基于三维视觉伺服的智能装配系统设计

Three-dimensional Visual Servo Assemble Based on PID Control

在机械制造中,零件装配是生产整合的重要环节。目前装配任务主要依靠人工完成,效率低下,部分自动化装配平台虽然可通过机器人开环控制系统完成装配任务,但灵活性欠缺。为了提高柔性和生产效率,对装配进行研究后,以三角轴为对象,设计了三维视觉伺服装配平台,采用图像处理和PNP算法进行位姿估计,设计基于位置的视觉伺服控制器。实验结果表明,该平台在不同轴孔零件位姿的装配任务中,均能进行精确装配,满足轴孔装配的要求。

In mechanical manufacturing,part assembly is an important link of production integration.At present,most assembly tasks are mainly completed by manual labor,which is inefficient.Although some automatic assembly platforms can complete assembly tasks through the robot open-loop control system,they lack flexibility.In order to improve the flexibility and efficiency,in this paper,after the study of assembly,we designed a 3D visual servoing assembly platform for triangular shaft assembly,used the algorithm of image processing and PNP to pose estimation,and design the position based visual servo controller.The experimental results show that the platform can accurately assemble in the different posture of the shaft hole parts assembly task,and meet the requirements of shaft hole assembly.