CN-121670140-B - LSW laser welding method and system

CN121670140BCN 121670140 BCN121670140 BCN 121670140BCN-121670140-B

Abstract

The invention provides an LSW laser welding method and a system, and relates to the technical field of laser welding, wherein the method comprises the steps of obtaining three-dimensional information of a large structural member and information of a plurality of welding robots; the method comprises the steps of determining a plurality of high-stress areas and a plurality of low-stress areas by using a stress analysis model based on three-dimensional information of a large structural member, constructing a collaborative task map based on a plurality of segmentation sub-paths, processing the collaborative task map based on a graph neural network to determine preliminary welding information of one or more welding sub-paths corresponding to each welding robot, and controlling the plurality of welding robots to carry out LSW laser welding based on target welding information of one or more welding sub-paths corresponding to each welding robot.

Inventors

LIANG XIANGYI
XU BO
CHEN CHANGYI

Assignees

四川福摩斯工业技术有限公司

Dates

Publication Date: 20260505
Application Date: 20260209

Claims (6)

1. A method of LSW laser welding comprising: acquiring three-dimensional information of a large structural member and information of a plurality of welding robots; Determining a plurality of high stress areas and a plurality of low stress areas by using a stress analysis model based on the three-dimensional information of the large structural member; determining a plurality of split sub-path information based on the three-dimensional information of the large structural member, the plurality of high stress regions, the plurality of low stress regions, the plurality of welding robot information, the determining a plurality of split sub-path information based on the three-dimensional information of the large structural member, the plurality of high stress regions, the plurality of low stress regions, the plurality of welding robot information comprising: Determining a plurality of welding passing points and a plurality of welding avoidance areas based on the three-dimensional information of the large structural member, the plurality of high stress areas and the plurality of low stress areas; determining total welding planning path information based on the three-dimensional information of the large structural member, the plurality of welding necessary points and the plurality of welding avoidance areas; determining a plurality of split sub-path information based on the total welding planning path information, the plurality of welding robot information; constructing a collaborative task map based on the plurality of segmentation sub-paths, wherein the collaborative task map comprises a plurality of task nodes and edges among the task nodes, each task node corresponds to one segmentation sub-path, the node characteristics of the task nodes comprise segmentation sub-path information and a plurality of welding robot information, and the edges among the nodes represent the spatial position relationship among the segmentation sub-paths; processing the collaborative task map based on a graph neural network to determine preliminary welding information of one or more welding sub-paths corresponding to each welding robot; Determining target welding information of one or more welding sub-paths corresponding to each welding robot based on the preliminary welding information of the one or more welding sub-paths corresponding to each welding robot, the determining the target welding information of the one or more welding sub-paths corresponding to each welding robot based on the preliminary welding information of the one or more welding sub-paths corresponding to each welding robot comprises: Controlling a plurality of welding robots to perform LSW laser welding on a large-scale structural member under test based on the preliminary welding information of one or more welding sub-paths corresponding to each welding robot, and acquiring a plurality of welding point information on each welding sub-path after welding is completed; Clustering based on the welding point information on each welding sub-path to obtain a plurality of welding information groups on each welding sub-path; Determining adjustment welding point information of each welding sub-path based on the plurality of welding information groups on each welding sub-path; Determining target welding information of one or more welding sub-paths corresponding to each welding robot based on the preliminary welding information of the one or more welding sub-paths corresponding to each welding robot and the adjustment welding point information of each welding sub-path; And controlling a plurality of welding robots to perform LSW laser welding based on the target welding information of one or more welding sub-paths corresponding to each welding robot.
2. The LSW laser welding method of claim 1, wherein said stress analysis model is a deep neural network model.
3. An LSW laser welding system comprising: The acquisition module is used for acquiring three-dimensional information of the large structural member and information of a plurality of welding robots; The stress analysis module is used for determining a plurality of high stress areas and a plurality of low stress areas by using a stress analysis model based on the three-dimensional information of the large structural member; A sub-path determining module configured to determine a plurality of split sub-path information based on the three-dimensional stereo information of the large structural member, the plurality of high stress areas, the plurality of low stress areas, the plurality of welding robot information, the sub-path determining module further configured to: Determining a plurality of welding passing points and a plurality of welding avoidance areas based on the three-dimensional information of the large structural member, the plurality of high stress areas and the plurality of low stress areas; determining total welding planning path information based on the three-dimensional information of the large structural member, the plurality of welding necessary points and the plurality of welding avoidance areas; determining a plurality of split sub-path information based on the total welding planning path information, the plurality of welding robot information; The map construction module is used for constructing a collaborative task map based on the plurality of segmentation sub-paths, the collaborative task map comprises a plurality of task nodes and edges between the task nodes, each task node corresponds to one segmentation sub-path, the node characteristics of the task nodes comprise segmentation sub-path information and a plurality of welding robot information, and the edges between the nodes represent the spatial position relationship between the segmentation sub-paths; the preliminary welding information determining module is used for processing the collaborative task map based on a graph neural network to determine preliminary welding information of one or more welding sub-paths corresponding to each welding robot; A target welding information determining module, configured to determine target welding information of one or more welding sub-paths corresponding to each welding robot based on the preliminary welding information of the one or more welding sub-paths corresponding to each welding robot, where the target welding information determining module is further configured to: Controlling a plurality of welding robots to perform LSW laser welding on a large-scale structural member under test based on the preliminary welding information of one or more welding sub-paths corresponding to each welding robot, and acquiring a plurality of welding point information on each welding sub-path after welding is completed; Clustering based on the welding point information on each welding sub-path to obtain a plurality of welding information groups on each welding sub-path; Determining adjustment welding point information of each welding sub-path based on the plurality of welding information groups on each welding sub-path; Determining target welding information of one or more welding sub-paths corresponding to each welding robot based on the preliminary welding information of the one or more welding sub-paths corresponding to each welding robot and the adjustment welding point information of each welding sub-path; and the welding control module is used for controlling the welding robots to carry out LSW laser welding based on the target welding information of one or more welding sub-paths corresponding to each welding robot.
4. The LSW laser welding system of claim 3 wherein said stress analysis model is a deep neural network model.
5. An electronic device comprising a processor, a memory, and a computer program, wherein the computer program is stored in the memory and configured to be executed by the processor to implement the LSW laser welding method of any of claims 1 to 2.
6. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the LSW laser welding method as claimed in any one of claims 1 to 2.

Description

LSW laser welding method and system Technical Field The invention relates to the technical field of laser welding, in particular to an LSW laser welding method and system. Background Laser Scanning Welding (LSW) has become one of the key technologies in the manufacture of large structural members by virtue of its high precision, high efficiency technology. However, the existing multi-robot LSW welding method still faces a number of prominent problems in practical applications. The large structural member has complex geometric form and uneven inherent stress distribution, the traditional welding path planning is not fully combined with stress distribution characteristics, the defects of cracks, deformation and the like are easy to occur in a high-stress area, and the integrity and the use safety of the structural member are seriously affected. When a plurality of robots work cooperatively, a scientific total welding path segmentation and task allocation mechanism is lacked, and the problems of low working efficiency, motion interference and the like are often caused, wherein the problems of robot load balancing and space position cooperation are difficult to consider. Meanwhile, parameters of welding information are mostly preset fixed values, influences of dynamic factors such as thermal deformation, working condition change and the like in the actual welding process are not considered, and welding information parameter adjustment depends on manual experience, and an adaptive optimization means based on actual welding data is lacked, so that welding quality consistency is poor, and high-precision manufacturing requirements of large structural parts are difficult to meet. These problems together limit the efficient application of multi-robot LSW welding technology in the manufacture of large structural members. Therefore, how to efficiently and accurately determine the multi-robot LSW welding information adaptation scheme of a large structural member is a current problem to be solved. Disclosure of Invention The invention mainly solves the technical problem of how to efficiently and accurately determine the multi-robot LSW welding information adaptation scheme of a large-scale structural member. According to a first aspect, the invention provides an LSW laser welding method, which comprises the steps of obtaining three-dimensional information of a large structural member and a plurality of welding robot information, determining a plurality of high-stress areas and a plurality of low-stress areas by using a stress analysis model based on the three-dimensional information of the large structural member, determining a plurality of pieces of split sub-path information based on the three-dimensional information of the large structural member, the plurality of high-stress areas, the plurality of low-stress areas and the plurality of welding robot information, constructing a collaborative task map based on the plurality of split sub-paths, wherein the collaborative task map comprises a plurality of task nodes and edges between the task nodes, each task node corresponds to one split sub-path, the node characteristics of the task nodes comprise split sub-path information and a plurality of welding robot information, the edges between the nodes represent spatial position relations between the split sub-paths, processing and determining preliminary welding information of one or more welding sub-paths corresponding to each welding robot based on a graph neural network, determining a preliminary welding sub-path corresponding to one or more welding robot based on the preliminary welding sub-path corresponding to each welding robot, and performing laser welding control welding robot welding target welding robot to each welding sub-path based on the preliminary welding sub-path or the plurality of welding robot welding sub-information. In one possible implementation manner, the determining the target welding information of one or more welding sub-paths corresponding to each welding robot based on the preliminary welding information of one or more welding sub-paths corresponding to each welding robot includes controlling a plurality of welding robots to perform LSW laser welding on a large-sized structural member under test and obtaining a plurality of welding point information on each welding sub-path after welding is completed, clustering based on the plurality of welding point information on each welding sub-path to obtain a plurality of welding information groups on each welding sub-path, determining the adjustment welding point information of each welding sub-path based on the plurality of welding information groups on each welding sub-path, and determining the target welding information of one or more welding sub-paths corresponding to each welding sub-path based on the preliminary welding information of one or more welding sub-paths corresponding to each welding robot and the adjustment welding point information of e