摘要
针对RRT(rapidly-exploring random tree)算法路径规划时间长,采样点利用率低,最终生成的路径曲折等问题,提出了一种改进RRT算法。采用基于动态概率的采样策略,避免机器人在采样的过程中陷入局部极小值;同时提出了变步长的随机树扩展策略,减少了采样点数量;最后,使用五次贝塞尔曲线对路径进行平滑处理,使最终生成的路径利于机器人移动。在MATLAB平台上进行仿真分析,并使用基于ROS的移动机器人进行实验,将改进RRT算法与RRT算法、目标偏向RRT算法进行对比。仿真结果表明改进RRT算法规划的路径长度减少了23.09%,规划时间减少了87.16%,并且路径更平滑。
Aiming at the problems of long-time cost,low utilization of sample points and unsuitable path of RRT algorithm,an improved RRT algorithm is proposed.The sampling strategy based on dynamic probability is adopted to avoid the robot falling into local minima in the process of sampling.A random tree scaling strategy with variable step size is proposed,which reduces the number of sampling points.Finally,the path is smoothed using quintic Bezier curve to make it suitable for robots to move.Simulations are performed on MATLAB and experiments are conducted using ROS-based mobile robots.The improved RRT algorithm is compared with RRT algorithm and Goal-biased RRT algorithm.The simulation results show that the length of the path planned by the improved RRT algorithm is reduced by 23.09%,the planning time is reduced by 87.16%,and the path is smoother.
作者
刘冲
刘本学
吕桉
李霞
LIU Chong;LIU Benxue;LYU An;LI Xia(School of Mechanical and Power Engineering,Zhengzhou University,Zhengzhou 450001,China)
出处
《组合机床与自动化加工技术》
北大核心
2023年第10期20-23,29,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
河南省重点研发及推广专项(科技攻关)项目(222102210051)。
