摘要
针对矢量场直方图(VFH+)算法在路径规划过程中容易陷入环境死区,生成的路径不能满足车辆运动学限制的问题,提出方向引导的VFH+路径规划算法。首先在双向快速随机树(Bi-RRT)节点扩展中引入车辆的运动学约束,在去除路径冗余节点的基础上,使用三次B样条曲线得到平滑引导路径。其次,在VFH+算法中引入车辆的最大转角约束与引导路径的离散点方向,来限制VFH+的候选方向范围,并修改代价函数获取合适的前进方向。最后,在MATLAB软件上进行算法的仿真对比以及基于ROS平台的实验验证。结果表明,改进后的VFH+算法能够在满足车辆运动学约束的情况下,生成一条避开环境死区的有效路径。
Aiming at the problem that the vector field histogram(VFH+)algorithm is easy to fall into the environment dead zone in the path planning process,and the generated path cannot satisfy the vehicle kinematics constraints,this paper proposed a direction guided VFH+path planning algorithm.Firstly,it introduced the kinematic constraints of vehicles into the node expansion of Bi-RRT.On the basis of removing redundant nodes on the path,it obtained a smooth guidance path by using cubic B-spline curves.Secondly,in the VFH+algorithm,it introduced the maximum angle constraint of the vehicle and the discrete point direction of the guidance path to limit the range of candidate directions of VFH+,and modified the cost function to obtain the appropriate forward direction.Finally,it performed the simulation comparison and experimental verification on MATLAB software and the robot operating system(ROS)platform.The results show that the improved VFH+algorithm can gene-rate an effective path to avoid the environment dead zone under the condition of meeting the vehicle kinematics constraints.
作者
朱茂飞
贺晨辰
张春鹏
吴琼
朱守力
Zhu Maofei;He Chenchen;Zhang Chunpeng;Wu Qiong;Zhu Shouli(College of Advanced Manufacturing Engineering,Hefei University,Hefei 230601,China;Anhui Intelligent Vehicle Control&Integrated Design Technology Engineering Research Center,Hefei 230601,China;Anhui Jianghuai Automotive Group,Hefei 230601,China;Hefei ChangAn Automobile Co.,Ltd.,Hefei 230001,China)
出处
《计算机应用研究》
CSCD
北大核心
2023年第7期2090-2095,共6页
Application Research of Computers
基金
安徽省高校自然科学研究项目(KJ2021A0988)
合肥学院人才科研基金资助项目(20RC06)
合肥学院研究生创新项目(21YCXL11)
安徽省新能源汽车暨智能网联汽车产业技术创新工程项目(wfgcyh2020477,wfgcyh2021439)。
关键词
路径规划
双向快速扩展随机树
矢量场直方图
转角约束
方向引导
智能车辆
path planning
bi-directional fast extended random tree
vector field histogram
corner constraint
direction guidance
intelligent vehicle
