作　　者： ; ; ; ; ;

机构地区： 深圳信息职业技术学院

出　　处： 《深圳信息职业技术学院学报》 2012年第3期20-27,共8页

摘　　要： 针对机器人避碰运动规划问题,提出了基本RRT的改进算法。首先,研究了基本RRT算法思想,进而融合J-空间采样和C-空间建模技术,提出了提升性能稳定性和收敛速度的改进RRT算法;其次,利用分离轴原理,研究了基于OBB的碰撞检测算法,并将其成功集成到改进RRT判定C-空间的机器人与障碍物是否碰撞;最后,设计和实现了三维仿真系统软件框架,通过计算仿真验证了本文提出算法的有效性和稳定性。 For collision avoidance motion planning problem for robot, improved algorithm of basic RRT （Rapidly Random-Exploring Trees） has been proposed. First, basic RRT algorithm has been researched, and thus by integration of the joint space（J-Space） sampling and the configuration space（C-Space） modeling techniques, improved RRT algorithm has been put forward in order to enhance performance stability and convergence speed. Second, collision detection algorithm based OBB（Oriented Bounding Box） has been studied by use of separating axis theorem, and has been successfully integrated into improved RRT to determine whether robot and obstacles collide in C-Space or not. Finally, three-dimensional simulation software framework has been designed and implemented, effectiveness and stability of the proposed algorithm has been verified by computer simulation.

关 键 词： 运动规划 快速扩展随机树 机器人 避碰 计算机仿真

领　　域： [自动化与计算机技术] [自动化与计算机技术]