With a view to ensuring the welding quality of crude oil storage tanks,having a visual correc-tion based to study the welding robot for floating ship plates of crude oil storage tanks was implemented.Firstly,having the welding robot vision system built and calibrated and both welding seam image processing and feature point extraction methods designed;then,having a 3D welding robot model,a welding gun mod-el and a simulation model of a structured light visual sensor established,including having an iterative near-est point registration algorithm based to implement welding visual correction so as to obtain the real welding seam position;finally,having an off-line correction experiment conducted on the weld trajectory.The results show that,the off-line welding deviation correction can solve the problem that the welding robot doesn't accurately work in a traditional teaching mode because of the workpiece's inaccurate positioning of the floating ship plate.
维普
万方数据