CN-121373703-B - Visual positioning method and device for projection welding of robot for welding vehicle body, electronic equipment and computer readable storage medium

CN121373703BCN 121373703 BCN121373703 BCN 121373703BCN-121373703-B

Abstract

The application provides a robot projection welding visual positioning method and device for welding a vehicle body, electronic equipment and a computer readable storage medium, and relates to the field of robot visual positioning. The method comprises the steps of constructing priori positions and geometric constraints of welding spots based on vehicle body design information data and tool information data, constructing a three-dimensional reference model based on the priori positions and the geometric constraints, outputting a geometric mapping relation, extracting a welding spot positioning feature set in a welding spot image, performing feature constraint association on the welding spot positioning feature set and the geometric mapping relation, outputting a first space pose of the welding spot in a three-dimensional space through the three-dimensional reference model, performing precision verification operation on the first space pose and correcting the first space pose to a second space position, converting the second space pose to a robot coordinate system, and controlling a target robot to perform projection welding visual positioning. The application solves the problem of lower positioning precision of the traditional robot projection welding visual positioning method for welding the vehicle body under the complex working condition.

Inventors

WANG LEI
MEI ZHIYUAN
WANG JIAWEI
ZHANG MIN
WANG HAOYU

Assignees

安徽德恒工业智能科技有限公司

Dates

Publication Date: 20260512
Application Date: 20251013

Claims (9)

1. A robotic projection welding visual positioning method for vehicle body welding, the method comprising: Acquiring vehicle body design information data and tooling information data of a vehicle to be welded, and constructing priori positions and geometric constraints of welding spots based on the vehicle body design information data and the tooling information data; Constructing a three-dimensional reference model based on the priori position and the geometric constraint, and outputting a geometric mapping relation, wherein the geometric mapping relation comprises a spatial relation among a target robot, a target camera and the welding spot; acquiring a welding spot image based on the target camera, and extracting a welding spot positioning feature set in the welding spot image; Performing feature constraint association on the welding spot positioning feature set and the geometric mapping relation, and outputting a first space pose of the welding spot in a three-dimensional space through the three-dimensional reference model; The method comprises the steps of carrying out precision verification operation on a first space pose and correcting the first space pose into a second space position based on welding disturbance factors, and concretely comprises the steps of calculating first welding precision corresponding to welding spots based on the first space pose, outputting the welding disturbance factors through a welding disturbance recognition model if the first welding precision is determined to be smaller than preset precision, outputting a welding spot correction strategy corresponding to the welding disturbance factors through the welding disturbance recognition model based on the welding disturbance factors, correcting the first space pose based on the welding spot correction strategy to obtain a second space pose, calculating second welding precision corresponding to welding spots based on the second space pose, and completing the precision verification operation when the second welding precision is determined to be larger than or equal to the preset precision; And converting the second space pose into a robot coordinate system, and controlling the target robot to finish projection welding visual positioning through the converted second space pose.
2. The method according to claim 1, wherein the obtaining the body design information data and the tooling information data of the vehicle to be welded and constructing the prior position and the geometric constraint of the welding spot based on the body design information data and the tooling information data specifically comprises: Extracting a vehicle body structure model based on the vehicle body design information data, and marking nominal space coordinates and local normal directions corresponding to the welding spots in the vehicle body structure model; taking the nominal spatial coordinates and the local normal direction as the prior locations; determining limiting conditions of a clamp, a jig and a process reference surface based on the tool information data, and correcting the prior position under the limiting conditions to obtain a corrected position conforming to the constraint; and combining the correction position with the clamping direction, the normal direction of the reference plane and the adjacent relation in the tool information data to generate the geometric constraint.
3. The method according to claim 1, wherein the constructing a three-dimensional reference model based on the prior position and the geometric constraint and outputting a geometric mapping relation specifically comprises: Combining the prior position and the geometric constraint through space fusion operation, and establishing three-dimensional space distribution of the welding spots in a vehicle body coordinate system according to a combination result; and constructing the three-dimensional reference model based on the three-dimensional space distribution to output the geometric mapping relation.
4. The method of claim 1, wherein the acquiring a solder joint image based on the target camera and extracting a set of solder joint locating features in the solder joint image specifically comprises: carrying out reflection inhibition and multi-exposure fusion treatment on the welding spot image so as to enhance the imaging quality of the area where the welding spot is positioned; extracting edge contour features and local geometric features of the welding spots from the reinforced welding spot image, and carrying out sub-pixel refinement treatment on the local geometric features; And combining the edge contour features with the local geometric features subjected to refinement through feature fusion operation, and outputting a welding spot positioning feature set, wherein the welding spot positioning feature set comprises shape features, position features and normal features.
5. The method of claim 4, wherein the feature constraint correlating the set of weld spot positioning features with the geometric mapping relationship specifically comprises: Matching the shape features in the welding spot positioning feature set with the nominal geometric constraints in the geometric mapping relation to establish the geometric corresponding relation of the welding spots; registering based on the position features in the welding spot positioning feature set and the priori positions in the geometric mapping relation to establish a spatial correspondence of welding spots; consistency constraint is carried out on the basis of normal features in the welding spot positioning feature set and the local normal direction in the geometric mapping relation so as to establish a posture corresponding relation of welding spots; and forming characteristic constraint association through the geometric corresponding relation, the spatial corresponding relation and the gesture corresponding relation.
6. The method of claim 1, wherein constructing the weld disturbance recognition model specifically comprises: acquiring product line environment monitoring information, wherein the product line environment monitoring information comprises an environment illumination parameter, a vehicle body assembly deviation parameter, a robot operation thermal drift parameter and a camera imaging stability parameter; Taking the historical welding spot positioning feature set, the geometric mapping relation and the product line environment monitoring information as training data, and constructing welding disturbance features, wherein the welding disturbance features comprise feature residual error distribution features, pose drift quantization index features, assembly tolerance deviation features and imaging deviation features; constructing a first corresponding relation between the welding disturbance characteristics and the welding disturbance factors; Based on the welding disturbance factors, constructing a welding spot correction strategy through a self-adaptive correction generation mechanism, wherein the welding spot correction strategy comprises an imaging parameter compensation strategy, a pose optimization compensation strategy, a track constraint correction strategy and a calibration drift compensation strategy; Constructing a second corresponding relation between the welding disturbance factors and the welding spot correction strategy; and constructing the welding disturbance recognition model based on the first corresponding relation and the second corresponding relation.
7. A visual positioning device for projection welding of a robot for welding a vehicle body is characterized by comprising an acquisition module and an output module, wherein, The acquisition module is used for acquiring vehicle body design information data and tooling information data of a vehicle to be welded, constructing priori positions and geometric constraints of welding spots based on the vehicle body design information data and the tooling information data, constructing a three-dimensional reference model based on the priori positions and the geometric constraints, and outputting a geometric mapping relation, wherein the geometric mapping relation comprises a spatial relation among a target robot, a target camera and the welding spots; The output module is used for carrying out characteristic constraint association on the welding spot positioning characteristic set and the geometric mapping relation, outputting a first space pose of the welding spot in a three-dimensional space through the three-dimensional reference model, carrying out precision verification operation on the first space pose, correcting the first space pose into a second space position based on a welding disturbance factor, specifically, calculating first welding precision corresponding to the welding spot based on the first space pose, outputting the welding disturbance factor through a welding disturbance recognition model if the first welding precision is determined to be smaller than preset precision, outputting a welding spot correction strategy corresponding to the welding disturbance factor through the welding disturbance recognition model based on the welding disturbance factor, correcting the first space pose based on the welding spot correction strategy to obtain a second space pose, calculating second welding precision corresponding to the welding spot based on the second space pose, completing the precision verification operation when the second welding precision is determined to be larger than or equal to the preset precision, converting the second welding precision to a second coordinate system, and then completing visual machine vision of a welding target by a machine.
8. An electronic device comprising a processor, a communication bus, a user interface, a network interface, and a memory, the memory for storing instructions, the user interface and the network interface for communicating to other devices, the processor for executing instructions stored in the memory to cause the electronic device to perform the method of any of claims 1-6.
9. A computer readable storage medium storing instructions which, when executed, perform the method of any one of claims 1 to 6.

Description

Visual positioning method and device for projection welding of robot for welding vehicle body, electronic equipment and computer readable storage medium Technical Field The application relates to the field of robot vision positioning, in particular to a robot projection welding vision positioning method and device for welding a vehicle body, electronic equipment and a computer readable storage medium. Background In the context of intelligent manufacturing and continuous upgrades in the automotive industry, vehicle body welding automation is evolving toward high precision, flexibility, and intelligence. With the increasing complexity of the vehicle body structure and the wide application of lightweight materials, the conventional welding mode relying on a fixed tool and a rigid path has been difficult to meet the production requirements. The traditional method generally takes a tool reference point and fixed compensation parameters as cores to realize the visual positioning of the projection welding robot for welding the vehicle body. But in the real scene of the car body welding production line, different types of positioning errors can be generated due to various interference factors. For example, the edge of a welding spot image is blurred due to strong reflection of the outer plate of the automobile body and local high light spots, so that feature extraction is unstable, hand-eye calibration accuracy is reduced due to thermal drift and joint clearance accumulation of a robot in a long-term operation process, so that pose solving deviation is caused, candidate areas are overlapped due to dense distribution of welding spots and surface scratch texture, so that false detection or omission detection is caused, and the like. At this time, only by the method, errors generated by the factors cannot be corrected, so that the overall accuracy of the visual positioning of the projection welding robot is reduced. Accordingly, there is a need for a robotic projection welding vision positioning method, apparatus, electronic device, and computer readable storage medium for vehicle body welding. Disclosure of Invention The application provides a robot projection welding visual positioning method and device for welding a vehicle body, electronic equipment and a computer readable storage medium, which solve the problem that the traditional robot projection welding visual positioning method for welding the vehicle body has lower positioning precision under complex working conditions. The application provides a robot projection welding visual positioning method for welding a vehicle body, which comprises the steps of obtaining vehicle body design information data and tool information data of the vehicle to be welded, constructing priori positions and geometric constraints of welding spots based on the vehicle body design information data and the tool information data, constructing a three-dimensional reference model based on the priori positions and the geometric constraints, outputting geometric mapping relations including space relations among a target robot, a target camera and the welding spots, obtaining welding spot images based on the target camera, extracting a welding spot positioning feature set in the welding spot images, performing feature constraint association on the welding spot positioning feature set and the geometric mapping relations, outputting a first space pose of the welding spots in a three-dimensional space through the three-dimensional reference model, performing precision verification operation on the first space pose, correcting the first space pose to a second space position based on welding disturbance factors, converting the second space pose to a robot coordinate system, and controlling the target robot to finish projection welding visual positioning through the converted second space pose. Optionally, acquiring vehicle body design information data and tooling information data of a vehicle to be welded, and constructing priori positions and geometric constraints of welding spots based on the vehicle body design information data and the tooling information data, wherein the method specifically comprises the steps of extracting a vehicle body structure model based on the vehicle body design information data, and marking nominal space coordinates and local normal directions corresponding to the welding spots in the vehicle body structure model; the method comprises the steps of taking a nominal space coordinate and a local normal direction as prior positions, determining limiting conditions of a clamp, a jig and a process reference surface based on tool information data, correcting the prior positions under the limiting conditions to obtain corrected positions conforming to constraints, and combining the corrected positions with clamping directions, reference plane normal directions and adjacent relations in the tool information data to generate geometric constraints. The method comprises the s