引用本文:廖道争,刘孝楠,程俊,席磊.智能汽车匝道合流区域紧急避障综合运动控制[J].控制理论与应用,2025,42(7):1426~1434.[点击复制]
LIAO Dao-zheng,LIU Xiao-nan,CHENG Jun,XI Lei.Integrated motion control of intelligent vehicle for emergency obstacle avoidance on confluence area[J].Control Theory & Applications,2025,42(7):1426~1434.[点击复制]
智能汽车匝道合流区域紧急避障综合运动控制
Integrated motion control of intelligent vehicle for emergency obstacle avoidance on confluence area
摘要点击 3256  全文点击 226  投稿时间:2023-05-26  修订日期:2025-02-11
查看全文  查看/发表评论  下载PDF阅读器   HTML
DOI编号  10.7641/CTA.2023.30365
  2025,42(7):1426-1434
中文关键词  紧急避障  轨迹规划  速度跟踪  自适应模型预测控制
英文关键词  obstacle avoidance  trajectory planning  speed tracking  adaptive model predictive control
基金项目  国家自然科学基金项目(52277108)资助.
作者单位E-mail
廖道争* 三峡大学 电气与新能源学院 liaodaozen@ctgu.edu.cn 
刘孝楠 三峡大学 电气与新能源学院  
程俊 千乘研究院  
席磊 三峡大学 电气与新能源学院  
中文摘要
      针对智能汽车匝道合流区域紧急换道避障路径规划与跟踪控制问题, 本文提出了一种自适应换道避障跟 踪控制方法. 首先, 依据当前车道、目标车道车辆行驶状态提出紧急换道避障的安全约束条件, 进行换道避障路径 规划和速度规划; 然后, 基于模型预测控制(MPC)设计车辆纵向运动控制器, 实现对车速的精确跟踪; 最后, 根据车 辆实时车速提出了最优预测时域参数的选择方法, 进行了自适应预测时域MPC横向跟踪控制器的设计. 仿真结果 表明, 所提控制方法比固定时域参数模型预测控制方法具有更高的跟踪精度和行驶稳定性.
英文摘要
      Aiming at the problem of path planning and tracking control for emergency lane change obstacle avoidance in ramp confluence area of intelligent vehicle, an adaptive tracking control method is proposed, which can avoid obstacles use lane change. First of all, according to the vehicle running state at current lane and target lane, the safety constraint conditions of emergency lane change obstacle avoidance are proposed, and path planning and speed planning are carried out for lane change obstacle avoidance. Then, based on the model predictive control (MPC), the longitudinal motion controller is designed to track the vehicle’s speed accurately. Finally, according to the real-time speed, the optimal prediction time domain parameter selection method is proposed, and the adaptive prediction time domain MPC lateral tracking controller is designed. Simulation results show that the proposed control method has higher tracking accuracy and driving stability than the fixed time domain parameter method.