以关节型机械臂避开垂直茎秆或撑杆采摘番茄为研究对象,提出了一种基于构形空间的关节型机械臂避障路径规划方法。利用空间映射原理,将关节型机械臂工作空间的三维避障问题转换为平面R-R机械臂避开障碍圆的问题,用临界碰撞关节角建立C-障碍空间的映射计算模型,将工作空间的位置避障转换为构形空间连杆关节角的计算。以能量最优函数优选避障规划的关节终点角度,利用A*算法计算平面R-R机械臂的避障关节角路径,获得一系列表示空间连杆位置的相交竖直面,并在竖直面内进行其余关节角的规划。避障采摘番茄的试验表明,机械臂带动夹持器能成功绕过直线状障碍物,引导夹持器到达果实目标位置,证明对平面R-R机械臂的空间映射建模是正确的,验证了提出的避障路径规划方法是可行的,可以应用于番茄的自动收获。
It is chosen as research object that joint robot arm avoids erect stem or strut to pick tomatoes, and a kind of obstacle-avoidance path planning method based on configuration space of joint robot arm was presented. Space mapping principle was used to transform obstacle-avoidance problem of joint robot arm in 3-D workspace into avoiding obstacle circle problem of planar R - R robot arm. Critical collision joint angle was used to build up mapping computation models of configuration obstacle space of planar R - R robot arm, and obstacle-avoidance position computation of joint robot arm in 3-D workspace was transformed into joint angles computation of linkages in configuration space. After terminal joint angles of obstacle-avoidance path planning were optimized according to optimal energy function, the obstacle- avoidance joint angle path of planar R - R robot arm based on A* arithmetic was computed. A series of intersectant erect planes described space link position were obtained and the planning of other joint was carried out in the intersectant erect planes. The test of avoiding erect obstacle to pick tomatoes showed that joint robot arm carried with the gripper can move round linear obstacle and guide the gripper to reach object position. The test proved that mapping computation models of configuration obstacle space of planar R-R robot arm were correct. It validated that the obstacle-avoidance path planning method was feasible and may be applied into automatic harvesting of tomato.