欠信息正交视觉装配机器人机械系统研究

Research on the Mechanical System of Orthogonal Visual Assembly Robot with Less Information

为解决机器视觉在零件的自动装配领域三维重构精度不高且计算量大的问题,提出了一种基于正交视觉装配机器人系统。运用了两个工业相机构成正交视觉,将动态视觉问题转化为静态问题,将立体视觉转化为平面视觉;采用三个平移滑台和三个转动自由度的机械臂构成装配机器人的主体,达到了降低装配机器人的控制难度和提高装配机器人的空间到达能力的目的;建立了正交双目视觉模型,演算了零件在两个正交下获取其空间位置和姿态的基本原理;最后通过对执行件末端定位精度的柔性仿真,验证了装配机器人的机械系统的可行性。

In order to solve the problem that the machine vision is not accurate and with large calculation amount in the three-dimensional reconstruction of the automatic assembly field of parts,a robot system based on the orthogonal vision assembly is proposed. Two industrial cameras are used to form the orthogonal vision,which transforms the dynamic visual problems into static problems and transforms stereoscopic vision into plane vision. The main body of the assembly robot is made up of three translational sliding tables and three mechanical arms with rotational degrees of freedom,which achieves the goal of reducing the control difficulty and improving the space arrival ability of the assembly robot. The orthogonal binocular vision model is established,and the basic principle of obtaining the space position and attitude of the part under two orthogonal conditions is deduced. Finally,the feasibility of mechanical system of the assembly robot is validated by flexible simulation of the end positioning accuracy of the actuator.