基于视觉传感器焊缝跟踪轨迹自示教方法研究

Investigation on Seam Tacking Path Self-teaching Method Based on Visual Sensor

介绍了一种基于视觉传感器在焊前采集焊缝信息,动态存储相关数据,从而记住从初始状态到目标状态的控制路径.施焊时,调用轨迹记录数据库,按照行走路径进行自动跟踪,从而达到自动焊接.进一步介绍了自示教的设计步骤及实现过程,并设计了图像处理流程和系统的监控界面.试验结果表明:该系统无需复杂的几何建模、仿真、示教编程,方法简单可行,无需考虑电弧、飞溅和烟尘等信号的干扰,提高了图像处理速度,从而可细化路径,提高焊缝跟踪精确度.

A seam tracking system based on visual sensor is developed in this study. The processing of the selfteaching system is as follows： before welding, CCD visual sensor detects the seam image, database dynamically stores coherent data, thereby remembering the control route from initial state to goal state; During welding course, the system calls the database and arrives at automatic welding according to the preceding route. The procedure and realization of the self-teaching system is introduced in detail. The procedure of image processing and the monitor and control interface are designed. Experiment results show： the system adopts a kind of simplicity practicable means without complicated geometrical modelling, simulation, programming, and need not considering the interferences such as arc light, spatter and fume and so on. It simplifies the mechanical structure and enhances image-processing speed. So it refines route and improves tracking accuracy.