摘要
针对快速探索随机树RRT^*算法在移动机器人路径规划过程中收敛速度慢,产生的取样空间大而密集的不足,提出改进的RRT^*算法。该算法先利用目标偏置和动态步长延伸找到最初始路径,然后快速瞄准目标区域,生成一个长椭圆区域,通过直接对该区域进行椭圆采样来集中搜索,减少采样过程的计算量,还可灵活的调整采样区域的大小。同时限制节点数,并以最大节点数受限的情况下运用少量时间规划出一条较优路径。最后,运用仿真实验证明提出的改进的RRT^*算法的有效性。
Aiming at the shortcomings of the fast exploration random tree RRT^*algorithm in the path planning process of mobile robots,and the large and dense sampling space,an improved RRT^*algorithm is proposed.The algorithm first uses the target offset and dynamic step extension to find the initial path,then quickly targets the target area,generates a long elliptical area,and performs centralized search by directly el lipsing the area,reducing the calculation amount of the sampling process.Flexible adjustment of the size of the sampling area.At the same time,limit the number of nodes,and use a small amount of time to plan a better path with the maximum number of nodes limited.Finally,the effectiveness of the proposed improved RRT^*algorithm is proved by simulation experiments.
作者
宋林忆
严华
SONG Lin-yi;YAN Hua(College of Electronics and Information Engineering,Sichuan University,Chengdu 610065)
出处
《现代计算机》
2020年第7期3-8,17,共7页
Modern Computer
基金
国家自然科学基金资助项目(No.61172181)。
