复杂结构条件下的装配路径求解与优化技术

Research on Assembly Path Planning and Optimization of Complex Structures

针对复杂结构条件下的零部件装配路径自动求解困难的问题,提出基于障碍和贪心规则的快速扩展随机树(Rapidly-exploring random tree,RRT)算法。该算法以基本RRT算法为基础,采用随机采样、终点采样、局部采样相结合的采样方式,利用目标零件与障碍物的碰撞面片法向量和碰撞点位置来引导随机树的扩展方向,在每个扩展方向上按贪心规则进行扩展,并提出先平移后旋转的扩展策略。对求解得到的初始装配路径,提出运用分段线性拟合的方法进行路径自动优化。设计并开发了装配路径求解软件原型系统,进行了算例测试和实例应用,结果验证了算法的高效可行。

In order to solve the difficult problem of assembly path planning of complex structures, an obstacle and greedy rule based rapidly-exploring random tree （OG-RRT） algorithm is proposed. The algorithm is based on basic RRT algorithm, adopts the combination of random sampling, goal sampling and local sampling, guides the extended direction of the random tree with the collision information （collision normal vector and collision point） of the target part and obstacles, extends nodes by greedy rule along each direction with translation-after-rotation expansion strategy. The piecewise linear approximation method is utilized to optimize the original assembly path. An assembly path planning prototype system is developed and some examples are tested to verify feasibility of the proposed algorithrn.