中文摘要：

现有的基于C空间的无碰撞轨迹规划算法需要求解C空间以获得c空间障碍边界。对于多自由度的冗余机械臂，求解过程需要消耗很大的计算量和内存，不适用于计算资源紧张的空间机械臂轨迹的快速规划。文章提出了一种不需要求解C空间的试探性规划算法，包含4个子算法：碰撞检测算法、无碰撞目标构型求解算法、无碰撞路径搜索算法和路径平滑算法。已知期望的末端作用器位置和姿态，利用目标构型求解算法得到无碰撞的目标构型，然后由路径搜索算法在C空间障碍边界未知的情况下，利用碰撞检测算法，采用一定的试探规律，在C空间中搜索出一条无碰撞路径，最终由路径平滑算法使该路径平滑，易于实现。仿真算例表明，该算法是快速有效的，适用于冗余空间机械臂粗捕获段的快速轨迹规划。

英文摘要：

The existing collision-free motion planning algorithms based on configuration space （C-Space） need to find the boundaries of C-space obstacles by solving C-space, which will cost much computation and memories for redundant manipulators with larger degrees of freedom （DOF）. A novel algorithm was proposed to find the collision-free path in C-space by testing motion directions without solving C-space. This algorithm contained four sub-algorithms： the collision detection algorithm, the collision-free target configuration solver, the collision-free path-searching algorithm and the path-smoothing algorithm. Given the desired position and attitude of the end-effector, the target configuration solver calculated a collision-free target configuration. Then the path-searching algorithm found a collision-free path with C-space obstacle boundaries unknown. Finally, the path-smoothing algorithm smoothed the obtained path. To verify the effectiveness of the algorithm, two numerical examples were carried out. And the results indicate that the algorithm is efficient and can be used in fast motion planning for redundant manipulators in space applications.