Method and system for quickly optimizing welding sequence of structural parts

Abstract

The invention provides a method and system for rapidly optimizing the welding sequence of structural parts, and relates to the technical field of structural part welding.The method comprises the steps that an assembly model of the structural parts is obtained; identifying the type of each structural part in the assembly body model, and defining the junction of each part as a standard welding seam; user requirements are manually input, a standard weld joint needing to be welded is selected, and the user requirements comprise the material type of the structural part, the performance requirements for the structural part and the welding method; and according to the input user demand, a machine learning method is utilized to preferably select a structural part welding sequence meeting the user demand, and a recommended optimal welding sequence and a welding process scheme are given. The welding sequence scheme meeting the requirement is obtained through computer calculation, the experiment frequency and the design production cost are reduced, the machine learning method is adopted, the design process can be simplified, and the design threshold is reduced.

Description

Technical Field

[0001] The invention relates to the technical field of structural component welding, and in particular to a method and system for quickly optimizing the welding sequence of structural components. Background Art

[0002] Structural welding plays a vital role in modern manufacturing, especially in high-tech fields such as aerospace, automobile manufacturing and shipbuilding. These industries have extremely strict requirements on welding quality, and the performance of welded joints is directly related to the strength, corrosion resistance and reliability of the overall structural parts. For example, wings and engine casings in the aerospace field, body frames in automobile manufacturing, hulls in ships, etc., all need to withstand high strength and complex environmental conditions. Therefore, a reasonable welding process is essential.

[0003] However, the complexity of structural parts brings challenges to welding. In actual operation, when the geometric shape of the structural parts is complex, the selection of welding sequence is particularly difficult. Inappropriate welding sequence may cause welding deformation, uneven internal stress and welding defects, which in turn affect the overall performance of the structural parts. In previous engineering practices, the selection of welding sequence mainly depends on empirical methods. Engineers determine the order and method of welding based on their own experience and past cases. Although this method is effective to a certain extent, it inevitably has subjectivity and limitations.

[0004] In recent years, with the development of computing technology, the finite element method has gradually become a common means of optimizing the welding sequence. This method can accurately simulate the temperature field, stress field and deformation behavior during the welding process by performing thermal-mechanical coupling analysis on the welding process, thereby providing a theoretical basis for optimizing the welding sequence. However, the amount of calculation of the finite element method is usually very large, requiring a large number of complex calculations, resulting in a significant increase in calculation time and cost. This poses a dual challenge of efficiency and accuracy in practical applications. Summary of the invention

[0005] In view of the shortcomings of existing methods for optimizing the welding sequence of structural parts, the present invention provides a method for quickly optimizing the welding sequence of structural parts, aiming to provide an efficient and accurate solution for optimizing the welding sequence of structural parts through a comprehensive database, machine learning algorithms and a convenient design process.

[0006] Specifically, the first aspect of the present invention provides a method for quickly optimizing the welding sequence of a structural part, comprising: Obtain assembly model of structural parts; Identify the type of each structural part in the assembly model, and define the junction of each part as a quasi-weld; Manually input user requirements and select the quasi-weld to be welded, wherein the user requirements include the material type of the structural component, the performance requirements of the structural component, and the welding method; Based on the input user needs, the machine learning method is used to select the welding sequence of structural parts that meets the user needs, and the recommended optimal welding sequence and welding process plan are given.

[0007] As a further explanation of the present invention, the assembly model is an assembly format file generated by 3D modeling software, which retains the assembly information of each structural part in the assembly.

[0008] As a further illustration of the present invention, after manually inputting user requirements and selecting the quasi-weld to be welded, the method further includes: The selected welds are defined as welds to be welded, and the remaining unselected quasi-welds are merged with the adjacent parts into one entity.

[0009] As a further illustration of the present invention, the performance requirements include requirements for post-weld residual stress and post-weld deformation performance indicators.

[0010] As a further explanation of the present invention, the identifying the type of each structural part in the assembly model specifically includes: Disassembling the assembly model into individual structural parts; The disassembled structural parts are matched with the basic structural part types in the basic structural part model database by using the part model recognition method, and the basic structural part types to which the structural parts in the assembly model belong are recognized.

[0011] As a further illustration of the present invention, the part model recognition method includes an image recognition method and a model parameter extraction method based on geometric features.

[0012] As a further explanation of the present invention, based on the input user requirements, a machine learning method is used to select a welding sequence of structural parts that meets the user requirements, and a recommended optimal welding sequence and welding process plan are given, specifically including: Using the finite element method, the basic structural parts in the basic structural parts model database are subjected to finite element welding simulation using different materials to obtain a large number of welding simulation results, wherein the welding simulation results include post-welding deformation and post-welding residual stress; Through the welding experiment method, the same structure, material, welding type, welding heat input and welding sequence as the finite element welding simulation process are used for experimental verification, and a large number of verification results are obtained, and the verification results include post-weld deformation and post-weld residual stress; The data sets obtained from the welding simulation and the welding experiment are used as the training set and the validation set in the machine learning model respectively, and the machine learning model is trained to obtain a machine learning model whose input is material type, structural part type, welding type, post-welding deformation requirement, and post-welding residual stress requirement, and whose output is welding heat input and welding sequence; The trained machine learning model is used to select the welding sequence of structural parts that meets user needs, and the recommended optimal welding sequence and welding process plan are given.

[0013] As a further illustration of the present invention, the welding heat input includes: welding current, welding voltage, and welding efficiency.

[0014] A second aspect of the present invention provides a system for rapidly optimizing a welding sequence of a structural part, comprising: Model import module, which is used to obtain the assembly model of the structural parts; A basic structural component model recognition module, which is used to identify the type of each structural component in the assembly model and define the junction of each component as a quasi-weld; A user requirement module, which is used to manually input user requirements and select the quasi-welds to be welded, wherein the user requirements include the material type of the structural part, the performance requirements of the structural part, and the welding method; The welding sequence optimization module is used to select the welding sequence of structural parts that meets the user's needs based on the input user needs and use machine learning methods, and give a recommended optimal welding sequence and welding process plan.

[0015] Compared with the prior art, the present invention has the following beneficial technical effects: 1. The existing structural parts welding sequence design optimization uses empirical methods that consume a lot of experimental costs. The present invention can use computer calculations to obtain a welding sequence solution that meets the needs, reduce the number of experiments, and reduce design and production costs.

[0016] 2. The existing structural parts welding sequence design optimization adopts the finite element method and has a long design cycle. The present invention can quickly design a welding sequence solution that meets the requirements and reduce the time cost of the design.

[0017] 3. The existing structural parts welding sequence design optimization method often requires the use of multiple software and manual statistical processing of the design calculation results. The present invention adopts a machine learning method to simplify the design process and lower the design threshold.

[0018] Other features and advantages of the technical solution will be described in the subsequent description, and partly become apparent from the description, or understood by implementing the technical solution. The purpose and other advantages of the technical solution can be achieved and obtained through the structures specifically pointed out in the written description and the drawings.

[0019] The technical solution of the present technical solution is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the technical solution and constitute a part of the specification. Together with the embodiments of the technical solution, they are used to explain the technical solution and do not constitute a limitation of the technical solution. In the accompanying drawings: Figure 1 A flow chart of a method for rapidly optimizing the welding sequence of structural parts provided by the present invention.

[0021] Figure 2 This is the E-shaped pipeline model recognition result diagram in the present invention.

[0022] Figure 3 Schematic diagram of the machine learning model training process in the present invention.

[0023] Figure 4 1 is the post-welding deformation in the finite element simulation result of the E-shaped pipeline in the embodiment of the present invention.

[0024] Figure 5 The figure shows the stress distribution after welding in the finite element simulation result of the E-shaped pipeline in the embodiment of the present invention. DETAILED DESCRIPTION

[0025] The preferred embodiments of the present technical solution are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present technical solution, and are not used to limit the present technical solution.

[0026] The embodiment of the present invention takes the welding sequence design of an E-type pipe joint as an example to illustrate the optimization method of the present invention.

[0027] like Figure 1 As shown, an embodiment of the present invention provides a method for quickly optimizing the welding sequence of a structural part, comprising the following steps: Step 1: Get the assembly model of the structural parts.

[0028] Specifically, by importing assembly format files generated by external 3D modeling software, such as .SLDASM, .x_t, and .ASM, the system can obtain the assembly model of the structural parts and retain the assembly information of each structural part in the assembly.

[0029] Step 2: Identify the type of each structural part in the assembly model and define the junction of each part as a quasi-weld.

[0030] The identification process specifically includes the following steps: Step 201: Disassemble the assembly model into individual structural parts.

[0031] Step 202: using a part model recognition method, the disassembled structural parts are matched with the basic structural part types in the basic structural part model database, and the basic structural part types to which the structural parts in the assembly model belong are identified.

[0032] Among them, the part model recognition method includes the use of image recognition method, model parameter extraction method based on geometric features and other methods. The above-mentioned basic structural parts model database is a pre-built database containing basic parts of various sizes and types, such as round tubes, square tubes, beams, etc. Figure 2 As shown, using the part model recognition method, the E-type pipe can be identified as a combination of a three-way pipe, an elbow pipe, and a straight pipe.

[0033] Step 3: Manually input user requirements and select the quasi-weld to be welded. User requirements include the material type of the structural part, the performance requirements of the structural part, and the welding method.

[0034] After the user selects the quasi-welds that need to be welded, the system will define the selected welds as welds to be welded, and the remaining unselected quasi-welds will be merged with the adjacent parts into one entity.

[0035] Specifically, the material type is the selected material grade, such as Q235, TC4, Ta1, etc.; the performance requirement is the requirement for post-weld residual stress and post-weld deformation performance indicators; the welding method can be, for example, GTAW welding, PAW welding, GMAW welding and other welding methods suitable for structural parts.

[0036] Step 4: Based on the input user requirements, use machine learning methods to select the welding sequence of structural parts that meets user requirements, and give the recommended optimal welding sequence and welding process plan.

[0037] Specifically, the above step 4 includes the following process: Step 401: Using the finite element method, the basic structural parts in the basic structural parts model database are subjected to finite element welding simulation using different materials to obtain a large number of welding simulation results, which include post-welding deformation and post-welding residual stress. Taking the E-type pipe joint as an example, the results obtained by the finite element welding simulation are as follows: Figure 4 and Figure 5 shown.

[0038] Step 402: Through the welding experiment method, the same structure, material, welding type, welding heat input, and welding sequence as the finite element welding simulation process are used for experimental verification, and a large number of verification results are obtained, including post-weld deformation and post-weld residual stress. (Welding heat input includes: welding current, welding voltage, and welding efficiency.) Step 403: The data sets obtained from the welding simulation and the welding experiment are used as the training set and the validation set in the machine learning model, respectively, to train the machine learning model, and obtain a machine learning model whose input is material type, structural part type, welding type, post-welding deformation requirement, and post-welding residual stress requirement, and whose output is welding heat input and welding sequence. The machine learning process is as follows: Figure 3 shown.

[0039] Step 404: Utilize the trained machine learning model to select the structural parts welding sequence that meets the user's needs, and give a recommended optimal welding sequence and welding process plan.

[0040] The present invention also provides a structural component welding sequence rapid optimization system, comprising: Model import module, which is used to obtain assembly models of structural parts.

[0041] The basic structural component model recognition module is used to identify the type of each structural component in the assembly model and define the junction of each component as a quasi-weld.

[0042] The user requirement module is used to manually input user requirements and select the quasi-welds that need to be welded. The user requirements include the material type of the structural part, the performance requirements of the structural part, and the welding method.

[0043] The welding sequence optimization module is used to select the welding sequence of structural parts that meets the user's needs based on the input user needs and use machine learning methods, and give the recommended optimal welding sequence and welding process plan.

[0044] Obviously, those skilled in the art can make various changes and modifications to the technical solution without departing from the spirit and scope of the technical solution. Thus, if these modifications and variations of the technical solution fall within the scope of the claims of the technical solution and their equivalents, the technical solution is also intended to include these modifications and variations.

Claims

1. A method for rapid optimization of the welding sequence of structural parts, characterized in that: include: Obtain assembly model of structural parts; Identify the type of each structural part in the assembly model, and define the junction of each part as a quasi-weld; Manually input user requirements and select the quasi-weld to be welded, wherein the user requirements include the material type of the structural component, the performance requirements of the structural component, and the welding method; Based on the input user needs, the machine learning method is used to select the welding sequence of structural parts that meets the user needs, and the recommended optimal welding sequence and welding process plan are given.

2. The method for rapid optimization of structural parts welding sequence according to claim 1, characterized in that: The assembly model is an assembly format file generated by 3D modeling software, which retains the assembly information of each structural part in the assembly.

3. The method for rapid optimization of structural parts welding sequence according to claim 1, characterized in that: After manually inputting user requirements and selecting the quasi-weld seam to be welded, the method further includes: The selected welds are defined as welds to be welded, and the remaining unselected quasi-welds are merged with the adjacent parts into one entity.

4. The method for rapid optimization of structural parts welding sequence according to claim 1, characterized in that: The performance requirements include requirements for post-weld residual stress and post-weld deformation performance indicators.

5. The method for rapid optimization of structural parts welding sequence according to claim 1, characterized in that: The identifying the type of each structural part in the assembly model specifically includes: Disassembling the assembly model into individual structural parts; The disassembled structural parts are matched with the basic structural part types in the basic structural part model database by using the part model recognition method, and the basic structural part types to which the structural parts in the assembly model belong are recognized.

6. The method for rapid optimization of structural parts welding sequence according to claim 5, characterized in that: The part model recognition method includes an image recognition method and a model parameter extraction method based on geometric features.

7. The method for rapid optimization of structural parts welding sequence according to claim 5, characterized in that: The method of machine learning is used to select the welding sequence of structural parts that meets the user's needs based on the input user needs, and give a recommended optimal welding sequence and welding process plan, specifically including: Using the finite element method, the basic structural parts in the basic structural parts model database are subjected to finite element welding simulation using different materials to obtain a large number of welding simulation results, wherein the welding simulation results include post-welding deformation and post-welding residual stress; Through the welding experiment method, the same structure, material, welding type, welding heat input and welding sequence as the finite element welding simulation process are used for experimental verification, and a large number of verification results are obtained, and the verification results include post-weld deformation and post-weld residual stress; The data sets obtained from the welding simulation and the welding experiment are used as the training set and the validation set in the machine learning model respectively, and the machine learning model is trained to obtain a machine learning model whose input is material type, structural part type, welding type, post-welding deformation requirement, and post-welding residual stress requirement, and whose output is welding heat input and welding sequence; The trained machine learning model is used to select the welding sequence of structural parts that meets user needs, and the recommended optimal welding sequence and welding process plan are given.

8. The method for rapid optimization of structural parts welding sequence according to claim 7, characterized in that: The welding heat input includes: welding current, welding voltage, and welding efficiency.

9. A system for rapid optimization of structural parts welding sequence, characterized in that: include: Model import module, which is used to obtain the assembly model of the structural parts; A basic structural component model recognition module, which is used to identify the type of each structural component in the assembly model and define the junction of each component as a quasi-weld; A user requirement module, which is used to manually input user requirements and select the quasi-welds to be welded, wherein the user requirements include the material type of the structural part, the performance requirements of the structural part, and the welding method; The welding sequence optimization module is used to select the welding sequence of structural parts that meets the user's needs based on the input user needs and use machine learning methods, and give a recommended optimal welding sequence and welding process plan.

Images