摘要
针对传统RRT算法在高维空间和复杂环境中收敛速度慢、采样效率低等问题,提出一种基于动态自适应采样与局部优化的改进算法(DA-RRT)。该算法根据区域重要性评估结果和环境信息来调整采样密度,优先在狭窄通道进行密集采样。DA-RRT引入局部优化机制来进一步优化路径质量,采用多树并行生长来加速路径搜索与优化过程。实验结果表明:DA-RRT算法相较于传统RRT算法,其路径代价、运行时间、节点数以及迭代次数均分别减少了22.0%、92.1%、43.7%以及94.5%。该算法的优势也在机械臂路径规划中得到验证。
Concerning the slow convergence and low sampling efficiency of the traditional RRT algorithm in high-dimensional spaces and complex environments,this paper proposes an improved algorithm based on dynamic adaptive sampling and local optimization(DA-RRT),which adjusts the sampling density based on regional importance evaluation and environmental information and prioritizes dense sampling in narrow passages.DA-RRT is introduced into a local optimization mechanism further enhancing path quality.Multi-tree parallel growth is employed to accelerate the path search and optimization process.Experiments demonstrate that the improved DA-RRT algorithm,compared to the traditional RRT one,reduces path cost,runtime,number of nodes and iteration count by 22.0%,92.1%,43.7%,and 94.5%respectively,and its advantages have been verified in robotic arm path planning.
作者
李瑶
周刚
李捍东
LI Yao;ZHOU Gang;LI Handong(Tianfu New Area Power Supply Company,State Grid Sichuan Electric Power Company,Chengdu 610213,China;School of Electrical Engineering,Guizhou University,Guiyang 550025,China)
出处
《机械制造与自动化》
2025年第4期159-164,共6页
Machine Building & Automation
基金
国家自然科学基金项目(61663005)。
关键词
RRT算法
动态自适应
多树并行
局部优化
RRT algorithm
dynamic adaptive
multi-tree parallel
local optimization
