虚拟机器手的研究开始于上个世纪90年代末期，是人工智能、机器入学、仿生学、控制理论和电子技术等多种技术学科交叉的产物。随着科技的发展和社会的进步，机器手已经走出实验室，在深海勘测、汽车制造、医学科研、能源开采、军事现代化等领域中都占有着其不可取代的位置。机器手的应用越来越广，越来越深入，各应用领域对机器手的要求也越来越高，因而虚拟机器手的研究逐渐成为趋势，成为高新技术应用领域的热点。 目前，对于虚拟机器手相关技术的研究，虽然已取得了大量的成果和突破性的进展，但是在很多关键理论和实际问题上，还有待解决和完善。因此，进行虚拟机器手相关技术的研究，实现虚拟机器手的全智能化具有非常重要的现实意义。
本文讨论的机器手环境为静态全局环境已知，通过栅格法对已知环境进行抽象，建立机器手工作空间模型，并采用虚拟现实技术，模拟虚拟觅食行为，根据优化条件搜索出一条从指定起点到终点的最优或近似最优路径，即全局路径规划。机器手利用自身视觉传感器按照规划出的最优路径自动导航，无碰撞地移动到目标点。
通过VC++和OpenGL库，Matlab平台实施仿真，实验结果表明，在环境地图已知的情况下，该算法能迅速规划出较优的全局路径，并且算法简单有效。与传统的搜索算法比较，它可以避免陷入过早收敛，能实现虚拟机器手在较短时间内搜索到最佳路径并规避障碍。
关键词：虚拟现实技术；组合优化；机器手；路径规划
Virtual robot began in the late 90s of the last century, artificial intelligence, machine school, bionics, control a variety of technical disciplines theory and electronic technology and other cross-product. With the advancement of technology and the development of society, the robot has been out of the lab, in the deep sea exploration, automobile manufacturing, medical research, energy, mining, military modernization and other areas are occupied with their irreplaceable position. Robot applications more widely, more deeply, the various applications of the robot are increasingly high requirements, thus Virtual robot gradually become a trend, it has become a hot high-tech applications. Currently, the study of virtual robot-related technologies, while there has been a lot of achievements and breakthroughs, but in many key theoretical and practical issues remain to be resolved and improved. Therefore, the study of virtual robot related technology, to achieve full of intelligent virtual robot has a very important practical significance.
This article discusses the robot environment is static global environment is known, by the grid method known abstract environment, the establishment of working space robot model, and the use of virtual reality technology, virtual simulation of foraging behavior, according to a search of the optimum conditions from Specifies the beginning to the end of the optimal or near-optimal path, namely global path planning. Robot using its own visual sensor according to plan the optimal path automatic navigation, collision-free move to the target point.
By VC ++ and OpenGL libraries, Matlab simulation platform for the implementation, results show that, in the environment map is known, the algorithm can quickly plan the optimum path to global, and the algorithm is simple and effective. Compared with traditional search algorithm, which avoids into premature convergence can be achieved virtual robot in a relatively short period of time to search for the best path and avoid obstacles.
Key words: virtual reality technology; combinatorial optimization; robot; path planning
