无人驾驶路径规划算法研究.pdf
文本预览下载声明
算法语言 信息与电脑 2023年第2期
Information Computer
无人驾驶路径规划算法研究
*
王双桥 王星苏
( 桂林电子科技大学 信息与通信学院,广西 桂林 541004)
摘 要:快速扩展随机树(Rapidly-exploring Random Trees,RRT)是路径规划中常用到的算法之一,具有
结构简单、搜索能力强、搜索效率高的优点,但也具有随机性强、算法路径不平滑的缺点。文章通过引入动态步
长和增加平滑算法改进 RRT 算法。首先,改进动态步长,通过设置固定最大步长和最小步长,计算目标节点与障
碍物的距离;其次,计算具体每一步的步长;最后,比较最小二乘法、二次指数平滑法、三次 B 样条曲线 3 种平
滑算法的平滑性能,采用三次 B 样条曲线改进原 RRT 算法。通过 MATLAB 仿真软件对传统 RRT 算法和改进 RRT 算法
进行仿真测试,得出改进后的 RRT 在路径搜索和路径平滑方面都有一定的提升。
关键词:无人驾驶;路径规划;快速扩展随机树(RRT);动态步长;三次 B样条曲线
中图分类号:TP391 文献标识码:A 文章编号:1003-9767(2023)02-064-03
Research on Driverless Path Planning Algorithm
*
WANG Shuangqiao , WANG Xingsu
(School of Information and Communication, Guilin University of Electronic Technology, Guilin Guangxi 541004, China)
Abstract: Rapidly-exploring Random Trees(RRT) is one of the algorithms commonly used in path planning. This
algorithm has the advantages of simple structure, strong search ability and high search efficiency. But it also has the
disadvantages of strong randomness and unsmooth algorithm path. This paper improves RRT algorithm by introducing dynamic
step size and increasing smoothing algorithm. The improvement of dynamic step size is to first set a fixed maximum step size
and minimum step size, calculate the distance between the target node and the obstacle, and then calculate and determine the
specific step size of each step according to these three parameters; By comparing the smoothing performance of three smoothing
algorithms, namely least square method, quadratic exponential smoothing method and cubic B-spli
显示全部