摘要
为解决针对RRT*算法的搜索效率低下,冗余点过多,路径代价等问题,提出一种改进型RRT*算法。改进步长确定机制,通过全局自适应取代人工设置固定步长;采用分区采样点拒绝策略,通过减少重复地区的采样,减少搜索树冗余节点,提高搜索效率。对规划路径进行路径优化,减少路径代价与路径节点。通过设计两张障碍散乱与路径狭窄的地图与一张三维空间地图进行仿真实验,实验结果表明,改进型RRT*算法的搜索能力显著高于标准RRT*,对于环境有良好的鲁棒性与适应性。
To solve the inefficiency of searching for RRT*algorithm,too many redundant points and the cost of path,an improved RRT*algorithm was presented.The step determination mechanism was improved,the fixed step set manually was replaced by global adaptive.By using the partitioned sampling point rejection strategy,the redundant nodes of the search tree were reduced and the search efficiency was improved by reducing the sampling in the repeating areas.The planning path was optimized to reduce the cost of the path and the route nodes.Two maps with scattered barriers and narrow paths were designed and a three-dimensional spatial map was simulated.Experimental results show that the search ability of the improved RRT*algorithm is significantly hi10:522024-08-30gher than that of the standard RRT*,and it has good robustness and adaptability to the environment.
作者
罗济雨
孙丙宇
LUO Ji-yu;SUN Bing-yu(Heifei Institutes of Physical Science,Chinese Academy of Sciences,Hefei 230031,China;Science Island Branch,Graduate School of USTC,Hefei 230026,China)
出处
《计算机工程与设计》
北大核心
2024年第8期2357-2363,共7页
Computer Engineering and Design
作者简介
罗济雨(2000-),男,四川南充人,硕士研究生,研究方向为移动机器人路径规划;孙丙宇(1974-),男,安徽淮北人,博士,研究员,博士生导师,研究方向为模式识别与智能系统。E-mail:xiaomi@mail.ustc.edu.cn。
