摘要
为给智能车规划一条安全、平滑且满足车辆非完整性约束的路径,针对混合A*算法规划所得路径安全性不足的问题,提出惩罚式图搜索方法,能够得到更为安全、合理的路径。另外,提出一种基于多目标约束的后处理方法以生成曲率连续变化的可行驶路径。通过仿真与实车实验,验证了算法的时效性、实用性与可靠性。
To plan a safe,smooth and non-holonomy restricted path for an intelligent vehicle,the paper studies a graph search method with penalty to improve the safety of hybrid A*algorithm,and puts forward a post-processing method based on multi-object constraint to generate a drivable path with a continuous smooth curve.Both the simulation experiment and real running validate this method’s timeliness,practicability and reliability.
作者
齐尧
徐友春
李华
王任栋
QI Yao;XU Youchun;LI Hua;WANG Rendong(Fifth Team of Cadets,Army Military Transportation University,Tianjin300161,China;Institute of Military Transportation,Army Military Transportation University,Tianjin300161,China)
出处
《军事交通学院学报》
2018年第8期85-90,共6页
Journal of Military Transportation University
关键词
路径规划
智能车
惩罚式搜索算法
混合A*
path planning
intelligent vehicle
penalty search algorithm
hybrid A*
