基于VRML节点树的双臂移动机器人碰撞检测及优化  

Collision Detection and Optimization of Dual-Arm Mobile Robot Based on Node Tree of VRML

作　　者：刘忠[1] 曹其新[1] 朱笑笑[1] 王雯珊[1] 

机构地区：[1]上海交通大学机械系统与振动国家重点实验室,上海200240

出　　处：《上海交通大学学报》2011年第7期985-989,共5页Journal of Shanghai Jiaotong University

基　　金：机械系统与振动国家重点实验室重点项目(MSV-MS-2010-01);教育部博士点基金资助项目(20090073110037);教育部重大项目培育资金项目(708035)

摘　　要：针对双臂移动机器人三维仿真中存在大量非凸模型而无法精确进行实时碰撞检测,提出了基于虚拟现实建模语言(VRML)节点树实现凸分解的方法.利用分解得到的单位形体的顶点数据,使用Quickhull算法构造应用于V-Clip算法的凸多面体,实现虚拟环境中模型与机器人之间的精确碰撞检测.为了提高碰撞检测的效率,优化仿真过程中碰撞检测的实时性,采用了包围盒技术.实验结果表明:该方法具有良好的实时性和较高的碰撞检测精度.A large number of non-convex models exist in three-dimensional simulation environment for dual-arm mobile robots,it is impossible to realize accurate real-time collision detection.This paper proposed a method to implement convex decomposition of non-convex models based on node tree of VRML.Pass vertexes returned by convex decomposition to Quickhull algorithm to create convex polyhedron that used for V-Clip collision detection algorithm to realize accurate real-time collision detection.In order to improve the efficiency of collision detection to enhance real-time simulation,bounding box is used.The results show that this method can make real-time collision detection and has great precision.

关 键 词：双臂移动机器人 碰撞检测 虚拟现实建模语言 V-Clip算法 包围盒 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]