基于方向匹配的装配路径规划算法

Research on Path Planning in Assembly Planning

描述了装配规划中的最短无碰装配路径的自动生成问题。首先由零件在装配体中的几何约束生成局部拆卸方向,在考虑局部拆卸方向的前提下,运用位姿空间方法推导当前装配零件在固定姿态下由装配起点到装配最终位置的最短无碰路径。文中用具有位置分量和方向分量的特征元素描述物体,用方向分量相互匹配的特征元素对,即所谓方向匹配法计算三维离散物体的C-空间障碍。搜索空间用具有动态密度的栅格表达,以改进的具有目标可见性测试和变步长的A算法搜索最短无碰装配路径。该算法可处理复杂结构的装配体。

Introduces the automatic generation of assembly path. Firstly, the local disassembly direction is generated by the geometric constrains of the part in the assembly model. For the motion of the part at a fixed configuration from the start position to the goal position, the shortest collision free assembly path is produced by the local assembly direction and the configuration space. The objects are represented by the feature primitives with position and direction vectors, the C obstacles of 3D discrete objects can be computed by feature primitive pair with the mated direction vectors, named as direction mate method. A modified A * algorithm is employed to search the shortest collision free assembly path, the search algorithm has a goal visible test and dynamic search steps, the search space representation has the dynamic grid density. The algorithm can deal with the assembly object with complicated construct and has less computation efforts.