针对动态障碍物的7R机械臂实时避障方法
Real-time Avoidance Method of 7R Manipulator for Dynamic Obstacle
-
摘要: 为解决在有动态障碍物环境下的避障问题, 针对七自由度冗余机械臂提出了一种实时避障方法, 该方法将传统复杂的机械臂避障规划转化为对单根连杆进行避障引导的运动规划. 通过连杆与障碍物的位置关系划分主次危险杆件, 并对杆件进行实时的避障规划, 引导机械臂在快速避开障碍物的同时向目标点运动. 在避障完成后, 基于自运动的逆解算法, 构建了安全距离势能指标确定自运动解集中的最优位形解并进行达点运动. 结果表明:该算法在动静态环境下进行达点运动时皆能及时有效地避开障碍物, 且避障过程中平均单步迭代运算时间约25 ms, 远小于传统避障算法, 验证了算法的实时性和有效性.Abstract: A real-time collision avoidance algorithm for 7-DOF redundant manipulator was proposed to solve the problem of avoidance in a dynamic obstacle environment. In this paper, the traditional complex obstacle avoidance planning of the manipulator was transformed into the motion planning of obstacle avoidance guidance for a single rod. First, the location relationship between the rod and the obstacle was used to determine the main and secondary dangerous rods. Then, real-time avoidance plan was implemented to guide the manipulator to move towards the target point while avoiding obstacles quickly. After obstacle avoidance was completed, on the basis of the self-motion inverse solution, the optimal configuration solution of the self-motion solution set was determined by the potential index of safe distance, and the reaching point movement was carried out. Results show that the new algorithm can avoid obstacles in time in both dynamic and static environments. In the process of obstacle avoidance, the average single-step iterative operation time is 25 ms, which is far less than the traditional obstacle avoidance algorithm and verifies the real time and effectiveness of the proposed algorithm.