications,such as fleet control and adaptive cruise control.