自动驾驶车辆交叉口协调控制策略

发布时间：2018-04-02 15:05

本文选题：自动驾驶　切入点：车路协同系统　出处：《哈尔滨工业大学》2017年硕士论文

【摘要】：交叉口是城市道路交通的重要组成部分,具有疏散车流及疏导人流的作用,同时也是阻碍城市道路通行效率提升,制约城市道路通行能力增加的瓶颈之处。随着私家车数量的迅速增加,城市道路交叉口的运行状况堪忧,特别是在车流量较大、交通组成复杂的交叉口,交通拥堵屡见不鲜,交通事故亦常有发生。智能车路协同系统的出现使得自动驾驶车辆在通过交叉口时,车辆间发生信息交互,采用集中控制的方式,按照一定的通行规则,自动避免交通冲突,无障碍通过交叉口。从而有效缓解交通拥堵,减少交通事故,提高通行效率,增大通行能力。本文主要研究的是自动驾驶车辆在交叉口的协调控制策略,该策略包括构建自动驾驶车辆优先通行规则库、自动驾驶车辆路段控制模型及自动驾驶车辆交叉口协调控制算法。本文研究对象均为自动驾驶车辆,行驶至交叉口的车辆均受到中央协同控制器的集中控制。以“可控间隙”理论为基础,将研究区域分为路段控制区、协调控制区和核心冲突区。通过设定自动驾驶车辆在核心冲突区的目标位置,构建自动驾驶车辆优先通行规则库。当自动驾驶车辆通过路段控制区时,根据其启动及行驶过程的特点,构建同步启动模型和车队合并模型。将车辆进入协调控制区域的状态进行初始化,使车辆以车队的形式行驶,保持车队行车间隔相同,速度相等,即消除车辆进入协调控制区域的随机性,简化后续算法设计及模型构建的复杂度。当自动驾驶车辆通过协调控制区域时,通过车路协同技术,将信息传送至中央协调控制器进行集中式处理,当预测到交叉口内的自动驾驶车辆即将发生冲突时,依据本文构建的车辆优先通行规则库,改变即将发生冲突的车辆的运行速度,合理安排自动驾驶车辆的通行顺序,实现车辆冲突消解,保障车辆安全、高效地通过交叉口。因此,从微观层面上讲,本论文提出的通行规则及交叉口协调控制算法,将为交叉口控制方式提供新的途径,为车辆在交叉口的通行开辟新的领域;从宏观层面上讲,也为自动驾驶车辆未来的发展及普及提供理论依据和基础支撑,为未来的城市交通管理模式提供借鉴等。
[Abstract]:Intersection is an important part of urban road traffic, which has the function of evacuating traffic flow and diverting people flow. It is also the bottleneck that hinders the improvement of urban road traffic efficiency and restricts the increase of urban road traffic capacity.With the rapid increase of the number of private cars, the operation of urban road intersections is worrying, especially in the intersections with large traffic flow and complex traffic composition, traffic congestion is common, traffic accidents often occur.With the appearance of intelligent vehicle-road coordination system, the information exchange occurs between the vehicles when the autonomous vehicle passes through the intersection. According to certain traffic rules, the traffic conflict can be avoided automatically and the traffic can pass through the intersection without obstacles by means of centralized control.In order to effectively alleviate traffic congestion, reduce traffic accidents, improve traffic efficiency, increase traffic capacity.This paper mainly studies the coordinated control strategy of self-driving vehicles at intersections. The strategy includes the construction of self-driving vehicle priority rules library, self-driving vehicle section control model and self-driving vehicle intersection coordination control algorithm.The objects of this paper are all autonomous vehicles, and the vehicles travelling to the intersection are all under the centralized control of the central cooperative controller.Based on the theory of controllable gap, the research area is divided into road section control area, coordinated control area and core conflict area.By setting the target position of the autonomous vehicle in the core conflict area, the automatic driving vehicle priority rule base is constructed.When the self-driving vehicle passes through the road control area, according to the characteristics of its start-up and driving process, the synchronous starting model and the vehicle fleet merging model are constructed.The state of vehicle entering the coordinated control area is initialized to make the vehicle travel in the form of a motorcade, keeping the vehicle running at the same interval and at the same speed, that is, eliminating the randomness of the vehicle entering the coordinated control area.Simplify the complexity of subsequent algorithm design and model building.When the self-driving vehicle passes through the coordinated control area, the information is transmitted to the central coordination controller for centralized processing through the vehicle-road coordination technology, and when it is predicted that the self-driving vehicle in the intersection is about to clash,According to the vehicle priority rule base which is constructed in this paper, the speed of vehicles about to conflict is changed, the traffic order of self-driving vehicles is arranged reasonably, the conflicts of vehicles are resolved, the safety of vehicles is ensured and the intersection is passed efficiently.Therefore, from the micro level, the traffic rules and the coordinated control algorithm proposed in this paper will provide a new way for the intersection control mode, open up a new field for the traffic of vehicles at the intersection; from the macro level,It also provides theoretical basis and basic support for the future development and popularization of autonomous vehicles, and provides reference for the future urban traffic management model.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.54

【参考文献】