CN-122007747-A - Teaching-free welding method and system for thick-wall pipe groove

CN122007747ACN 122007747 ACN122007747 ACN 122007747ACN-122007747-A

Abstract

The invention relates to the technical field of industrial welding robots and automatic welding control, in particular to a teaching-free welding method and a teaching-free welding system for thick-wall pipe grooves. The method comprises the steps of obtaining a global visual unit and an assembled thick-wall pipe groove area, obtaining an actual pose of a workpiece and a geometric central line vector of a welding seam through partition scanning, point cloud registration and model registration, extracting geometric characteristic parameters of the cross section of the welding seam, matching groove form codes and binding preset welding layer channel information to form a welding seam attribute set, performing path planning, generating a multi-layer multi-channel decomposition and welding program to obtain a welding program and a theoretical track, and finally correcting subsequent interpolation points on line, scanning the welding seam again under a triggering condition, and updating the welding seam attribute set. According to the invention, by constructing the weld joint attribute set capable of driving multi-layer and multi-channel welding, the variable cross-section welding problem caused by uneven assembly of the thick-wall pipe is effectively solved, and the welding self-adaption capability and the welding quality are obviously improved.

Inventors

DU ZHUO
GAO SHENG
KONG ZHENGWEI
ZHANG ENMING
YANG YINGLI

Assignees

东北石油大学

Dates

Publication Date: 20260512
Application Date: 20260408

Claims (10)

1. The teaching-free welding method for the groove of the thick-wall pipe is characterized by comprising the following steps of: S100, acquiring an overall vision unit, an industrial camera, laser line structured light and an assembled thick-wall pipe groove area, and performing partition scanning, point cloud registration processing and model and point cloud registration processing to obtain actual pose of a workpiece and geometric central line vectors of a welding seam, wherein the overall vision unit comprises the industrial camera, the laser line structured light, a mounting bracket, a control interface and a collection position which is separated from an automatic welding module, the industrial camera is used for receiving reflected light information of the surface of the thick-wall pipe groove area, the laser line structured light is used for forming a continuous light band profile on the surface of the thick-wall pipe groove area, and the assembled thick-wall pipe groove area comprises a groove opening part to be welded, base metal surfaces at two sides of the groove and a positioning reference area adjacent to the groove; s200, based on the actual pose of the workpiece and the geometric central line vector of the welding seam, extracting geometric characteristic parameters of the cross section of the welding seam, matching a groove form code and binding preset welding layer channel information to obtain a welding seam attribute set; s300, carrying out path planning algorithm, multi-layer multi-channel decomposition and welding program generation processing based on the weld attribute set to obtain a welding program and a theoretical track; S400, based on the welding program and the theoretical track, performing automatic welding module welding operation, online correcting subsequent interpolation points and rescanning welding seams to obtain updated welding seam attribute sets.
2. The method of claim 1, wherein the process of the partition scan process comprises: dividing the groove area of the thick-wall pipe into a plurality of adjacent scanning sections according to the length of the groove area and the view field range, respectively acquiring three-dimensional contours and RGB images for each scanning section, and processing the three-dimensional contours and RGB images through a splicing algorithm to form continuous weld joint area point clouds.
3. The method of claim 2, wherein the process of point cloud registration and model-to-point cloud registration comprises: performing downsampling on the weld region point cloud to unify spacing, filtering to screen out discrete noise points, and edge extraction based on the RGB image to determine groove boundaries; and carrying out transformation matrix calculation and model and point cloud registration to obtain the actual pose of the workpiece and the geometric center line vector of the welding seam continuously extracted along the opening area of the groove.
4. A method according to claim 3, wherein the process of weld cross-section geometric feature parameter extraction processing comprises: a plurality of local sections are cut along the geometric center line vector of the welding seam according to preset intervals; extracting the root gap width, the groove angle, the blunt edge thickness and the base material thickness of each section; And performing section merging according to the parameter proximity degree to obtain one or more parameter sections.
5. The method of claim 1, wherein the process of groove form code matching and pre-set weld layer information binding process comprises: the matching processing of the groove form codes comprises the steps of executing matching of four types of parameters of each parameter section with a process database, and generating groove form codes and process parameter index codes of each section; The preset welding layer channel information binding processing comprises the step of carrying out section-level binding on groove form codes, process parameter index codes, preset welding layer channel information and a welding seam geometric center line vector of the same parameter section, and forming a layer channel record chain as a welding seam attribute set, wherein when a certain section is not matched with the preset welding layer channel information, adjacent section information is copied to form temporary binding or marked as a section to be updated.
6. The method of claim 5, wherein the path planning algorithm processing comprises: and reading starting point coordinates, end point coordinates, geometric center line vectors of the welding seam, groove form codes, process parameter index codes and preset welding layer channel information of each section according to the section identification sequence in the welding seam attribute set, and executing four-dimensional coupling operation of layer thickness, width, speed and current in each section to generate an automatic welding track changing along with the sections.
7. The method of claim 6, wherein the multi-layer multi-pass decomposition and welding procedure generation process comprises: decomposing the automatic welding track into independent welding path tracks according to sections, layers and channels, and generating corresponding multiaxial motion interpolation points; the welding program generation process comprises the step of binding the multi-axis motion interpolation points with the welding current, the arc voltage, the welding speed and the swing amplitude corresponding to the technological parameter index codes point by point to generate a welding program, and synchronously generating theoretical tracks corresponding to each layer of channels.
8. The method of claim 7, wherein the process of the automated welding module welding job process comprises: and calling a visual guiding module according to a welding program to synchronously acquire welding seam error data, and forming continuous records of the current position deviation, the direction deviation and the welding bead boundary deviation.
9. The method of claim 8, wherein the process of online correcting subsequent interpolation points and rescanning the weld includes: the online subsequent interpolation point correction processing comprises the steps of performing position feedback and feedforward compensation based on weld error data, and generating corrected subsequent interpolation points; When the triggering conditions of the current layer welding completion, the current channel welding completion, the existence of the section mark to be updated or the continuous abnormal acquisition mark reaching the preset times are met, the global visual unit or the visual guiding module is called to rescan the current section, the geometric characteristic parameters of the welding line cross section of the current section are updated, the groove form code and the process parameter index code are regenerated, the groove form code and the process parameter index code are rebind with preset welding layer channel information, an updated welding line attribute set is obtained, and the updated welding line attribute set is returned to the path planning algorithm for processing.
10. A teaching-free welding system for thick-wall pipe grooves is characterized by comprising a panoramic scanning module, a point cloud registration and coordinate conversion module, a cross section parameter extraction module, an attribute binding module, a path planning and interpolation generation module, a welding program generation module, a welding operation and error acquisition module and an online correction and attribute updating module, wherein the modules are sequentially connected and used for realizing the method according to any one of claims 1-9.

Description

Teaching-free welding method and system for thick-wall pipe groove Technical Field The invention relates to the technical field of industrial welding robots and automatic welding control, in particular to a teaching-free welding method and a teaching-free welding system for thick-wall pipe grooves. Background In the technical field of industrial welding robots and automatic welding control, the existing scheme of the thick-wall pipe groove area generally adopts the technical routes of manual teaching programming, model library calling, welding line area point cloud acquisition, groove center line fitting, automatic welding track generation and welding line error data tracking, and has the limitations that the corresponding relation between the welding line area point cloud and the actual pose of a workpiece is unstable, the geometric characteristic parameters of the welding line cross section are disjointed with preset welding layer channel information, the welding line error data are difficult to write back to a subsequent track processing link and the like. In the existing method, the welding procedure and theoretical track generation are finished by depending on a three-dimensional model of a workpiece, a procedure template or a single scanning result, the root gap width, the groove angle, the blunt edge thickness and the base material thickness are inconsistent with the actual working condition easily to occur in the groove area of the assembled thick-wall pipe, and the conditions of multi-layer multi-channel decomposition and unmatched process parameter index code call are difficult to meet the continuous realization of multi-layer multi-channel welding. Aiming at joint processing of geometric characteristic parameters of a welding line cross section, groove form codes, process parameter index codes, preset welding layer channel information and updating of a welding line attribute set, the prior art generally lacks a unified link between point cloud registration, characteristic extraction algorithm processing, attribute binding, path planning algorithm processing, on-line correction of subsequent interpolation points and rescanning of the welding line, and is difficult to form a consistent flow of point cloud acquisition of a welding line area, actual pose determination of a workpiece, construction of the welding line attribute set, automatic welding track generation, welding program and theoretical track generation, welding line error data acquisition and updating of the welding line attribute set in a thick-wall pipe groove area, so that subsequent interpolation points are adjusted and disjointed with the updating of the welding line attribute set in a multi-layer multi-channel welding process, and continuous connection among welding operation, process switching and subsequent updating is difficult. Disclosure of Invention In order to solve the technical problems, the invention provides a teaching-free welding method for a groove of a thick-wall pipe, which comprises the following steps: S100, acquiring an overall vision unit, an industrial camera, laser line structured light and an assembled thick-wall pipe groove area, and performing partition scanning, point cloud registration and model and point cloud registration processing to obtain actual pose of a workpiece and geometric central line vectors of a welding seam, wherein the overall vision unit comprises the industrial camera, the laser line structured light, a mounting bracket, a control interface and a collection position which is separated from an automatic welding module, the industrial camera is used for receiving reflected light information of the surface of the thick-wall pipe groove area, the laser line structured light is used for forming a continuous light band profile on the surface of the thick-wall pipe groove area, and the assembled thick-wall pipe groove area comprises a groove opening part to be welded, base metal surfaces at two sides of a groove and a positioning reference area adjacent to the groove; s200, based on the actual pose of the workpiece and the geometric central line vector of the welding seam, extracting geometric characteristic parameters of the cross section of the welding seam, matching a groove form code and binding preset welding layer channel information to obtain a welding seam attribute set; s300, carrying out path planning algorithm, multi-layer multi-channel decomposition and welding program generation processing based on the weld attribute set to obtain a welding program and a theoretical track; S400, based on the welding program and the theoretical track, performing automatic welding module welding operation, online correcting subsequent interpolation points and rescanning welding seams to obtain updated welding seam attribute sets. Further, the process of the partition scanning process includes: dividing the groove area of the thick-wall pipe into a plurality of adjacent scanning sections according to