汽车工程

汽车工程 ›› 2022, Vol. 44 ›› Issue (8): 1153-1161.doi: 10.19562/j.chinasae.qcgc.2022.08.005

所属专题： 智能网联汽车技术专题-规划&控制2022年

• • 上一篇    下一篇

面向混合自动驾驶车流的协同自适应巡航控制

彭理群1,2,王依婷1,马育林2,许述财3()   

  1. 1.华东交通大学交通运输工程学院，南昌 　330013
    2.清华大学苏州汽车研究院（相城），苏州 　215132
    3.清华大学，汽车安全与节能国家重点实验室，北京 　100084
  • 收稿日期:2022-02-06 修回日期:2022-04-07 出版日期:2022-08-25 发布日期:2022-08-25
  • 通讯作者: 许述财 E-mail:xushc@tsinghua.edu.cn
  • 基金资助:
    国家重点研发计划项目(2018YFE0204302);国家自然科学基金(52062015);江西省研究生创新基金(YC2020-S331);交通行业重点实验室开放课题（JTZL1904）资助

Cooperated Adaptive Cruise Control for Mixed Autonomous Traffic Flow

Liqun Peng1,2,Yiting Wang1,Yulin Ma2,Xü Shucai3()   

  1. 1.School of Transportation Engineering，East China Jiaotong University，Nanchang 　330013
    2.Suzhou Automotive Research Institute，Tsinghua University，Suzhou 　215132
    3.State Key Laboratory of Automotive Safety and Energy Tsinghua University，Beijing 　100084
  • Received:2022-02-06 Revised:2022-04-07 Online:2022-08-25 Published:2022-08-25
  • Contact: Xü Shucai E-mail:xushc@tsinghua.edu.cn

RichHTML

19

PDF (PC)

661

赞

摘要：

车联网V2V环境下能实时获取自车和周围车辆的运动状态、驾驶工况和道路环境，为汽车自适应巡航控制系统提供更准确的信息。为消除自动驾驶汽车（AV）和人工驾驶汽车（MV）混合行驶工况下的车头时距干扰对汽车纵向巡航控制的影响，提出了一种基于车联网V2V的协同自适应控制方法。通过车联网V2V实时采集车辆跟驰过程中车辆基本安全信息（basic safety message，BSM），进而获得车辆相对运动状态和驾驶行为序列；应用线性最优二次型方法建立驾驶操纵序贯链优化目标函数，再对扰动作用下的汽车运动状态改变量进行短时预测；在此基础上，以混合车流车头时距的最优均衡状态为目标，构建了车辆跟驰间距的滚动优化模型和协同自适应控制方法。实验结果表明，在头车加/减速行驶工况下，改进后的车辆控制器能更快响应前车运动状态的变化量，并在保证车辆安全跟驰间距的情况下，降低了车头时距，提高了道路通行能力。

关键词: 自适应巡航控制, 混合异质车流, 协同控制, 模型预测控制

Abstract:

In the V2V environment of the internet of vehicles， the real-time movement status， driving conditions and road environment of the vehicle and its surrounding vehicles can be collected so as to provide more accurate information for vehicle adaptive cruise control （ACC） system. In order to eliminate the interference of headway fluctuation on longitudinal ACC under the mixed driving conditions of autonomous vehicles （AV） and manual vehicles （MV） ， a V2V based cooperative adaptive control method is proposed. Firstly， through V2V of the internet of Vehicles， the basic safety message （BSM） of the vehicle in the process of vehicle following is collected in real time so as to obtain the relative motion state and driving behavior sequence of the vehicles. Then， the linear quadratic regulator （LQR） is applied to infer the sequential driving maneuvers and short-term predict the vehicle motion state under the disturbance of traffic vehicle acceleration and deceleration. On this basis， a rolling optimization model and a cooperative adaptive control method for vehicle following distance are established to acquire the optimal equilibrium state of the headway of the mixed traffic flow. The test results show that under the acceleration/deceleration driving conditions of the preceding vehicle， the improved vehicle controller can respond faster to the change of the moving state of the preceding vehicle， and reduces the headway while ensuring the safe car-following distance and high traffic capacity.

Key words: adaptive cruise control, mixed heterogeneous fleet, cooperative control, model predictive control