利用世界空间几何学信息的关节式机器人路径规划

Path planning for a manipulator using geometric information of a worldspace

针对关节式机器人提出了一种直接利用机器人世界空间几何学信息的有效的路径规划算法。该算法使用机器人周围的局部世界空间信息建立势函数，基于势函数的导数规划机器人路径。而且，针对常见的旋转关节机器人，分析了路径规划过程中的停留状态，提出了一种有效的消除势函数局部最小值的启发式方法。该算法避免了构形空间显式描述的复杂计算，在一般情况下，即使针对多自由度机器人，也能很快获得无冲突路径。最后，计算机仿真结果验证了该算法的有效性。

A useful planning method, which uses geometrical information of a worldspace directly to create a collision free path for a manipulator, is proposed. The algorithm uses local geometric information of a workspace of a manipulator to found an artificial potential function and plans the path of a manipulator based on the differentiation of an artificial potential function. To the common manipulator, dead lock case of the artificial potential function is analysed and the heuristic of escaping the local minimum value of the artificial potential function is proposed in this paper. The algorithm can avoid complicated calculations of explicit description of configuration space, which means that generally, a collision free path can be created quickly even for a multi degree of freedom manipulator. Finally, simulation results show effectiveness of the proposed method.