汽车工程 ›› 2023, Vol. 45 ›› Issue (7): 1174-1183.doi: 10.19562/j.chinasae.qcgc.2023.07.008

所属专题: 智能网联汽车技术专题-规划&决策2023年

• 专题:汽车智能化关键技术 • 上一篇    下一篇

基于自适应拟合的智能车换道避障轨迹规划

李军1,2(),周伟1,2,唐爽1,2   

  1. 1.重庆交通大学机电与车辆工程学院,重庆 400074
    2.轨道交通车辆系统集成与控制重庆市重点实验室,重庆 400074
  • 收稿日期:2022-12-16 修回日期:2023-01-17 出版日期:2023-07-25 发布日期:2023-07-25
  • 通讯作者: 李军 E-mail:cqleejun@163.com
  • 基金资助:
    国家自然科学基金(51305472);重庆市轨道交通车辆系统集成与控制重庆市重点实验室项目(CSTC2015yfpt-zdsys30001);重庆市研究生培养项目(JDLHPYJD2018003)

Lane Change and Obstacle Avoidance Trajectory Planning of Intelligent Vehicle Based on Adaptive Fitting

Jun Li1,2(),Wei Zhou1,2,Shuang Tang1,2   

  1. 1.School of Mechanical and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074
    2.Chongqing Key Laboratory of Rail Vehicle System Integration and Control,Chongqing 400074
  • Received:2022-12-16 Revised:2023-01-17 Online:2023-07-25 Published:2023-07-25
  • Contact: Jun Li E-mail:cqleejun@163.com

摘要:

研究动态环境下的自动驾驶车辆换道避障模型预测轨迹规划。首先定义坐标系及动态换道避障场景,确定车辆换道避障轨迹规划的行驶约束条件,搭建被控车辆和障碍物车辆模型;然后设计自适应分段贝塞尔曲线拟合算法用以拟合换道避障轨迹规划的离散点序列;进一步采用模型预测控制算法,设计自动驾驶车辆换道避障轨迹规划方法。最后,对动态环境下自动驾驶车辆换道避障模型预测轨迹规划进行仿真分析,结果表明:本文的换道避障轨迹规划方案能够使被控车辆在动态环境下完成障碍物的规避并保持稳定状态;所设计的自适应分段贝塞尔曲线拟合算法拟合轨迹连续且光滑,换道避障轨迹规划提供的离散点序列与贝塞尔曲线对应点的拟合残差保持在设定阈值内。

关键词: 自动驾驶车辆, 贝塞尔曲线, 模型预测控制, 动态避障, 换道轨迹规划

Abstract:

In this paper, the predictive trajectory planning of intelligent vehicles of lanes changing and obstacle avoidance in dynamic environment is studied. Firstly, the coordinate system and dynamic lane change and obstacle avoidance scenarios are defined, to determine the driving constraints of vehicle lane change and obstacle avoidance trajectory planning, and the controlled vehicle and obstacle vehicle models are built. Then, an adaptive piecewise Bezier curve-fitting algorithm is designed to fit the discrete point sequence of the lane change and obstacle avoidance trajectory planning. Further, the model predictive control algorithm is used to design the trajectory planning method for intelligent vehicles for lane changing and obstacles avoidance. Finally, the prediction trajectory planning method of the intelligent vehicle lane change and obstacle avoidance model in a dynamic environment is simulated and analyzed. The results show that the trajectory planning scheme of lane change and obstacle avoidance in this paper can enable the controlled vehicle to complete obstacle avoidance and maintain a stable state in dynamic environment. The adaptive piecewise Bezier curve fitting algorithm designed in this paper has a continuous and smooth fitting trajectory, and the fitting residual between the discrete point sequence provided by the lane change and obstacle avoidance trajectory planning and the corresponding point of the Bezier curve is kept within the set threshold.

Key words: intelligent vehicles, Bezier curves, model predictive control, dynamic obstacle avoidance, lane change trajectory planning