A sampling-based planning algorithm is one of the most powerful tools for collision avoidance in the motion planning of manipulators. However, this algorithm takes a long time to generate motions of the manipulator. This work proposes a goal-oriented (GO) sampling method for the motion planning of a manipulator. The GO sampling method can identify the initial solution in a shorter time than other sampling-based algorithms, leading to significant improvement in computational efficiency. Based on the GO sampling method, cases involving configuration space and collision checking are implemented based on the proposed equations in the planning of manipulator motion. Different combinations of configuration space settings are mainly analyzed and compared through experiments using a six-degree-of-freedom manipulator.

中文翻译：

---

基于采样的机械手基于采样的运动计划

基于采样的规划算法是机械手运动规划中避免碰撞的最强大工具之一。但是，该算法需要很长时间才能生成操纵器的运动。这项工作提出了一种针对目标的（GO）采样方法，用于机械手的运动计划。与其他基于采样的算法相比，GO采样方法可以在更短的时间内识别出初始解，从而显着提高了计算效率。基于GO采样方法，在机器人运动计划中，基于提出的方程，实现了涉及配置空间和碰撞检查的情况。主要通过使用六自由度操纵器的实验来分析和比较配置空间设置的不同组合。