Cooperative Control Of Autonomous Driving Vehicles In Networked Intersection Environment Algorithm Research And Simulation Implementation

Level intersections are traffic bottlenecks in urban road networks,and an efficient intersection control system can effectively improve the efficiency of traffic flow at intersections.Traditional traffic signal control methods ensure the safety of traffic flow through phase prohibition.However,under the background of increasing traffic demand and increasingly complex space-time conflicts,traditional traffic signal control methods cannot efficiently guide traffic flow and make full use of the space-time resources of the road.The rapidly developing Connected and Autonomous Vehicle(CAV)provides a new model for collaborative vehicle control at intersections.Through the advantages of vehicle-road cooperative information interaction and the high controllability of CAV,it can avoid the problem of unreasonable driving behavior leading to reduced signal control efficiency and realize multi-vehicle collaborative eco-driving in different traffic scenarios.This thesis makes full use of the accuracy of CAV vehicle control technology,based on two typical traffic scenarios of signalized intersections and non-signalized intersections,comprehensively considers the overall traffic efficiency of intersections and vehicle fuel characteristics,and analyzes the coordination of autonomous vehicles in networked intersections.Exploratory research has been carried out on the control method.The main work includes the following three aspects:(1)On the basis of investigating the current situation of vehicle-road collaborative application research at home and abroad and analyzing and summarizing intersection control methods,by analyzing the mechanism of vehicle collaborative control at cooperative intersections under the intelligent network environment,the multi-vehicle collaborative control task is disassembled into vehicle arrival time The determination of and the optimization of the ecological velocity trajectory of the vehicle in a given time slot.(2)Aiming at the cooperative control task of autonomous driving vehicles in the networkconnected signal light intersection environment,a vehicle platoon trajectory optimization framework based on leader-following hierarchical control is proposed.First,on the basis of modeling the traffic scene at the intersection of networked signal lights,the fleet is divided based on the state of road traffic flow,and the arrival scene is divided according to the rules under the premise of obtaining the signal light state and the running state of downstream vehicles,and the non-stop running time range of the vehicle is obtained;Secondly,the leading vehicle builds a multi-objective trajectory optimization function with fuel consumption and travel time as weighted items and solves it based on the genetic algorithm.A standard queue can be formed in front of the intersection and pass through the intersection at the maximum speed,realizing the ecological driving of the CAV queue at the signalized intersection under the intelligent network.(3)Aiming at the cooperative control task of autonomous driving vehicles in the environment of networked unsignalized intersections,a multi-stage timing optimization vehicle traffic framework(Multi-Stage Timing Optimization,MTO)based on area division is proposed,and the task is decomposed into vehicle state observation,There are three main tasks of arrival time optimization and trajectory tracking control.Firstly,a rule is set on the vehicle speed in the observation section to realize the estimation of the vehicle arrival time.Secondly,the conflict-free arrival time of the vehicle is optimized based on the goal of minimum delay,and the simulated annealing algorithm is used to improve the solution efficiency of the optimization model;finally,the three-stage trajectory modeling method is used to optimize the velocity profile of the vehicle to realize the trajectory under a fixed time slot track control.The simulation results show that the MTO method can achieve global vehicle coordinated traffic under various traffic flow scenarios,and significantly improve traffic efficiency in medium and high traffic conditions.(4)Based on the SUMO simulation platform,the simulation test scenarios of signalized intersections and non-signalized intersections are respectively built,and the performance of the two methods under different traffic conditions is tested and analyzed.Among them,fixed time control(FTC)is used as a comparison method in the signalized intersection scene,and the sensitivity analysis of the PTO method is carried out with parameter variables such as departure time,initial speed,control interval length,and number of vehicles in the queue;In the unsignalized intersection scene,FTC and PTO method are used as comparison methods,and the effectiveness of the proposed control method is verified under various working conditions from low saturated traffic flow to high saturated traffic flow.