论文部分内容阅读
本文实现了一个基本的机器人规划系统。它能自动生成一系列避免与障碍物发生碰撞的机器人动作轨迹。其视觉子系统可获取环境知识。算法对可能引起碰撞的障碍物进行从直角坐标空间到机器人关节坐标空间的转换。并采用分级式方式对自由空间进行先“粗”后“细”的两级描述。合理的数据结构既减少了存储量又给出了充分的启发信息。在此基础上又采用了先“全局”后“局部”的两级路径优化,从而能以较快速度决策出一能避免碰撞且运行时间短的优化路径。实验表明,该系统可在一般 PC 机上实现。
This article realizes a basic robot planning system. It automatically generates a series of robot trajectories that avoid collisions with obstacles. Its vision subsystem acquires environmental knowledge. The algorithm transforms the obstacles that may cause collision into the joint space of the robot from Cartesian coordinate space. And adopt a hierarchical way to describe the free space first after “coarse” and “fine”. Reasonable data structure not only reduces the amount of storage but also gives ample inspiration. On this basis, the two-level path optimization of “global” and “local” is adopted, so that an optimized path that avoids collisions and has a short running time can be quickly decided. Experiments show that the system can be realized in the general PC.