Based On The Visual Sensing V Groove Weld Automatic Tracking System Research

With the development of science and technology, welding technology is developing toward the direction of the automatic and intelligent. Welding robot is emeged in order to satisfy the actual needs. Based on a v-shaped groove weld, in order to realize goal of the automatic identification and automatic tracking of weld, we proposed the algorithms which can identify the weld seam and image processing autonomous ly, established a image processing system based on DSP and a seam tracking servo system based on electric sliding platform and motion control card.The whole tracking system is composed of three subsystems : image acquisition, image processing and weld seam automatic tracking. Image acquisition system is mainly composed of a CCD camera and a laser transmitter component. Seam image processing system is mainly based on DSP development board which uses UCO TMS320DM642 as the core. Seam tracking system has been developed on the core of PCI-1245 controller card made by Advantech.The topic mainly focused on a lgorithm design of seam tracking system and system software and hardware design. Algorithm design included image processing algorithms and weld seam tracking control algorithm two parts. In the algorithm of weld image, weld filtering was dealed with median filtering algorithm, image edge detection used binary morphological edge detection method, and image feature information extraction used the detection method combining slope analysis and least squares method. On the part of racking control algorithm, by studying and combinating traditional classical PID control and fuzzy control, a control algorithm—dual- mode control algorithm segmented control was proposed.In the hardware development platform, the algorithms used in the paper were verified by experiments resulted in good results, which demonstrated the feasibility of the system.