Body-in-white welding fixture design method based on multi-modal large model
By using collaborative reasoning and physics field optimization of multimodal large models, the problems of manual dependence and low efficiency in the design of welding fixtures for body-in-white are solved, and automated and efficient welding fixture design is realized.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI PHARMATECHS CO LTD
- Filing Date
- 2026-04-03
- Publication Date
- 2026-07-10
AI Technical Summary
Existing body-in-white welding fixture designs rely on manual experience, have long design cycles, and are prone to repeated modifications in complex scenarios, resulting in poor design efficiency and engineering consistency.
A multimodal large model-based approach is adopted, which maps multi-source design data for body-in-white welding through a unified semantic index, performs multimodal semantic representation and engineering constraint modeling, combines multimodal large model for collaborative reasoning, and uses physical field optimization objectives for iterative optimization to generate parametric welding fixture design schemes and models.
It improves the efficiency and accuracy of welding fixture design for body-in-white, realizes the automated generation from abstract semantics to specific engineering drawings, reduces reliance on manual experience, and enhances design consistency.
Smart Images
Figure CN122365749A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of computer technology, and more specifically, to a design method for a welding fixture for a body-in-white based on a multimodal large model. Background Technology
[0002] In the automotive manufacturing industry, the design of welding fixtures for body-in-white is a key process for achieving precise positioning, stable clamping, and welding quality control of body components. With the development of lightweight and platform-based automotive design, body-in-white structures are exhibiting a diversification of materials, which places higher demands on welding fixture design.
[0003] Currently, the design of welding fixtures for body-in-white mainly relies on a combination of manual experience and computer-aided design software. Designers typically determine the positioning reference (RPS), clamping method, and fixture structure layout manually based on a 3D digital model of the body, combined with welding process planning documents and production line layout information. The design is then gradually improved through multiple design modifications and experimental verifications. The design process is highly dependent on the designer's accumulated experience and understanding of welding processes, resulting in a long design cycle. In complex welding scenarios, repeated modifications or even complete rework are common, severely restricting design efficiency and engineering consistency. Summary of the Invention
[0004] The problem addressed by this invention is how to improve the design efficiency of welding fixtures for body-in-white.
[0005] To address the aforementioned problems, this invention provides a design method for a welding fixture for a body-in-white based on a multimodal large model.
[0006] In a first aspect, the present invention provides a method for designing a welding fixture for a body-in-white based on a multimodal large model, comprising: Multi-source design data for welding body-in-white is mapped using a unified semantic index to obtain a multimodal semantic representation; Engineering constraint modeling is performed based on the multimodal semantic representation to obtain clamping and welding engineering constraints; The multimodal semantic representation and the clamping and welding engineering constraints are input into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model. The parametric CAD model of the welding fixture is obtained through the parametric welding fixture design scheme and the parametric welding fixture model.
[0007] Optionally, the clamping and welding engineering constraints include clamping degree-of-freedom constraints and welding torch accessibility constraints; The clamping freedom constraints include: , in, To locate the constraint matrix, A spiral vector with n degrees of freedom constrained; The reachability constraints of the welding torch include: , in, For the reachability constraint of the welding torch, The collection of the motion envelope of the welding torch. The boundary of the envelope set of the welding torch motion. For the boundary point, Let x be a sign function of the point inside or outside the envelope.
[0008] Optionally, the clamping and welding engineering constraints also include clamping stiffness field constraints and welding deformation sensitivity field constraints; The clamping stiffness field constraint includes: , in, The clamping stiffness field is constrained. The total number of clamping points. Let x be the equivalent stiffness and x be the location of the point to be queried. Let i be the position of the i-th clamping point. The influence range coefficient of the clamping stiffness field constraint; The welding deformation sensitivity field constraint includes: , in, The welding deformation sensitivity field is constrained. This represents the total number of weld locations. Let j be the unit heat input intensity of the j-th weld. For the spatial location of the weld, This is the influence range coefficient of the welding deformation sensitivity field constraint.
[0009] Optionally, the step of inputting the multimodal semantic representation and the clamping and welding engineering constraints into a multimodal large model for collaborative reasoning to obtain a parametric welding fixture design scheme includes: The workpiece freedom constraint state represented by the multimodal semantics is analyzed based on the clamping freedom constraints to obtain the positioning unit type; The functional unit type is obtained by utilizing the clamping stiffness field constraint and the welding deformation sensitivity field constraint, through the multimodal semantic representation and the positioning unit type; The parametric welding fixture design scheme is obtained through the welding torch accessibility constraint and the multimodal semantic representation.
[0010] Optionally, the iterative optimization of the parametric welding fixture design using a physics field optimization objective to obtain a parametric welding fixture model includes: Obtain the physical potential energy function; The physical potential energy function includes: , in, Let be the physical potential energy function. For the reachability potential energy term of the welding torch, For the fixture stiffness term, For welding variable energy term, , , These are the weighting coefficients; The physical field optimization objective is constructed based on the physical potential energy function; The physical field optimization objectives include: , in, Density of the structural material; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model.
[0011] Optionally, after obtaining the parametric CAD model of the welding fixture from the parametric welding fixture model, the method further includes: The constraint correction vector is obtained by modifying the vector according to the designer. The constraint correction vector is used to optimize the multimodal large model by inputting it in the form of context bias. The constraint correction vector includes: , in, The constraint correction vector is... Modify the vector for the designer. This is to correct the vector before.
[0012] Optionally, the step of mapping the multi-source design data for body-in-white welding through a unified semantic index to obtain a multimodal semantic representation includes: The multi-source design data for welding the body-in-white is mapped using a unified semantic index to obtain the multimodal semantic representation; The multimodal semantic representation includes: , in, This is the multimodal semantic representation. Geometric eigenvectors For welding process semantic vectors, This is the spatial constraint vector.
[0013] Secondly, the present invention provides a design device for welding fixtures for body-in-white based on a multimodal large model, comprising: a multimodal data fusion module, used to map multi-source design data for body-in-white welding through a unified semantic index to obtain a multimodal semantic representation; The engineering constraint modeling module is used to perform engineering constraint modeling based on the multimodal semantic representation to obtain clamping and welding engineering constraints; The multimodal large model reasoning module is used to input the multimodal semantic representation and the clamping and welding engineering constraints into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The physics-driven optimization module is used to iteratively optimize the parametric welding fixture design scheme using the physics optimization objective to obtain the parametric welding fixture model. The parametric reconstruction and drawing module is used to obtain a parametric CAD model of the welding fixture from the parametric welding fixture design scheme and the parametric welding fixture model.
[0014] Thirdly, the present invention provides an electronic device, including a memory and a processor; The memory is used to store computer programs; The processor is configured to, when executing the computer program, implement the white body welding fixture design method based on a multimodal large model as described in the first aspect.
[0015] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the design method for a body-in-white welding fixture based on a multimodal large model as described in the first aspect.
[0016] The beneficial effects of the multimodal large-scale model-based body-in-white welding fixture design method of this invention are as follows: By establishing a unified semantic index to map multi-source design data for body-in-white welding, a multimodal semantic representation is obtained, effectively solving the information silo problem caused by inconsistent data formats and semantic fragmentation, and significantly improving data consistency. Engineering constraint modeling is performed based on the multimodal semantic representation to obtain clamping and welding engineering constraints, avoiding reliance on manual experience and improving constraint accuracy. A multimodal large-scale model with engineering semantic understanding capability is introduced, and a parametric welding fixture design scheme is obtained by combining the multimodal semantic representation and clamping and welding engineering constraints, generating the parametric welding fixture design scheme in a hierarchical and collaborative manner. The parametric welding fixture design scheme is automatically refined and optimized using a physics-driven approach based on physical field optimization objectives, resulting in a parametric welding fixture model that is more suitable for the actual working conditions of body-in-white welding fixtures. A parametric CAD model of the welding fixture is obtained by reconstructing the parametric welding fixture design scheme and the parametric welding fixture model, realizing the automated generation from abstract semantics to specific engineering drawings, thereby improving the design efficiency of body-in-white welding fixtures. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating a method for designing a welding fixture for a body-in-white based on a multimodal large model, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a white body welding fixture design device based on a multimodal large model according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation
[0018] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.
[0019] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.
[0020] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.
[0021] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".
[0022] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.
[0023] In related technologies, with the development of lightweight and platform-based design in automobiles, body-in-white structures are showing trends such as material diversification, the mixed use of high-strength steel and aluminum alloys, thinner component walls, and more complex curved surfaces, which places higher demands on welding fixture design. On the one hand, the fixture needs to reliably position and clamp complex curved parts within a limited space; on the other hand, it must also fully consider the heat input, force path, and welding torch movement space during the welding process to avoid problems such as welding deformation, welding torch interference, and insufficient fixture rigidity. Therefore, the design of body-in-white welding fixtures is essentially a complex engineering design problem that is simultaneously subject to geometric, process, and physical constraints.
[0024] To address the problems existing in the aforementioned related technologies, this embodiment provides a design method for a welding fixture for a body-in-white based on a multimodal large model.
[0025] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for designing a welding fixture for a body-in-white based on a multimodal large model, comprising: Step 110: Map the multi-source design data for welding the body-in-white using a unified semantic index to obtain a multimodal semantic representation.
[0026] Specifically, multi-source design data for body-in-white welding in the welding scenario is acquired. This multi-source design data includes three-dimensional geometric models of body components or assemblies, welding process documents, clamping freedom constraint information, and production line spatial layout information. The multi-source design data is parsed, aligned, and fused to construct a multimodal semantic unified representation model for welding tooling design. The three-dimensional geometric models of the body in the multi-source design data include one or more of B-Rep models, surface models, or point cloud models, and contain hole positions, feature lines, flanges, and assembly datum information. The welding process diagram is represented in graph structure form, with nodes representing welds or weld points, and edges representing welding sequence, process dependencies, or heat-affected zone relationships.
[0027] Step 120: Perform engineering constraint modeling based on the multimodal semantic representation to obtain clamping and welding engineering constraints.
[0028] Specifically, the clamping and welding engineering constraints include a set of constraints on the six degrees of freedom of the body-in-white components in the welding state. The integrity and stability of the clamping freedom constraints are explicitly described through the spatial configuration of positioning points, support points, and clamping points. It also includes the force transmission path caused by welding sequence, heat input parameters, and welding posture during the welding process, which is used to evaluate the ability of the fixture structure to suppress welding deformation.
[0029] Step 130: Input the multimodal semantic representation and the clamping and welding engineering constraints into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme.
[0030] Specifically, the multimodal large model adopts a hierarchical reasoning approach with structure-parameter decoupling. First, it generates the functional unit types and structural topology relationships of the welding fixture. Then, under the topological constraints, it generates the continuous parameter value ranges for each functional unit. Through this process, a parametric welding fixture design scheme is obtained. Each functional unit includes at least one or more of the following: positioning unit, clamping unit, support unit, adjustment unit, and connection unit. Each functional unit is associated with corresponding engineering functional semantics and assembly interface constraints.
[0031] Step 140: Iteratively optimize the parametric welding fixture design scheme using the physical field optimization objective to obtain the parametric welding fixture model.
[0032] Specifically, the physical field optimization objective is to comprehensively optimize the overall stiffness, positioning error, welding accessibility, structural complexity, and assembly feasibility of the fixture. It introduces constraint penalty terms based on engineering design specifications to suppress structural schemes that violate welding process specifications or standardized fixture design guidelines.
[0033] Step 150: Obtain the parametric CAD model of the welding fixture through the parametric welding fixture design scheme and the parametric welding fixture model.
[0034] Specifically, through the parametric welding fixture model, the structural generation process optimizes the main structure of the fixture using a continuous parametric field or material distribution function. By extracting the skeleton and geometrically fitting this continuous structure result, the system automatically identifies: the structural centerline or main force path; key node locations; and the length, direction, and cross-sectional variation trends of connecting components. For positioning and clamping units, the system further combines the clamping stiffness field and welding deformation sensitivity field to automatically calculate their working parameters. Based on the structural topology generated in the parametric welding fixture design scheme, the system automatically infers the assembly interface type and connection method between each functional unit. Structural parameters are automatically calculated from the preceding structural generation and physical optimization process. Based on the above parameter combinations, a parameter configuration vector is obtained. Standard functional parts and connectors are instantiated from the parametric parts library, and a three-dimensional assembly of the welding fixture is automatically constructed based on assembly constraint inference. The three-dimensional assembly is parametrically reconstructed to generate a fully parametric CAD model with a complete feature history tree, dimension-driven relationships, and assembly constraints, and a two-dimensional engineering drawing conforming to the vehicle body welding drawing specifications is automatically generated. The automatic assembly step infers based on the assembly constraint diagram between fixture components, which includes: , in, This is an assembly constraint diagram, where P represents a set of parts and C represents assembly constraint relationships. The assembly constraint diagram describes the connection types, assembly sequence, and assembly tolerance relationships between components. The two-dimensional engineering drawing includes a welding fixture assembly drawing, part drawings, and welding process annotation information, and the welding process annotation information is consistent with the welding process diagram.
[0035] In this embodiment, a unified semantic index is established to map multi-source design data for body-in-white welding, resulting in a multimodal semantic representation. This effectively solves the problem of information silos caused by inconsistent data formats and semantic fragmentation, significantly improving data consistency. Engineering constraint modeling is performed based on the multimodal semantic representation to obtain clamping and welding engineering constraints, avoiding reliance on manual experience and improving constraint accuracy. A large multimodal model with engineering semantic understanding is introduced, and combined with the multimodal semantic representation and clamping and welding engineering constraints, a parametric welding fixture design scheme is obtained, which is generated hierarchically and collaboratively. The parametric welding fixture design scheme is automatically refined and optimized using a physics-driven approach, utilizing physical field optimization objectives, to obtain a parametric welding fixture model that is more suitable for the actual working conditions of body-in-white welding fixtures. A parametric CAD model of the welding fixture is obtained by reconstructing the parametric welding fixture design scheme and the parametric welding fixture model, realizing the automated generation from abstract semantics to specific engineering drawings, thereby improving the design efficiency of body-in-white welding fixtures.
[0036] Optionally, the clamping and welding engineering constraints include clamping degree-of-freedom constraints and welding torch accessibility constraints; The clamping freedom constraints include: , in, To locate the constraint matrix, A spiral vector constrained by n degrees of freedom; The reachability constraints of the welding torch include: , in, For the reachability constraint of the welding torch, The collection of the motion envelope of the welding torch. The boundary of the envelope set of the welding torch motion. For the boundary point, Let x be a sign function of the point inside or outside the envelope.
[0037] Specifically, the system constructs a degree-of-freedom constrained spiral vector for each positioning point or support point. This forms the clamping degree of freedom constraint. The rank of the system calculation matrix includes: , Among them, when Furthermore, if the matrix condition number is less than a preset threshold, the workpiece is determined to meet the stable positioning requirements under welding conditions; otherwise, feedback is sent to the subsequent inference stage to adjust the positioning layout. The welding torch reachability constraint field is constructed based on the motion trajectory and attitude envelope of the welding torch under each welding process, used to constrain the interference relationship between the fixture structure and the welding torch motion space. The system constructs a set of welding torch motion envelopes based on the welding torch geometric model and the motion trajectory of the welding torch end effector during welding. The reachability constraint field of the welding torch is defined using a Signed Distance Function (SDF). When... At this time, the spatial region is defined as the unsuitable area for welding torch placement, and the welding torch obstacle avoidance penalty potential energy function is defined as follows: , in, For safe distance threshold, Let the potential energy function be the penalty for obstacle avoidance by the welding torch. For areas where deployment is not allowed, This represents the distance to the obstacle. It is used as an exclusion constraint during the structure generation phase.
[0038] Optionally, the clamping and welding engineering constraints also include clamping stiffness field constraints and welding deformation sensitivity field constraints; The clamping stiffness field constraint includes: , in, The clamping stiffness field is constrained. The total number of clamping points. Let x be the equivalent stiffness and x be the location of the point to be queried. Let i be the position of the i-th clamping point. The influence range coefficient of the clamping stiffness field constraint; The welding deformation sensitivity field constraint includes: , in, The welding deformation sensitivity field is constrained. This represents the total number of weld locations. Let j be the unit heat input intensity of the j-th weld. For the spatial location of the weld, This is the influence range coefficient of the welding deformation sensitivity field constraint.
[0039] Specifically, the system abstracts each clamping point as an equivalent spring model, with an equivalent stiffness of... The system defines a clamping stiffness field constraint, which guides the fixture structure to preferentially generate in areas that contribute significantly to workpiece stability. Based on welding thermal input parameters and weld location, the system constructs a welding deformation sensitivity field. In areas of high deformation sensitivity, the system automatically enhances structural stiffness or shortens the force transmission path. The clamping stiffness field and welding deformation suppression field, constructed based on welding thermal input parameters, material properties, and clamping force boundary conditions, guide the fixture structure to generate or reinforce in low-deformation regions.
[0040] In this optional embodiment, existing body-in-white welding fixture design processes suffer from high reliance on manual labor, difficulty in collaboratively utilizing multimodal design information, insufficient consideration of clamping and welding physical constraints, and the inability to directly apply the generated results to engineering manufacturing. Based on the aforementioned three-dimensional geometric model and welding process semantics, a system is established to describe the workpiece's degree-of-freedom constraints, clamping force direction, welding force transmission path, and welding deformation suppression requirements in the welding state, providing an engineering constraint basis for subsequent fixture generation.
[0041] Optionally, the step of inputting the multimodal semantic representation and the clamping and welding engineering constraints into a multimodal large model for collaborative reasoning to obtain a parametric welding fixture design scheme includes: The workpiece freedom constraint state represented by the multimodal semantics is analyzed based on the clamping freedom constraints to obtain the positioning unit type; The functional unit type is obtained by utilizing the clamping stiffness field constraint and the welding deformation sensitivity field constraint, through the multimodal semantic representation and the positioning unit type; The parametric welding fixture design scheme is obtained through the welding torch accessibility constraint and the multimodal semantic representation.
[0042] Specifically, multimodal semantic representations and clamping / welding engineering constraints are input into a multimodal large-scale model for collaborative reasoning. The multimodal large-scale model does not directly generate the complete fixture structure, but rather employs a hierarchical reasoning strategy prioritizing functional units, first determining "which types of functional units are needed." The system first analyzes the workpiece's degree-of-freedom constraint state under the current welding condition based on clamping freedom constraints. If there are incompletely constrained degrees-of-freedom directions, the system inputs these directions as explicit semantic prompts into the multimodal large-scale model. The model then integrates... The system infers the appropriate positioning unit type to compensate for the degree of freedom based on the geometric feature type. After inferring the positioning unit type, the system further infers the type and quantity of clamping functional units based on the welding deformation sensitivity field and clamping stiffness field. The multimodal large model analyzes the overlap between high deformation-sensitive areas and high stiffness-contributing areas to determine whether clamping units need to be introduced and infers the type of clamping units to be introduced. Secondly, after determining the functional unit type, the multimodal large model further infers the spatial arrangement area and relative positional relationship of each functional unit. The system first eliminates areas inaccessible to the welding torch from the geometric space based on the welding torch accessibility constraint field, forming a candidate arrangement space set. Then, the multimodal large model performs spatial distribution inference based on comprehensive factors to optimize performance, such as: positioning units should be preferentially close to the geometric reference plane or reference hole; clamping units should be preferentially distributed in areas with high welding deformation sensitivity and significant stiffness contribution; and functional units should meet assembly space and operation space constraints. Finally, based on the spatial adjacency relationships, force transmission paths, and assembly logic between functional units, the multimodal large model first generates the types and spatial distribution schemes of the positioning and clamping functional units, and outputs the structural topology relationships. , where node V represents a functional unit and edge E represents a connection relationship.
[0043] Optionally, the iterative optimization of the parametric welding fixture design using a physics field optimization objective to obtain a parametric welding fixture model includes: Obtain the physical potential energy function; The physical potential energy function includes: , in, Let be the physical potential energy function. For the reachability potential energy term of the welding torch, For the fixture stiffness term, For welding variable energy term, , , These are the weighting coefficients; The physical field optimization objective is constructed based on the physical potential energy function; The physical field optimization objectives include: , in, Density of the structural material; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model.
[0044] Specifically, through iterative optimization, the structure automatically avoids the welding torch envelope area and forms a stable force transmission path in areas with high stiffness and low deformation sensitivity.
[0045] Optionally, after obtaining the parametric CAD model of the welding fixture from the parametric welding fixture model, the method further includes: The constraint correction vector is obtained by modifying the vector according to the designer. The constraint correction vector is used to optimize the multimodal large model by inputting it in the form of context bias. The constraint correction vector includes: , in, The constraint correction vector is... Modify the vector for the designer. This is to correct the vector before.
[0046] Specifically, the correction vectors are grouped according to parameter type, including geometric correction components, constraint correction components, and logical correction components. For multimodal large-scale models, the constraint correction vectors are not directly used for supervised training, but are injected into the inference process of the multimodal large-scale model in the form of context bias, prioritizing the generation of solutions that are more in line with engineering practices and experience in subsequent inference. For engineering constraint parameters, the system maps the correction vectors back to the engineering constraint model and updates the relevant constraint weights.
[0047] Optionally, the step of mapping the multi-source design data for body-in-white welding through a unified semantic index to obtain a multimodal semantic representation includes: The multi-source design data for welding the body-in-white is mapped using a unified semantic index to obtain the multimodal semantic representation; The multimodal semantic representation includes: , in, This is the multimodal semantic representation. Geometric eigenvectors For welding process semantic vectors, This is the spatial constraint vector.
[0048] Specifically, the system parses the vehicle body geometry model into a geometric feature map with topological relationships, parses the welding process into a welding process atlas, and encodes the production line space into a spatial constraint domain. All modal information is mapped to the same high-dimensional feature space through a unified semantic index, forming a multimodal semantic representation for subsequent reasoning and optimization.
[0049] like Figure 2 As shown in the figure, an embodiment of the present invention provides a design device for a welding fixture for a body-in-white based on a multimodal large model, comprising: The multimodal data fusion module 10 is used to map the multi-source design data of body-in-white welding through a unified semantic index to obtain a multimodal semantic representation; The engineering constraint modeling module 20 is used to perform engineering constraint modeling based on the multimodal semantic representation to obtain clamping and welding engineering constraints. The multimodal large model reasoning module 30 is used to input the multimodal semantic representation and the clamping and welding engineering constraints into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The physics-driven optimization module 40 is used to iteratively optimize the parametric welding fixture design scheme using the physics optimization objective to obtain the parametric welding fixture model. The parametric reconstruction and drawing module 50 is used to obtain a parametric CAD model of the welding fixture through the parametric welding fixture design scheme and the parametric welding fixture model.
[0050] The body-in-white welding fixture design device based on a multimodal large model in this embodiment is used to implement the body-in-white welding fixture design method based on a multimodal large model as described above. Its advantages over the prior art are the same as those of the body-in-white welding fixture design method based on a multimodal large model as described above, and will not be repeated here.
[0051] like Figure 3 As shown in the figure, an electronic device 300 provided in this embodiment of the invention includes a memory 310 and a processor 320; the memory 310 is used to store a computer program; the processor 320 is used to implement the above-described design method for a body-in-white welding fixture based on a multimodal large model when the computer program is executed.
[0052] Alternatively, an electronic device 300 includes a memory 310 and a processor 320 coupled to the memory 310; the memory 310 is configured to store a computer program; and the processor 320 is configured to perform the following operations when the computer program is executed: Multi-source design data for welding body-in-white is mapped using a unified semantic index to obtain a multimodal semantic representation; Engineering constraint modeling is performed based on the multimodal semantic representation to obtain clamping and welding engineering constraints; The multimodal semantic representation and the clamping and welding engineering constraints are input into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model. The parametric CAD model of the welding fixture is obtained through the parametric welding fixture design scheme and the parametric welding fixture model.
[0053] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the design method for a body-in-white welding fixture based on a multimodal large model as described above.
[0054] Alternatively, a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the following operations: Multi-source design data for welding body-in-white is mapped using a unified semantic index to obtain a multimodal semantic representation; Engineering constraint modeling is performed based on the multimodal semantic representation to obtain clamping and welding engineering constraints; The multimodal semantic representation and the clamping and welding engineering constraints are input into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model. The parametric CAD model of the welding fixture is obtained through the parametric welding fixture design scheme and the parametric welding fixture model.
[0055] The present invention will now be described an electronic device 300 that can serve as a server or client of the present invention, which is an example of a hardware device that can be applied to various aspects of the present invention. Electronic device 300 is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 300 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.
[0056] Electronic device 300 includes a computing unit that can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) or a computer program loaded from a storage unit into random access memory (RAM). The RAM may also store various programs and data required for device operation. The computing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.
[0057] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention according to actual needs. Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units can be implemented in hardware or as software functional units.
[0058] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.
Claims
1. A design method for a welding fixture for a body-in-white based on a multimodal large model, characterized in that, include: Multi-source design data for welding body-in-white is mapped using a unified semantic index to obtain a multimodal semantic representation; Engineering constraint modeling is performed based on the multimodal semantic representation to obtain clamping and welding engineering constraints; The multimodal semantic representation and the clamping and welding engineering constraints are input into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model. The parametric CAD model of the welding fixture is obtained through the parametric welding fixture design scheme and the parametric welding fixture model.
2. The design method for welding fixtures for body-in-white based on a multimodal large model according to claim 1, characterized in that, The constraints on the clamping and welding process include clamping degree-of-freedom constraints and welding torch accessibility constraints. The clamping freedom constraints include: ， in, To locate the constraint matrix, A spiral vector with n degrees of freedom constrained; The reachability constraints of the welding torch include: ， in, For the reachability constraint of the welding torch, The collection of the motion envelope of the welding torch. The boundary of the envelope set of the welding torch motion. For the boundary point, Let x be a sign function of the point inside or outside the envelope.
3. The method for designing a welding fixture for a body-in-white based on a multimodal large model according to claim 2, characterized in that, The clamping and welding engineering constraints also include clamping stiffness field constraints and welding deformation sensitivity field constraints; The clamping stiffness field constraint includes: ， in, The clamping stiffness field is constrained. The total number of clamping points. Let x be the equivalent stiffness and x be the location of the point to be queried. Let i be the position of the i-th clamping point. The influence range coefficient of the clamping stiffness field constraint; The welding deformation sensitivity field constraint includes: ， in, The welding deformation sensitivity field is constrained. This represents the total number of weld locations. Let j be the unit heat input intensity of the j-th weld. For the spatial location of the weld, This is the influence range coefficient of the welding deformation sensitivity field constraint.
4. The method for designing a welding fixture for a body-in-white based on a multimodal large model according to claim 3, characterized in that, The step of inputting the multimodal semantic representation and the clamping and welding engineering constraints into a multimodal large model for collaborative reasoning to obtain a parametric welding fixture design scheme includes: The workpiece freedom constraint state represented by the multimodal semantics is analyzed based on the clamping freedom constraints to obtain the positioning unit type; The functional unit type is obtained by utilizing the clamping stiffness field constraint and the welding deformation sensitivity field constraint, through the multimodal semantic representation and the positioning unit type; The parametric welding fixture design scheme is obtained through the welding torch accessibility constraint and the multimodal semantic representation.
5. The method for designing a welding fixture for a body-in-white based on a multimodal large model according to claim 1, characterized in that, The step of iteratively optimizing the parametric welding fixture design scheme using a physics field optimization objective to obtain a parametric welding fixture model includes: Obtain the physical potential energy function; The physical potential energy function includes: ， in, Let be the physical potential energy function. For the reachability potential energy term of the welding torch, For the fixture stiffness term, For welding variable energy term, , , These are the weighting coefficients; The physical field optimization objective is constructed based on the physical potential energy function; The physical field optimization objectives include: ， in, Density of the structural material; The parametric welding fixture design scheme is iteratively optimized using the physical field optimization objective to obtain the parametric welding fixture model.
6. The method for designing a welding fixture for a body-in-white based on a multimodal large model according to claim 1, characterized in that, After obtaining the parametric CAD model of the welding fixture through the parametric welding fixture model, the method further includes: The constraint correction vector is obtained by modifying the vector according to the designer. The constraint correction vector is used to optimize the multimodal large model by inputting it in the form of context bias. The constraint correction vector includes: ， in, The constraint correction vector is... Modify the vector for the designer. This is to correct the vector before.
7. The method for designing a welding fixture for a body-in-white based on a multimodal large model according to claim 1, characterized in that, The process of mapping multi-source design data for body-in-white welding using a unified semantic index to obtain a multimodal semantic representation includes: The multi-source design data for welding the body-in-white is mapped using a unified semantic index to obtain the multimodal semantic representation; The multimodal semantic representation includes: ， in, For the multimodal semantic representation, Geometric eigenvectors For welding process semantic vectors, This is the spatial constraint vector.
8. A design device for a welding fixture for a body-in-white based on a multimodal large model, characterized in that, include: The multimodal data fusion module is used to map multi-source design data for body-in-white welding through a unified semantic index to obtain a multimodal semantic representation; The engineering constraint modeling module is used to perform engineering constraint modeling based on the multimodal semantic representation to obtain clamping and welding engineering constraints; The multimodal large model reasoning module is used to input the multimodal semantic representation and the clamping and welding engineering constraints into the multimodal large model for collaborative reasoning to obtain the parametric welding fixture design scheme; The physics-driven optimization module is used to iteratively optimize the parametric welding fixture design scheme using the physics optimization objective to obtain the parametric welding fixture model. The parametric reconstruction and drawing module is used to obtain a parametric CAD model of the welding fixture from the parametric welding fixture design scheme and the parametric welding fixture model.
9. An electronic device, characterized in that, Including memory and processor; The memory is used to store computer programs; The processor is configured to, when executing the computer program, implement the white body welding fixture design method based on a multimodal large model as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the design method for a body-in-white welding fixture based on a multimodal large model as described in any one of claims 1 to 7.