Parameter modeling method and system for intelligent generation of CAD drawings driven by large models

Abstract

The application provides a large model driven CAD drawing intelligent generation parameter modeling method and system, relates to the technical field of data processing, and the method acquires design requirement information, analyzes natural language description, multi-source engineering data and historical design samples, generates a structured semantic feature set; converts semantic constraints into parameter constraint relationships, extracts parameter variables and constructs a parameter constraint model; generates an adaptive parameter modeling sequence according to the parameter constraint model; gradually generates geometric features in a CAD modeling environment, and adjusts conflict parameter variables, parameter constraint relationships or geometric feature generation sequences when there is a constraint conflict, to obtain a CAD drawing; verifies the CAD drawing and updates the parameter constraint model; and thus adaptive parameter modeling generation from design requirements to the CAD drawing is realized.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for intelligent generation of parametric modeling for CAD drawings driven by large models. Background Technology

[0002] The closest existing technologies to this patent are typically based on parametric modeling and rule-driven CAD automatic modeling methods. Specifically, the system predefines geometric feature templates, constraints, and parametric expressions, and combines these with scripts or feature tree mechanisms to automatically generate 2D or 3D drawings after the user inputs key dimensional parameters. Simultaneously, some solutions incorporate knowledge bases or simple machine learning models to perform template matching and parameter filling for common structures, thereby improving modeling efficiency. These methods have been widely applied in fields such as mechanical parts design and architectural component drawing.

[0003] However, in the design of complex, irregularly shaped parts, when design requirements are input in the form of unstructured text or multi-source data, existing technologies struggle to accurately interpret the design intent, relying instead on manual conversion of requirements into parameters and gradual adjustments. Specifically, in the design of an aircraft support structure, engineers needed to repeatedly modify multiple parameters based on the description of "weight reduction while maintaining local strength." The system could not automatically establish a mapping between semantics and geometry, leading to multiple reconstructions of the feature tree and constraint conflicts, increasing the complexity of design iterations. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for intelligent generation of parametric modeling for CAD drawings driven by large models, in order to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a parametric modeling method for intelligent generation of CAD drawings driven by large models, the method comprising: The system acquires design requirement information, parses the natural language descriptions, multi-source engineering data, and historical design samples in the design requirement information, identifies semantic entities representing the objects to be modeled and semantic constraints representing design requirements; generates a structured semantic feature set based on semantic entities, semantic constraints, and the relationship between them, and establishes an initial mapping relationship between the structured semantic feature set and geometric feature types. Based on the structured semantic feature set, semantic constraints are converted into executable parameter constraint relationships, and parameter variables used to control the size, position, or shape of geometric features are extracted; the parameter constraint relationships are associated with the corresponding parameter variables, geometric features, and semantic constraints respectively to construct a parameter constraint model; Based on the parameter constraint model, the driving and dependency relationships between the parameter constraints are analyzed, and the execution priority is determined according to the importance of the semantic constraints. An adaptive parameter modeling sequence containing the geometric feature generation order and parameter constraint propagation path is generated. Based on the adaptive parametric modeling sequence, the corresponding geometric features are generated step by step in the CAD modeling environment by calling parameter variables, and constraint consistency is checked according to the parameter constraint relationship. When a constraint conflict is detected, the value range of the conflicting parameter variables, parameter constraint relationship or geometric feature generation order is adjusted according to the associated semantic constraints, and subsequent geometric feature generation continues to be executed to obtain the CAD drawing. The structural integrity and constraint satisfaction of CAD drawings are verified, and the structured semantic feature set and parameter constraint model are updated based on the verification results to generate optimized parameter modeling results.

[0006] Secondly, a large-model-driven intelligent parametric modeling system for generating CAD drawings, the system comprising: The requirements parsing module is used to obtain design requirements information, parse the natural language descriptions, multi-source engineering data and historical design samples in the design requirements information, identify the semantic entities representing the objects to be modeled and the semantic constraints representing the design requirements; generate a structured semantic feature set based on the semantic entities, semantic constraints and the relationship between them, and establish an initial mapping relationship between the structured semantic feature set and the geometric feature types. The parameter modeling module is used to convert semantic constraints into executable parameter constraint relationships based on a set of structured semantic features, and extract parameter variables used to control the size, position, or shape of geometric features; the parameter constraint relationships are associated with the corresponding parameter variables, geometric features, and semantic constraints to construct a parameter constraint model. The sequence generation module is used to parse the driving and dependency relationships between various parameter constraints based on the parameter constraint model, determine the execution priority according to the importance of semantic constraints, and generate an adaptive parameter modeling sequence containing the geometric feature generation order and parameter constraint propagation path. The drawing generation module is used to generate corresponding geometric features step by step in the CAD modeling environment by calling parameter variables according to the adaptive parameter modeling sequence, and to perform constraint consistency detection according to parameter constraint relationships. When constraint conflicts are detected, the value range of conflicting parameter variables, parameter constraint relationships or geometric feature generation order are adjusted according to the associated semantic constraints, and subsequent geometric feature generation is continued to obtain CAD drawings. The structural optimization module is used to verify the structural integrity and constraint satisfaction of CAD drawings, and update the structured semantic feature set and parameter constraint model based on the verification results, generating optimized parameter modeling results.

[0007] The above-described solution of the present invention has at least the following beneficial effects: First, by parsing natural language descriptions, multi-source engineering data, and historical design samples, semantic entities and semantic constraints are identified, and a set of structured semantic features is generated. This enables unstructured design requirements to be converted into data that can be used for CAD parametric modeling, reducing the need for manual conversion of design intent into modeling parameters.

[0008] Furthermore, by converting semantic constraints into parametric constraints and associating these parametric constraints with parameter variables, geometric features, and semantic constraints, a parametric constraint model is constructed. This enables a traceable relationship between design intent, parameter control, and geometric generation, thus solving the problem of semantics and geometry being difficult to correspond in complex irregular parts.

[0009] Furthermore, by analyzing the driving and dependency relationships between parameter constraints, an adaptive parametric modeling sequence containing the geometric feature generation order and parameter constraint propagation path is generated, so that the modeling process no longer depends on a fixed feature tree and can organize the modeling order according to design constraints.

[0010] Furthermore, during the generation of geometric features, constraint conflicts are identified through constraint consistency detection, and the value range of conflict parameter variables, parameter constraint relationships, or geometric feature generation order are adjusted according to the associated semantic constraints, so that conflict handling can be corrected around the original design requirements, reducing repeated reconstruction caused by constraint conflicts.

[0011] Finally, by verifying the structural integrity and constraint satisfaction of the CAD drawings, and updating the structured semantic feature set and parameter constraint model based on the verification results, the updated semantic mapping relationship and parameter constraint relationship can be reused in subsequent similar design tasks, thereby improving the continuity and consistency of the complex structure drawing generation process. Attached Figure Description

[0012] Figure 1 This is a flowchart of a large-model-driven intelligent generation parametric modeling method for CAD drawings provided in an embodiment of the present invention. Detailed Implementation

[0013] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0014] like Figure 1 As shown, embodiments of the present invention propose a large-model-driven intelligent generation parametric modeling method for CAD drawings, the method comprising: The system acquires design requirement information, parses the natural language descriptions, multi-source engineering data, and historical design samples in the design requirement information, identifies semantic entities representing the objects to be modeled and semantic constraints representing design requirements; generates a structured semantic feature set based on semantic entities, semantic constraints, and the relationship between them, and establishes an initial mapping relationship between the structured semantic feature set and geometric feature types. Based on the structured semantic feature set, semantic constraints are converted into executable parameter constraint relationships, and parameter variables used to control the size, position, or shape of geometric features are extracted; the parameter constraint relationships are associated with the corresponding parameter variables, geometric features, and semantic constraints respectively to construct a parameter constraint model; Based on the parameter constraint model, the driving and dependency relationships between the parameter constraints are analyzed, and the execution priority is determined according to the importance of the semantic constraints. An adaptive parameter modeling sequence containing the geometric feature generation order and parameter constraint propagation path is generated. Based on the adaptive parametric modeling sequence, the corresponding geometric features are generated step by step in the CAD modeling environment by calling parameter variables, and constraint consistency is checked according to the parameter constraint relationship. When a constraint conflict is detected, the value range of the conflicting parameter variables, parameter constraint relationship or geometric feature generation order is adjusted according to the associated semantic constraints, and subsequent geometric feature generation continues to be executed to obtain the CAD drawing. The structural integrity and constraint satisfaction of CAD drawings are verified, and the structured semantic feature set and parameter constraint model are updated based on the verification results to generate optimized parameter modeling results.

[0015] In this embodiment of the invention, by acquiring design requirement information and parsing natural language descriptions, multi-source engineering data, and historical design samples, the semantic entities corresponding to the objects to be modeled and the semantic constraints corresponding to the design requirements are identified. Then, a structured semantic feature set is generated based on the semantic entities, semantic constraints, and the relationship between them, and an initial mapping relationship is established between the structured semantic feature set and geometric feature types. This enables unstructured design requirements to be converted into unified description data that can be called by the modeling process, solving the problem that design requirements are difficult to directly enter the CAD modeling process and improving the accuracy of the correspondence between design input and modeling objects.

[0016] By converting semantic constraints into executable parameter constraint relationships based on a structured set of semantic features, extracting parameter variables that control the size, position, or shape of geometric features, and then associating the parameter constraint relationships with parameter variables, geometric features, and semantic constraints respectively, a parameter constraint model is constructed. This creates a traceable association structure between design requirements, parameter control, and geometric generation, solving the problems of difficulty in locating the affected objects after parameter modification and difficulty in synchronously transmitting design requirements, thereby improving the consistency of association during parameter adjustment.

[0017] By analyzing the driving and dependency relationships between various parameter constraints according to the parameter constraint model, and determining the execution priority based on the importance of semantic constraints, an adaptive parameter modeling sequence containing the geometric feature generation order and parameter constraint propagation path is generated. This allows the modeling order to be adjusted as the constraint relationship changes, solving the problems of feature failure, reference failure, and repeated adjustments caused by unreasonable modeling order under the fixed feature tree method, and improving the modeling coherence in the process of generating complex models.

[0018] By calling parameter variables step by step in the CAD modeling environment according to the adaptive parameter modeling sequence, corresponding geometric features are generated, and constraint consistency is checked according to parameter constraint relationships during the generation process. When constraint conflicts are detected, the value range of conflicting parameter variables, parameter constraint relationships or geometric feature generation order are adjusted according to the associated semantic constraints, and then the subsequent geometric feature generation is continued. This enables the modeling process to have conflict identification and adaptive correction capabilities, solving the problem of needing to remodel or manually check item by item after constraint conflicts, and improving the continuity of the drawing generation process and the stability of the results.

[0019] By verifying the structural integrity and constraint satisfaction of CAD drawings, and updating the structured semantic feature set and parameter constraint model based on the verification results, subsequent similar design tasks can call the updated constraint relationships and modeling paths, solving the problem of repeated rule setting in repetitive design tasks and improving the reuse efficiency of subsequent modeling tasks.

[0020] In a preferred embodiment of the present invention, design requirement information is obtained, and the natural language description, multi-source engineering data, and historical design samples in the design requirement information are parsed to identify semantic entities representing the object to be modeled and semantic constraints representing design requirements, including: The system acquires user-input text requirements, speech-to-text conversion, table parameter files, existing CAD model files, and historical design case files, and categorizes and stores them according to text data, numerical data, and model data to form the original design requirement data. The text data undergoes word segmentation, terminology standardization, and unit unification to generate standard text data.

[0021] Entity recognition is performed on standard text data to extract part names, structural parts, feature names, and reference objects, generating semantic entities. These semantic entities include mounting holes, reinforcing ribs, support surfaces, center lines, reference surfaces, and connecting grooves. Constraint recognition is also performed on the standard text data to extract dimensional requirements, positional requirements, symmetry requirements, and functional requirements, generating semantic constraints. The dimensional requirements include a hole diameter of 10mm and a plate thickness of 6mm; the positional requirements include a hole center distance of 20mm from the edge; the symmetry requirements include left-right symmetrical arrangement; and the functional requirements include no deformation under a load of 500N.

[0022] The system matches field names and values ​​in the table parameter file, mapping field names to corresponding semantic entities or semantic constraints, and writing field values ​​into the corresponding constraint content. It then performs feature tree parsing on existing CAD model files, extracting hole features, extrusion features, fillet features, and their historical parameter values. The extracted results are then written into the original design requirement data as supplementary semantic entities and constraints.

[0023] Historical design case files are categorized, and design templates consistent with the current part category are extracted. The part category is determined based on the part name and structural features. When the part name is consistent and the number of matching structural features reaches 3, they are identified as the same type of part. Constraint expressions from similar design cases are extracted as reference constraints, and parsing results are generated based on the pointing relationship between semantic entities and semantic constraints.

[0024] In a preferred embodiment of the present invention, a structured semantic feature set is generated based on semantic entities, semantic constraints, and the relationship between them, and an initial mapping relationship is established between the structured semantic feature set and geometric feature types, including: Extract semantic entities, semantic constraints, and pointing relationships from the parsed results, and create structured records according to the field structure of "entity number - constraint type - constraint content - target object". Write multiple semantic constraints corresponding to the same semantic entity into the same entity record. For example, write "mounting hole 1 - size constraint - hole diameter 10mm" and "mounting hole 1 - position constraint - distance from left edge 20mm" into the structured record of mounting hole 1 to generate a set of structured semantic features.

[0025] An initial mapping relationship between semantic entities and geometric feature types is established by matching semantic entity names against a preset feature type library. This preset feature type library consists of commonly used CAD modeling features, including hole features, extrusion features, cut features, fillet features, chamfer features, array features, and surface features. When a semantic entity is a mounting hole, positioning hole, or through hole, it is mapped to a hole feature; when a semantic entity is a boss, base plate, or support block, it is mapped to an extrusion feature; when a semantic entity is a groove or clearance opening, it is mapped to a cut feature; and when a semantic entity is a rounded edge, it is mapped to a fillet feature.

[0026] When a semantic entity matches multiple geometric feature types, the semantic constraints associated with that semantic entity are extracted for secondary judgment. When the semantic constraint includes a depth parameter, the semantic entity is mapped to a hole feature; when the semantic constraint includes a material removal description, the semantic entity is mapped to a cutting feature; when the semantic constraint includes an additive description, the semantic entity is mapped to a stretching feature. The structured records are combined and stored with their corresponding geometric feature types to generate a set of structured semantic features with initial mapping relationships.

[0027] In a preferred embodiment of the present invention, the structural integrity and constraint satisfaction of the CAD drawings are verified, and the structured semantic feature set and parameter constraint model are updated according to the verification results to generate optimized parameter modeling results, including: Extract all geometric features, dimensions, positional relationships, and assembly interface information from the generated CAD drawings to form verification input data. Perform structural integrity verification on the verification input data to determine whether all the geometric features required in the design requirements exist. When the geometric feature type, quantity, and associated objects corresponding to the semantic entity are consistent with the corresponding geometric features in the CAD drawing, the structural integrity verification is considered successful. When the required number of mounting holes is 4 and the number of mounting holes identified in the CAD drawing is 3, a structural deficiency is identified.

[0028] Constraint satisfaction verification is performed on the input data by comparing the actual parameter values ​​in the CAD drawings with the target parameter values ​​in the parameter constraint model. When the dimensional deviation does not exceed 0.1 mm and the positional deviation does not exceed 0.2 mm, the corresponding parameter is deemed to satisfy the constraint. For example, when the target hole diameter is 10 mm and the measured value is 10.05 mm, the hole diameter is deemed to satisfy the constraint; when the target hole center distance is 20 mm and the measured value is 20.35 mm, the hole center distance is deemed not to satisfy the constraint.

[0029] When structural gaps exist, an incomplete marker is added to the structured semantic feature set for the missing semantic entity, and the type of missing feature is recorded. When constraints are not met, the correction record for the corresponding parameter variable is updated in the parameter constraint model, and the actual deviation value, target value, and correction direction are written; when the actual value is greater than the target value, the correction direction is recorded as decreasing; when the actual value is less than the target value, the correction direction is recorded as increasing. Based on the incomplete marker and correction record, the structured semantic feature set and parameter constraint model are regenerated to generate the optimized parameter modeling result for the next modeling call.

[0030] In a preferred embodiment of the present invention, based on a structured semantic feature set, semantic constraints are converted into executable parameter constraint relationships, and parameter variables used to control the size, position, or shape of geometric features are extracted; the parameter constraint relationships are respectively associated with the corresponding parameter variables, geometric features, and semantic constraints to construct a parameter constraint model, including: The semantic constraints in the structured semantic feature set are parsed to generate size semantic constraints, positional semantic constraints and functional semantic constraints; Based on the semantic constraints of dimensions, the value restriction relationships of parameter variables are generated, and the parameter constraint relationships of dimension class are obtained; Based on positional semantic constraints, positional constraint relationships between geometric features are generated. These positional constraint relationships include symmetry relationships, alignment relationships, spacing relationships, or reference positioning relationships, resulting in positional parameter constraint relationships. Based on the functional semantic constraints, the functional constraint relationship of the geometric features is generated, and the functional class parameter constraint relationship is obtained. Based on the geometric feature type, extract the corresponding size parameter variables, position parameter variables, and shape parameter variables, establish the calling relationship between parameter variables and geometric features, and generate the corresponding results of parameter variables and geometric features; Based on the constraints of size parameters, position parameters, function parameters, and the correspondence between parameter variables and geometric features, a unique association identifier is generated for each parameter constraint relationship, and the unique association identifier is written into the association record of the corresponding semantic constraint, parameter variable, and geometric feature. Based on each associated record, semantic constraints, parameter constraint relationships, parameter variables, and geometric features are associated and organized to generate a parameter constraint model with bidirectional traceability paths.

[0031] In this embodiment of the invention, semantic constraints in the structured semantic feature set are parsed to generate size semantic constraints, positional semantic constraints, and functional semantic constraints. Different types of semantic constraints are then converted into corresponding size-type parameter constraint relationships, positional parameter constraint relationships, and functional parameter constraint relationships. Size parameter variables, positional parameter variables, and shape parameter variables are extracted in combination with geometric feature types to establish the calling relationship between parameter variables and geometric features. Subsequently, a unique association identifier is generated for each parameter constraint relationship and written into the association record of semantic constraints, parameter variables, and geometric features to form a parameter constraint model with a bidirectional traceability path. This enables a unified association structure between design requirements, parameter control, and geometric objects, solving the problem of scattered storage of design semantics and modeling parameters and difficulty in locating the scope of influence after modification. This improves the accuracy of parameter adjustment, constraint tracking, and subsequent conflict location.

[0032] In a preferred embodiment of the present invention, type parsing is performed on the semantic constraints in the structured semantic feature set, including: Extract semantic constraints from the structured semantic feature set, and read the constraint description, constraint object and constraint keywords corresponding to the semantic constraints; When the constraint description includes length, width, height, thickness, diameter, radius, angle, or numerical unit, the corresponding semantic constraint will be marked as a dimension semantic constraint. When the constraint description includes center distance, symmetry, alignment, parallelism, perpendicularity, datum plane, edge distance, or coordinate position, the corresponding semantic constraint is marked as a positional semantic constraint. When the constraint description includes load-bearing, weight reduction, avoidance, assembly, support, sealing, or strength maintenance, the corresponding semantic constraint will be marked as a functional semantic constraint. When the same semantic constraint contains both size content and functional content, the semantic constraint is written into the size semantic constraint and the functional semantic constraint respectively, and the source association is preserved to generate the semantic constraint classification result.

[0033] In a preferred embodiment of the present invention, the dimension-type parameter constraint relationship is obtained, including: Extract the dimension object, dimension type, target value, and unit information from the dimension semantic constraints to generate dimension parsing results; Based on the dimension analysis results, assign dimension parameter variables to the dimension objects, where the hole diameter corresponds to the diameter parameter variable, the plate thickness corresponds to the thickness parameter variable, and the fillet corresponds to the radius parameter variable; Write the target value and unit information into the value constraint relationship of the dimension parameter variable to generate dimension parameter constraint relationship; When the dimensional semantic constraint is "mounting hole diameter 10mm", the value of the generated mounting hole diameter parameter variable is limited to a diameter equal to 10mm; when the dimensional semantic constraint is "plate thickness not less than 6mm", the value of the generated plate thickness parameter variable is limited to a thickness greater than or equal to 6mm.

[0034] In a preferred embodiment of the present invention, the positional parameter constraint relationship is obtained, including: Extract the location object, reference object, and location description from the location semantic constraints to generate the location resolution result; When the position description is symmetrical, a mirror position constraint relationship is established based on the position object and the reference object to generate a symmetrical relationship; When the position description is alignment, the position constraint relationship is established based on the position object and the reference object to establish collinear, coplanar, or center alignment, and the alignment relationship is generated. When the location is described as spacing, extract the distance values ​​between the location objects, establish distance constraints, and generate spacing relationships; When the location description is based on a reference point, extract the reference plane, reference line, or reference point, and establish the positioning relationship between the location object and the reference object to generate the reference positioning relationship; When the positional semantic constraint is "the two mounting holes are symmetrical about the center line", a symmetrical relationship between the two mounting holes and the center line is generated; when the positional semantic constraint is "the center of the hole is 20mm from the left edge", a spacing relationship is generated where the distance between the center point of the mounting hole and the left edge is equal to 20mm.

[0035] In a preferred embodiment of the present invention, the functional class parameter constraint relationship is obtained, including: Extract the functional objectives, objects of action, and constraints from the functional semantic constraints to generate functional analysis results; Based on the functional analysis results, determine the geometric features associated with the functional objectives. Among them, the load-bearing function is associated with the support surface, stiffeners or connecting holes; the weight reduction function is associated with the cut-off features, groove features or thinning areas; and the assembly function is associated with the mounting holes, positioning grooves or mating surfaces. Establish functional constraint relationships based on the target and limiting conditions to obtain functional class parameter constraint relationships; When the functional semantic constraint is "reducing weight while maintaining local strength", the weight reduction target is associated with the cutting feature, the local strength maintenance target is associated with the stiffener or support surface, and a functional class parameter constraint relationship is generated that the cutting area must not destroy the stiffener connection area. When the functional semantic constraint is "satisfy bolt assembly", the assembly target is associated with the mounting hole, and functional class parameter constraint relationships are generated to ensure that the mounting hole diameter, position and mating object meet the bolt assembly requirements.

[0036] In a preferred embodiment of the present invention, generating the correspondence between parameter variables and geometric features includes: Extract the initial mapping relationship between the set of structured semantic features and the geometric feature types, and determine the geometric feature type corresponding to each semantic entity; When the geometric feature type is a hole feature, the hole diameter, hole depth, and hole center coordinates are extracted as parameter variables; when the geometric feature type is a stretching feature, the stretching length, cross-sectional width, cross-sectional height, and stretching direction are extracted as parameter variables; when the geometric feature type is a fillet feature, the fillet radius and fillet boundary are extracted as parameter variables. Label the parameter variables that control the size as size parameter variables, label the parameter variables that control the coordinates, direction and reference as position parameter variables, and label the parameter variables that control the contour, curvature or boundary shape as shape parameter variables. Establish the calling relationship between parameter variables and corresponding geometric features, so that the corresponding parameter variables are extracted as modeling input when the geometric features are generated; When a mounting hole is mapped to a hole feature, the hole diameter parameter variable, hole depth parameter variable, and hole center coordinate parameter variable are written into the calling relationship of the hole feature, generating the corresponding result of parameter variables and geometric features.

[0037] In a preferred embodiment of the present invention, generating a parameter constraint model with a bidirectional traceability path includes: Extract each associated record and group semantic constraints, parameter constraint relationships, parameter variables, and geometric features according to the unique association identifier; In each group, establish a forward tracing path from semantic constraints to parameter constraints, from parameter constraints to parameter variables, and from parameter variables to geometric features; In each group, establish a reverse tracing path from geometric features to parameter variables, from parameter variables to parameter constraint relationships, and from parameter constraint relationships to semantic constraints; When parameter variables are adjusted, the corresponding semantic constraints are located according to the reverse tracing path, and the affected geometric features are determined according to the tracing path from the parameter variables to the geometric features. The forward tracing path, the reverse tracing path, and the constraint relationships of each parameter are stored together to generate a parameter constraint model with bidirectional tracing paths.

[0038] In a preferred embodiment of the present invention, the process of generating an adaptive parameter modeling sequence includes: Based on the parameter constraint model, the parameter constraint relationships are extracted, and the interaction relationships between each parameter constraint relationship are analyzed to generate driving parameter constraint relationships, driven parameter constraint relationships, and coupled parameter constraint relationships. Based on the driving parameter constraints, the baseline geometric feature is determined as the starting point for modeling; based on the driven parameter constraints, the subsequent geometric features generated based on the modeling results of the preceding geometric features are determined, where the baseline geometric feature is the first preceding geometric feature; based on the coupling parameter constraints, the associated geometric features that need to be updated synchronously are determined, and the geometric feature dependency results are generated. Based on the type of semantic constraints, the execution priority of each parameter constraint relationship is determined by the order of the modeling control effects of size semantic constraints, symmetry semantic constraints, and functional semantic constraints, and parameter constraint priority results are generated. Based on the parameter constraint priority results, the preceding geometric features, subsequent geometric features, and associated geometric features in the geometric feature dependency results are sorted to generate the geometric feature generation order. Based on the generation order of geometric features, identify the parameter transfer relationship between the preceding and subsequent geometric features, and generate the parameter constraint propagation path based on the parameter transfer relationship; An adaptive parameter modeling sequence is generated based on the geometric feature generation order and parameter constraint propagation path.

[0039] In this embodiment of the invention, by extracting the parameter constraint relationships in the parameter constraint model and analyzing the interaction relationships between each parameter constraint relationship, driving parameter constraint relationships, driven parameter constraint relationships, and coupled parameter constraint relationships are generated. Then, based on different interaction relationships, the baseline geometric features, subsequent geometric features, and associated geometric features are determined. Combined with the order of semantic constraints on the modeling control, the parameter constraint priority results are generated, and the geometric feature generation order and parameter constraint propagation path are formed accordingly. Finally, an adaptive parameter modeling sequence is generated, which enables the modeling process to be dynamically organized according to design requirements and constraint dependencies. This solves the problem that a fixed modeling order is difficult to adapt to complex associated structures and that the overall process needs to be repeatedly adjusted after local modifications. It improves the order rationality and parameter transmission coherence in the process of generating complex drawings.

[0040] In a preferred embodiment of the present invention, generating driving parameter constraint relationships, driven parameter constraint relationships, and coupling parameter constraint relationships includes: Extract the parameter constraint relationships, parameter variables, geometric features, and unique association identifiers from the parameter constraint model to generate constraint data to be parsed; Based on the constraint data to be parsed, identify the parameter constraint relationship that can be determined without relying on other parameter variables, and generate the driving parameter constraint relationship, which includes the datum dimension, datum plane position or initial contour dimension; Based on the constraint data to be parsed, identify the parameter constraint relationships that need to call other parameter variables or geometric feature results to determine the values, and generate driven parameter constraint relationships. The driven parameter constraint relationships include the distance relationship between the hole position and the reference plane, and the connection relationship between the stiffener and the base plate. Based on the constraint data to be parsed, identify parameter constraint relationships that jointly call the same parameter variable or jointly restrict the same geometric feature, and generate coupled parameter constraint relationships. The coupled parameter constraint relationships include multiple mounting holes sharing hole spacing parameters and left and right reinforcing ribs sharing a symmetrical baseline.

[0041] In a preferred embodiment of the present invention, a baseline geometric feature is determined as the starting point for modeling based on the driving parameter constraint relationship; subsequent geometric features generated based on the modeling results of preceding geometric features are determined based on the driven parameter constraint relationship, wherein the baseline geometric feature is the first preceding geometric feature; and associated geometric features that need to be updated synchronously are determined based on the coupling parameter constraint relationship, generating geometric feature dependency results, including: Extract the geometric features corresponding to the driving parameter constraint relationship, and determine the geometric features that provide the reference plane, reference line, initial outer contour or main structure as the reference geometric features; When the base plate stretching feature provides a mounting hole positioning plane, the base plate stretching feature is determined as the reference geometry feature; Extract the geometric features corresponding to the driving parameter constraint relationship, and determine the geometric features that depend on the previous geometric features for localization, connection or removal as the subsequent geometric features; When the mounting hole is extracted from the upper surface of the base plate as the positioning surface, the mounting hole is determined as a subsequent geometric feature generated after the base plate stretching feature; Extract the geometric features corresponding to the coupling parameter constraint relationship, and identify the geometric features with shared parameter variables or controlled by the same position constraint relationship as associated geometric features; When the left and right stiffeners share a symmetrical relationship along the centerline, the left and right stiffeners are identified as associated geometric features that need to be updated synchronously. Based on the dependencies between baseline geometric features, subsequent geometric features, and associated geometric features, geometric feature dependency results are generated.

[0042] In a preferred embodiment of the present invention, generating parameter constraint priority results includes: Extract the semantic constraint type corresponding to each parameter constraint relationship, and determine the control attribute of the corresponding parameter constraint relationship based on the semantic constraint type; The parameter constraint relationship corresponding to the functional semantic constraint is determined as the first priority, the parameter constraint relationship corresponding to the size semantic constraint is determined as the second priority, and the parameter constraint relationship corresponding to the position semantic constraint is determined as the third priority. When the positional semantic constraint contains a symmetric relationship, the parameter constraint relationship corresponding to the symmetric relationship is determined as the preceding positional constraint in the third priority. When the same geometric feature is associated with both functional semantic constraints and dimensional semantic constraints, the parameter constraint relationship corresponding to the functional semantic constraints is retained, and the corresponding parameter variables are then limited according to the dimensional semantic constraints. When the same geometric feature is associated with both dimensional semantic constraints and positional semantic constraints, the dimensional parameter variables are determined first, and then the positional parameter variables are determined. Generate parameter constraint priority results based on the first priority, second priority, third priority and preceding position constraints.

[0043] In a preferred embodiment of the present invention, the geometric feature generation sequence includes: Extract the baseline geometric features, subsequent geometric features, and associated geometric features from the geometric feature dependency results, and extract the parameter constraint priority results; Arrange the baseline geometric features at the beginning of the geometric feature generation sequence; sort the subsequent geometric features according to the execution priority of the parameter constraint relationships corresponding to the subsequent geometric features; When multiple subsequent geometric features have the same execution priority, the subsequent geometric feature that will be referenced by other geometric features will be placed before the subsequent geometric feature that is not referenced. When there are coupling parameter constraints between associated geometric features, the associated geometric features are arranged into continuous generation units, and within the continuous generation units, they are arranged according to the order in which shared parameter variables are called. For example, if the base plate stretching feature is the base geometric feature, the mounting hole is referenced by the positioning slot, and the positioning slot is referenced by the reinforcing rib, then the geometric feature generation order is base plate stretching feature, mounting hole, positioning slot, and reinforcing rib; the geometric feature generation order is generated according to the sorting result.

[0044] In a preferred embodiment of the present invention, the parameter transfer relationship between preceding and subsequent geometric features is identified according to the geometric feature generation order, and a parameter constraint propagation path is generated based on the parameter transfer relationship, including: Extract the preceding geometric features, the following geometric features, and the parameter constraints between them in the geometric feature generation sequence; When a subsequent geometric feature calls the size parameter variables, position parameter variables, shape parameter variables, or generates a surface from a preceding geometric feature, it is determined that there is a parameter transfer relationship between the preceding and subsequent geometric features. When the mounting hole uses the upper surface of the base plate as the positioning surface and uses the base plate length parameter to determine the hole spacing range, the parameter transfer relationship from the base plate stretching feature to the mounting hole is established. When the stiffener calls the mounting hole avoidance area and base plate thickness parameters, establish the parameter transfer relationship from the mounting hole and base plate tension characteristics to the stiffener; Connect the parameter propagation relationships according to the geometric feature generation order to generate parameter constraint propagation path; When the same subsequent geometric feature receives parameter variables from multiple preceding geometric features, a multi-source input path is established for the subsequent geometric feature in the parameter constraint propagation path.

[0045] In a preferred embodiment of the present invention, an adaptive parameter modeling sequence is generated based on the geometric feature generation order and parameter constraint propagation path, including: Extract the geometric feature generation order and parameter constraint propagation path, and configure corresponding parameter variables, parameter constraint relationships and input sources for each geometric feature in the geometric feature generation order; According to the order of geometric feature generation, the generation operations of each geometric feature are written into the modeling sequence in sequence; Write the parameter constraint propagation path into the input field of the corresponding geometric feature so that the parameter variables passed by the preceding geometric feature are extracted before the subsequent geometric feature is generated. When associated geometric features in a continuously generated cell share the same parameter variable, the shared parameter variable is written into the common input field of the continuously generated cell. When the parameter constraint relationship is updated, the affected subsequent geometric features are determined according to the parameter constraint propagation path, and the input fields of the corresponding geometric features are updated. Based on the generation operations of each geometric feature, parameter variables, parameter constraints, and input sources, an adaptive parametric modeling sequence is generated.

[0046] In a preferred embodiment of the present invention, according to the adaptive parameter modeling sequence, corresponding geometric features are generated step by step in the CAD modeling environment by calling parameter variables, and constraint consistency detection is performed according to parameter constraint relationships; when a constraint conflict is detected, the value range of the conflicting parameter variables, parameter constraint relationships, or geometric feature generation order are adjusted according to the associated semantic constraints, and subsequent geometric feature generation continues to be executed to obtain a CAD drawing, including: Based on the adaptive parameter modeling sequence, the geometric feature generation order and parameter constraint propagation path are extracted, and the corresponding parameter variables are called to perform geometric feature generation operations according to the geometric feature generation order to generate the stage geometric model; Based on the stage geometric model, extract the parameter constraint relationships associated with the current geometric features, and detect the consistency of size, position, topological connectivity, and feature dependency, generating constraint detection results; Based on the constraint detection results, determine whether there is a constraint conflict; when a constraint conflict exists, locate the corresponding semantic constraints, conflict parameter variables and conflict parameter constraint relationships based on the associated records in the parameter constraint model, and generate a conflict location result based on the conflict detection items. Based on the conflict location results, determine the conflict type; when the conflict type is size conflict, adjust the value range of the conflict parameter variables according to the corresponding semantic constraints; when the conflict type is topological conflict, adjust the conflict parameter constraint relationship according to the corresponding semantic constraints; when the conflict type is dependency conflict, adjust the geometric feature generation order according to the corresponding semantic constraints, and generate conflict correction results; among them, position consistency anomalies are classified into topological conflicts or dependency conflicts according to the cause of the anomaly. Based on the conflict correction results, continue with the remaining geometric feature generation operation to obtain the CAD drawing.

[0047] In this embodiment of the invention, the geometric feature generation order and parameter constraint propagation path are extracted according to the adaptive parameter modeling sequence. The corresponding parameter variables are called to perform geometric feature generation operations according to the generation order to obtain a stage geometric model. Then, the stage geometric model is tested for size consistency, position consistency, topological connection consistency and feature dependency consistency. When there is a constraint conflict, the corresponding semantic constraint, conflicting parameter variable and conflicting parameter constraint relationship are located according to the associated records in the parameter constraint model. The parameter variable value range, parameter constraint relationship or geometric feature generation order are adjusted according to the conflict type. Then, the remaining geometric feature generation operations are continued to obtain CAD drawings. This enables the modeling process to generate, verify and correct at the same time, solving the problems of modeling interruption caused by constraint conflicts, difficulty in tracing the source of problems and overall rework, and improving the continuity of the drawing generation process and the consistency of the results.

[0048] In a preferred embodiment of the present invention, generating a stage geometric model includes: Extract the geometric feature generation order, parameter constraint propagation path, parameter variables and parameter constraint relationships corresponding to each geometric feature from the adaptive parameter modeling sequence, and generate the modeling data to be executed. The target geometric feature to be generated is determined according to the geometric feature generation order, and the size parameter variables, position parameter variables and shape parameter variables corresponding to the target geometric feature are extracted. When the target geometric feature has a prior input source, extract the parameter variables, reference plane, boundary contour or positioning object passed by the prior geometric feature according to the parameter constraint propagation path, and write them into the modeling input data of the target geometric feature; Constraint verification is performed on the parameter variables corresponding to the target geometric features. When the size parameter variables, position parameter variables, and shape parameter variables satisfy the corresponding parameter constraint relationships, the geometric feature generation operation is performed. When the target geometric feature is a hole feature, the hole feature is generated based on the hole diameter parameter variable, hole depth parameter variable, and hole center coordinate parameter variable; for example, when the hole diameter is 10mm, the hole depth is 15mm, and the hole center coordinate is (20mm, 30mm), the corresponding mounting hole is generated in the current model; When the target geometric feature is an extrusion feature, the extrusion feature is generated based on the cross-sectional profile parameter variables and the extrusion height parameter variables; for example, when the cross-sectional profile is a 100mm×60mm rectangle and the extrusion height is 8mm, a base plate solid is generated. When the target geometry is a rounded corner feature, the rounded corner feature is generated based on the rounded corner radius parameter variable and the target boundary; for example, when the rounded corner radius is 3mm, a rounded corner transition with a radius of 3mm is generated at the target boundary; After generating the target geometric features, the entities, faces, edges, coordinate datums, and updated parameter variables generated by the target geometric features are written into the parameter constraint propagation path as the input source for subsequent geometric features. All generated geometric features are combined to generate a stage geometric model.

[0049] In a preferred embodiment of the present invention, based on the conflict correction result, the remaining geometric feature generation operation is continued to obtain a CAD drawing, including: Extract the corrected parameter variables, corrected parameter constraint relationships, or corrected geometric feature generation order from the conflict correction results, and update the corresponding fields in the adaptive parameter modeling sequence to generate the corrected modeling sequence; The remaining geometric features that have not yet been executed when the conflict occurred are determined based on the corrected modeling sequence, and the remaining geometric features are arranged in the order of generation of the corrected geometric features to generate the remaining execution sequence; When the conflict correction result includes parameter variable adjustment, the corrected parameter variable is written into the modeling input data of the corresponding remaining geometric feature; for example, when the aperture parameter variable is corrected from 10mm to 9.5mm, 9.5mm is written into the modeling input data of the corresponding aperture feature. When the conflict correction result includes adjustments to parameter constraint relationships, the positioning relationship, connection relationship, or size limit of the remaining geometric features is determined based on the corrected parameter constraint relationship, and written into the modeling input data of the remaining geometric features; for example, when the hole center distance from the edge is corrected from 20mm to 18mm, 18mm is written into the corresponding positioning constraint. When the conflict correction result includes an adjustment to the geometric feature generation order, the remaining geometric feature generation operations are executed according to the corrected order; for example, if the original order was to generate mounting holes first and then generate reinforcing ribs, and the corrected order is to generate reinforcing ribs first and then generate mounting holes, the corrected order is executed. Execute the stretching feature, cut feature, hole feature, fillet feature, or surface feature generation operations one by one according to the remaining execution sequence, and update the parameter constraint propagation path after each generation is completed; Once all geometric features in the remaining execution sequence have been executed, the stage geometric model is combined with the geometric features generated by the remaining execution sequence to generate a complete CAD drawing; Save the complete CAD drawing as a preset drawing format file, which includes DWG, DXF, or STEP format. The drawing format is determined based on the compatibility format of the external calling software.

[0050] In a preferred embodiment of the present invention, based on the stage geometric model, parameter constraint relationships associated with the current geometric features are extracted, and size consistency, positional consistency, topological connectivity consistency, and feature dependency consistency are detected to generate constraint detection results, including: Based on the stage geometric model, the size parameters, position parameters, and shape parameters of the current geometric feature are identified, and the parameter constraint relationships associated with the current geometric feature are extracted to generate the current detection data; Based on the current detection data, the actual size status of the current geometric features is compared with the constraint relationship of the size parameters to identify whether there are size over-limits, size missing or size linkage failures, and generate size consistency detection results; Based on the current detection data, the actual position state of the current geometric features is compared with the position class parameter constraint relationship to identify whether there is symmetry failure, alignment deviation, spacing mismatch or reference offset, and generate position consistency detection results; Based on the current detection data, identify the connection status between the current geometric feature and its adjacent geometric features, match it with the connection requirements in the parameter constraint relationship, identify whether there are connection interruptions, cross interference or topological gaps, and generate topological connection consistency detection results; Based on the current detection data, identify the reference status of the current geometric feature to the preceding geometric feature and the call status of the subsequent geometric feature to the current geometric feature, and match them with the dependency relationships in the adaptive parameter modeling sequence to identify whether there are invalid references, circular dependencies or order misalignments, and generate feature dependency consistency detection results. Constraint detection results are generated based on the size consistency detection results, position consistency detection results, topology connection consistency detection results, and feature dependency consistency detection results.

[0051] In this embodiment of the invention, the size parameters, position parameters, and shape parameters of the current geometric features are identified based on the stage geometric model, and the parameter constraint relationships associated with the current geometric features are extracted to generate current detection data. Then, the actual size state is compared with the size-type parameter constraint relationship, the actual position state is compared with the position-type parameter constraint relationship, the connection state is matched with the connection requirements, and the feature reference state is matched with the dependency relationship in the modeling sequence to generate size consistency detection results, position consistency detection results, topology connection consistency detection results, and feature dependency consistency detection results, and form constraint detection results. This allows the geometric features to complete multi-dimensional verification during the generation stage, solving the problem of discovering size deviations, position mismatches, connection errors, or reference failures only after modeling is completed, improving the timeliness of problem discovery and the completeness of detection coverage.

[0052] In a preferred embodiment of the present invention, generating a size consistency detection result includes: Extract the actual size parameters from the current detection data. The actual size parameters include length, width, height, thickness, hole diameter, hole depth, corner radius, or angle. Extract the target dimension value, dimension range, or dimension linkage relationship from the dimension-type parameter constraint relationship to generate a dimension comparison benchmark; The actual size parameters are compared with the size comparison benchmark item by item. When the actual size parameter exceeds the target size value or size range, it is marked as size over-limit. For example, when the target hole diameter is 10mm, the allowable deviation is ±0.1mm, and the actual hole diameter is 10.25mm, it is marked as size over-limit. The allowable deviation is preset according to the dimensional tolerance of the machining drawing. When the dimension parameters required by the dimension-type parameter constraints are not generated in the current geometric feature, it is marked as missing dimensions; When there is a size linkage relationship and the associated size parameter is not updated synchronously with the size parameter being adjusted, it is marked as size linkage failure; Based on the marking results of out-of-limit dimensions, missing dimensions, or dimensional linkage failure, generate dimensional consistency detection results.

[0053] In a preferred embodiment of the present invention, generating a position consistency detection result includes: Extract the actual position parameters from the current detection data. The actual position parameters include coordinate values, center point position, distance from the reference plane, distance from the reference line, and feature direction. Extract the symmetry, alignment, spacing and reference positioning relationships from the positional parameter constraints to generate a positional comparison reference. When the positional parameter constraint relationship is symmetrical, the distance between the two geometric features and the symmetrical datum is extracted. When the difference between the two distances exceeds the preset positioning tolerance, it is marked as a symmetrical failure. When the positional parameter constraint relationship is an alignment relationship, extract the centerline, center point, or reference boundary of the corresponding geometric feature. When the centerline, center point, or reference boundary does not coincide, mark it as an alignment deviation. When the positional parameter constraint relationship is a spacing relationship, the actual distance between the two geometric features is extracted and compared with the target spacing; when the target spacing is 20mm, the allowable deviation is ±0.2mm, and the actual spacing is 20.4mm, it is marked as a spacing mismatch. The allowable deviation is preset according to the positioning dimension tolerance of the drawing. When the position parameter constraint relationship is a reference positioning relationship, extract the actual positioning value of the current geometric feature relative to the reference plane, reference line or reference point. When the actual positioning value is inconsistent with the target positioning value, mark it as a reference offset. Based on the marking results of symmetry failure, alignment deviation, spacing mismatch, or reference offset, position consistency detection results are generated.

[0054] In a preferred embodiment of the present invention, generating a topology connectivity consistency detection result includes: Extract geometric boundaries, adjacent faces, intersecting edges, connecting faces, and entity adjacency relationships from the current detection data to generate the current connection state; Extract the connection requirements from the parameter constraint relationship, including surface mating, edge connection, solid penetration, cut-through connection, or reinforcement of structural continuity; Match the current connection status with the connection requirements. If the two geometric features that are required to be connected do not have a common boundary or a common contact surface, mark it as a connection interruption. When the current geometric feature occupies the same solid space as an adjacent geometric feature and no cut, avoidance, or fusion relationship is configured, it is marked as an intersection interference; for example, when a stiffener overlaps with the avoidance area of ​​a mounting hole and no avoidance cut is configured, it is marked as an intersection interference. When there is an unclosed boundary, fracture surface, or unfilled region between the current geometric feature and its adjacent geometric features, it is marked as a topological gap; Based on the marking results of connection interruptions, cross-interferences, or topology gaps, generate topology connectivity consistency detection results.

[0055] In a preferred embodiment of the present invention, generating a feature dependency consistency detection result includes: Extract referenced objects, input parameter sources, output parameter destinations, and call records from the current detection data to generate current feature dependency data; Extract the geometric feature generation order and parameter constraint propagation path from the adaptive parameter modeling sequence to generate a dependent comparison benchmark; If the preceding geometric feature, datum plane, datum line, or parameter variable referenced by the current geometric feature does not exist, it is marked as an invalid reference; When the output parameters of the current geometric feature are used as the source of the preceding input of the current geometric feature, it is marked as a cyclic dependency; for example, when feature A depends on the localization surface of feature B, and feature B depends on the boundary contour of feature A, it is marked as a cyclic dependency. A geometric feature is marked as misaligned when the current geometric feature precedes its dependent objects in the geometric feature generation order. Generate feature dependency consistency detection results based on the marking results of invalid references, circular dependencies, or misordered sequences.

[0056] In a preferred embodiment of the present invention, constraint detection results are generated based on the size consistency detection results, position consistency detection results, topology connection consistency detection results, and feature dependency consistency detection results, including: Extract the size consistency detection results, position consistency detection results, topology connectivity consistency detection results, and feature dependency consistency detection results; Detection results without abnormal markers are written to the pass record, while detection results with out-of-limit dimensions, missing dimensions, dimension linkage failure, symmetry failure, alignment deviation, spacing mismatch, reference offset, connection interruption, cross interference, topological gap, failure reference, circular dependency, or sequence misalignment are written to the abnormal record. The abnormal records are bound to the corresponding current geometric features, parameter constraints, and parameter variables to generate abnormal association data; Constraint detection results are generated based on recorded and anomaly-related data.

[0057] In a preferred embodiment of the present invention, based on the constraint detection results, it is determined whether a constraint conflict exists; when a constraint conflict exists, the corresponding semantic constraints, conflict parameter variables, and conflict parameter constraint relationships are located according to the associated records in the parameter constraint model, and a conflict location result based on the conflict detection items is generated, including: Based on the constraint detection results, identify abnormal size consistency detection results, position consistency detection results, topology connection consistency detection results, and feature dependency consistency detection results, and generate conflict detection items; Based on the conflict detection items, determine the current geometric feature where the conflict has occurred and the target parameter constraint relationship associated with the current geometric feature, and generate the constraint conflict object; Extract the unique association identifier corresponding to the target parameter constraint relationship based on the constraint conflict object, and retrieve the associated record in the parameter constraint model based on the unique association identifier to locate the corresponding semantic constraint, parameter variable and geometric feature, and generate the first location result; Based on the parameter variables in the first positioning result, identify the conflicting parameter variables that are directly related to the target parameter constraint relationship, and identify the related parameter variables affected by the transmission along the parameter constraint propagation path to generate the second positioning result; Based on the semantic constraints in the first localization result, the original design requirements, modeling objects and constraint types that caused the conflict are extracted to generate semantic source results; Based on the first location result, the second location result, and the semantic source result, a conflict location result is generated.

[0058] In this embodiment of the invention, conflict detection items are generated by identifying abnormal size consistency detection results, position consistency detection results, topological connection consistency detection results, and feature dependency consistency detection results based on constraint detection results. Then, the current geometric features and target parameter constraint relationships where conflicts occur are determined based on the conflict detection items. The associated records are retrieved in the parameter constraint model using unique association identifiers to locate the corresponding semantic constraints, parameter variables, and geometric features. Subsequently, directly associated conflict parameter variables, associated parameter variables affected by propagation, and the original design requirements, modeling objects, and constraint types that caused the conflict are identified to form conflict location results. This enables the conflict handling process to trace from the abnormal phenomenon to the design source, affected parameters, and associated objects, solving the problem that constraint conflicts can only locate surface errors and it is difficult to determine the root cause and scope of influence, thus improving the accuracy and traceability of conflict analysis.

[0059] In a preferred embodiment of the present invention, generating conflict detection items includes: Extract pass records and anomalous records from the constraint detection results, and extract the anomalous type, current geometric features, parameter constraint relationships, and parameter variables corresponding to the anomalous records; When the anomaly type is size over-limit, size missing, or size linkage failure, the corresponding anomaly record will be marked as a size conflict detection item; When the anomaly type is symmetric failure, alignment deviation, spacing mismatch, or reference offset, the corresponding anomaly record will be marked as a positional conflict detection item. When the anomaly type is connection interruption, cross interference, or topology gap, the corresponding anomaly record will be marked as a topology conflict detection item; When the exception type is invalid reference, circular dependency, or misorder, the corresponding exception record will be marked as a dependency conflict detection item. The conflict detection items are generated by binding them with the corresponding anomaly type, current geometric features, and parameter variables.

[0060] In a preferred embodiment of the present invention, generating a constraint conflict object includes: Extract the current geometric feature identifier, anomaly type, and anomaly association data from the conflict detection item; Based on the current geometric feature identifier, determine the current geometric feature in the stage geometric model where the conflict occurs; Based on the abnormal correlation data, extract the parameter constraint relationship directly related to the current geometric feature, and determine the parameter constraint relationship as the target parameter constraint relationship; When a current geometric feature is associated with multiple parameter constraints, the parameter constraint that directly generates the anomaly in the anomaly record is determined as the target parameter constraint. For example, if a mounting hole is associated with both hole diameter and hole position constraints, and the anomaly type is spacing mismatch, then the hole position constraint is determined as the target parameter constraint. Combine the current geometric features and target parameter constraints to generate constraint conflict objects.

[0061] In a preferred embodiment of the present invention, generating a first positioning result includes: Extract the target parameter constraint relationships from the constraint conflict objects, and extract the unique association identifier carried by the target parameter constraint relationships; Using a unique association identifier as the retrieval field, the corresponding record is retrieved from the associated records in the parameter constraint model; When the unique association identifier is H001-P-D01, search for semantic constraint records, parameter variable records, and geometric feature records with the identifier field H001-P-D01 in the associated records; Extract the corresponding semantic constraints from the semantic constraint record, extract the corresponding parameter variables from the parameter variable record, and extract the corresponding geometric features from the geometric feature record; The corresponding semantic constraints, corresponding parameter variables, and corresponding geometric features are combined to generate the first localization result.

[0062] In a preferred embodiment of the present invention, generating a second positioning result includes: Extract the parameter variables and target parameter constraint relationships from the first positioning result, and determine the parameter variables directly limited by the target parameter constraint relationships; Parameter variables that directly cause abnormal records are identified as conflicting parameter variables; for example, when the hole center distance parameter variable is inconsistent with the target spacing, the hole center distance parameter variable is identified as a conflicting parameter variable. Extract the parameter constraint propagation path in the adaptive parameter modeling sequence and determine the subsequent propagation nodes of conflicting parameter variables; Extract the associated parameter variables that are passed from conflicting parameter variables to subsequent geometric features along the parameter constraint propagation path; for example, when the base plate length parameter variable affects the mounting hole spacing parameter variable and the stiffener length parameter variable, the mounting hole spacing parameter variable and the stiffener length parameter variable are identified as the associated parameter variables affected by the propagation. The conflicting parameter variables and the associated parameter variables affected by the transmission are combined to generate a second localization result.

[0063] In a preferred embodiment of the present invention, based on the semantic constraints in the first localization result, the original design requirements, modeling objects, and constraint types that caused the conflict are extracted to generate a semantic source result, including: Extract the semantic constraints from the first location result, and extract the source field, object field, and type field corresponding to the semantic constraints; Based on the source field, extract the original design requirement that generated the semantic constraint; for example, if the semantic constraint comes from "mounting hole distance from left edge 20mm", extract the text content as the original design requirement. Based on the object field, extract the modeling object to which the semantic constraint applies; for example, when the semantic constraint applies to a mounting hole, extract the mounting hole as the modeling object. Based on the type field, determine the constraint type of the semantic constraint. Constraint types include size semantic constraints, positional semantic constraints, and functional semantic constraints. The original design requirements, modeling objects, and constraint types are associated to generate semantic source results.

[0064] In a preferred embodiment of the present invention, a conflict location result is generated based on the first location result, the second location result, and the semantic source result, including: Extract the corresponding semantic constraints, corresponding parameter variables, and corresponding geometric features from the first localization result; Extract conflict parameter variables and related parameter variables affected by transmission from the second localization results; Extract the original design requirements, modeling objects, and constraint types from the semantic source results; The corresponding geometric features are identified as the objects of conflict, the conflict parameter variables are identified as the direct parameters of conflict, and the related parameter variables affected by the transmission are identified as the parameters of conflict impact. Write the corresponding semantic constraints, original design requirements, modeling objects, constraint types, conflict objects, direct conflict parameters, conflict impact parameters, and target parameter constraint relationships into the same location record; Conflict location results are generated based on the location records.

[0065] In a preferred embodiment of the present invention, the conflict type is determined based on the conflict localization result; when the conflict type is a size conflict, the value range of the conflict parameter variable is adjusted according to the corresponding semantic constraints; when the conflict type is a topological conflict, the conflict parameter constraint relationship is adjusted according to the corresponding semantic constraints; when the conflict type is a dependency conflict, the geometric feature generation order is adjusted according to the corresponding semantic constraints, and a conflict correction result is generated, including: Based on the conflict parameter variables, conflict parameter constraint relationships and corresponding semantic constraints in the conflict localization results, determine the conflict type corresponding to the constraint conflict and generate conflict classification results. The conflict types include size conflict, topology conflict and dependency conflict. Location consistency anomalies are classified as topology conflict or dependency conflict. When the conflict classification result is a size conflict, extract the size limitation requirements and functional limitation requirements from the corresponding semantic constraints, keep the parameter constraint relationship corresponding to the functional limitation requirements unchanged, redetermine the allowable value range of the conflict parameter variable, write the redetermined allowable value range into the corresponding parameter variable, and generate the size conflict correction result. When the conflict classification result is a topological conflict, the positional and connectivity constraints in the corresponding semantic constraints are extracted, the connectivity, positioning or constraint objects in the target parameter constraint relationship are adjusted, and the geometric features that have met the functional constraints are kept unchanged, and the topological conflict correction result is generated. When the conflict classification result is dependency conflict, extract the functional requirements and modeling object relationships from the corresponding semantic constraints, identify the current geometric features, previous geometric features and subsequent geometric features affected by the conflict, perform local order rearrangement of the geometric features affected by the conflict, and keep the order of the geometric features not affected by the conflict unchanged, and generate dependency conflict correction results. Generate conflict correction results based on size conflict correction results, topology conflict correction results, or dependency conflict correction results.

[0066] In this embodiment of the invention, the conflict type is determined based on the conflict parameter variables, conflict parameter constraints, and corresponding semantic constraints in the conflict location results, and a conflict classification result is generated. Then, for size conflicts, size limitation requirements and functional limitation requirements are extracted, and the allowable value range of the conflict parameter variables is adjusted while keeping the corresponding parameter constraint relationship of the functional limitation requirements unchanged. For topological conflicts, position limitation requirements and connection limitation requirements are extracted, and connection relationships, position relationships, or constraint objects are adjusted. For dependency conflicts, functional requirements and modeling object relationships are extracted, and the affected geometric features are locally rearranged. Finally, a conflict correction result is generated, so that conflicts from different sources adopt corresponding correction paths, solving the problem that a unified correction method leads to the destruction of functional requirements, instability of associated structures, or overall reconstruction, and improving the modeling effectiveness and result stability after conflict correction.

[0067] In a preferred embodiment of the present invention, based on the conflict parameter variables, conflict parameter constraint relationships, and corresponding semantic constraints in the conflict localization results, the conflict type corresponding to the constraint conflict is determined, and a conflict classification result is generated, including: Extract the conflict parameter variables, target parameter constraint relationships, corresponding semantic constraints, constraint types, and conflict occurrence objects from the conflict localization results; When the conflicting parameter variable is a parameter variable corresponding to length, width, height, thickness, hole diameter, hole depth, fillet radius or angle, and the target parameter constraint relationship is a dimension-type parameter constraint relationship, the constraint conflict is determined as a dimension conflict. When the constraint relationship of the target parameter involves the connection relationship, positioning relationship, symmetry relationship, alignment relationship, spacing relationship or reference positioning relationship between geometric features, the constraint conflict is identified as a topological conflict; When the preceding geometric features of the object referenced by the conflicting object are missing, there is a circular call relationship in the parameter constraint propagation path, or the object in conflict is located before its dependent object, the constraint conflict is identified as a dependency conflict. When a positional inconsistency is caused by a positioning relationship, symmetry relationship, alignment relationship, spacing relationship, or reference positioning relationship, the positional inconsistency is classified as a topological conflict. When a positional consistency anomaly is caused by missing preceding geometric features, failure to generate a baseline object, or misalignment of the modeling order, the positional consistency anomaly is classified as a dependency conflict. Based on the determination of size conflict, topology conflict, or dependency conflict, generate conflict classification results.

[0068] In a preferred embodiment of the present invention, generating the size conflict correction result includes: Extract the conflict classification results and conflict location results, confirm that the conflict type is size conflict, and extract the size limitation requirements, functional limitation requirements and conflict parameter variables from the corresponding semantic constraints; Set the parameter constraint relationship corresponding to the functional limitation requirement to a hold state so that the parameter constraint relationship does not change during the size correction process; Determine the target value or target value range of the conflict parameter variable based on the size limitation requirements, and determine the allowable value range in combination with the functional limitation requirements; When the size requirement is "mounting hole diameter 10mm" and the preset size allowable deviation is ±0.1mm, the allowable value range is 9.9mm to 10.1mm. The preset size allowable deviation is set according to the commonly used dimensional tolerances in machining drawings. When the functional requirement is "to satisfy bolt assembly", and the corresponding hole diameter range for bolt assembly is 9.8mm to 10.2mm, the intersection of the range corresponding to the size requirement and the range corresponding to the functional requirement is taken as the redefined allowable range, i.e., 9.9mm to 10.1mm. Write the newly determined allowed value range into the conflict parameter variable, and synchronously update the size class parameter constraint relationship directly associated with the conflict parameter variable; Based on the updated conflict parameter variables and dimensional parameter constraints, dimensional conflict correction results are generated.

[0069] In a preferred embodiment of the present invention, generating the topology conflict correction result includes: Extract the conflict classification results and conflict location results, confirm that the conflict type is a topological conflict, and extract the position constraints, connection constraints, target parameter constraints and conflict objects from the corresponding semantic constraints; Identify geometric features around the object where the conflict occurs that meet the functional constraints, and set the geometric features as retaining objects; When the positional constraints involve symmetry, alignment, spacing, or reference positioning, adjust the positioning relationship in the target parameter constraints; for example, when two mounting holes are symmetrical about the centerline but their actual positions do not satisfy the symmetry relationship, adjust the positioning distances from the two mounting holes to the centerline to the same value. When connection requirements involve surface mating, edge connection, solid penetration, cut-off connection, or reinforcement of structural continuity, adjust the connection relationships in the target parameter constraint relationships; for example, when the stiffener and the base plate are required to be surface mating and there is a connection interruption, constrain the bottom surface of the stiffener to the upper surface of the base plate. When the constraint object in the target parameter constraint relationship is inconsistent with the modeling object in the semantic constraint, the constraint object in the target parameter constraint relationship will be adjusted to the modeling object specified by the semantic constraint. When adjusting connection relationships, positioning relationships, or constraint objects, keep the parameter variables and parameter constraint relationships of geometric features that have met the functional requirements unchanged; Based on the adjusted target parameter constraints, the topology conflict correction results are generated.

[0070] In a preferred embodiment of the present invention, generating a dependency conflict correction result includes: Extract the conflict classification results and conflict location results, confirm that the conflict type is dependency conflict, and extract the functional requirements, modeling object relationships, conflicting objects and conflict impact parameters from the corresponding semantic constraints; Based on the relationships between the modeling objects, determine the preceding geometric features that the conflicting object depends on and the subsequent geometric features of the output results of the conflicting object. Based on the conflict impact parameters, determine the range of geometric features affected by the conflict, and write the conflict object, the preceding geometric features, and the following geometric features into the local rearrangement set; Preserve the order of geometric features not written into the local rearrangement set; In a local rearrangement set, the geometric features that provide reference planes, reference lines, positioning objects, or main structures are arranged before the geometric features that call those geometric features; When the mounting hole needs to use the upper surface of the base plate as a positioning surface, arrange the base plate stretching feature before the mounting hole; when the reinforcing rib needs to use the mounting hole clearance area, arrange the mounting hole before the reinforcing rib. Based on the adjustment results of the local rearrangement set, update the geometric feature generation order and parameter constraint propagation path in the adaptive parametric modeling sequence; Based on the updated geometric feature generation order and parameter constraint propagation path, dependency conflict correction results are generated.

[0071] In a preferred embodiment of the present invention, generating a conflict correction result based on a size conflict correction result, a topology conflict correction result, or a dependency conflict correction result includes: Extract the size conflict correction results, topology conflict correction results, or dependency conflict correction results; When the extraction result is a size conflict correction result, extract the updated parameter variables and size class parameter constraint relationships, and write them into the conflict correction record; When the extraction result is a topology conflict correction result, extract the adjusted connection relationship, positioning relationship, constraint object and target parameter constraint relationship, and write them into the conflict correction record; When the extraction result is a conflict correction result, extract the updated geometric feature generation order and parameter constraint propagation path, and write them into the conflict correction record. Bind the conflict correction record to the corresponding semantic constraints, the conflicting object, and the target parameter constraint relationship; Generate conflict correction results based on the bound conflict correction records.

[0072] Embodiments of the present invention also provide a large-model-driven intelligent generation parametric modeling system for CAD drawings, the system comprising: The requirements parsing module is used to obtain design requirements information, parse the natural language descriptions, multi-source engineering data and historical design samples in the design requirements information, identify the semantic entities representing the objects to be modeled and the semantic constraints representing the design requirements; generate a structured semantic feature set based on the semantic entities, semantic constraints and the relationship between them, and establish an initial mapping relationship between the structured semantic feature set and the geometric feature types. The parameter modeling module is used to convert semantic constraints into executable parameter constraint relationships based on a set of structured semantic features, and extract parameter variables used to control the size, position, or shape of geometric features; the parameter constraint relationships are associated with the corresponding parameter variables, geometric features, and semantic constraints to construct a parameter constraint model. The sequence generation module is used to parse the driving and dependency relationships between various parameter constraints based on the parameter constraint model, determine the execution priority according to the importance of semantic constraints, and generate an adaptive parameter modeling sequence containing the geometric feature generation order and parameter constraint propagation path. The drawing generation module is used to generate corresponding geometric features step by step in the CAD modeling environment by calling parameter variables according to the adaptive parameter modeling sequence, and to perform constraint consistency detection according to parameter constraint relationships. When constraint conflicts are detected, the value range of conflicting parameter variables, parameter constraint relationships or geometric feature generation order are adjusted according to the associated semantic constraints, and subsequent geometric feature generation is continued to obtain CAD drawings. The structural optimization module is used to verify the structural integrity and constraint satisfaction of CAD drawings, and update the structured semantic feature set and parameter constraint model based on the verification results, generating optimized parameter modeling results.

[0073] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0074] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0075] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0076] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent generation of parametric modeling parameters for CAD drawings driven by large models, characterized in that: The method includes: The system acquires design requirement information, parses the natural language descriptions, multi-source engineering data, and historical design samples in the design requirement information, identifies semantic entities representing the objects to be modeled and semantic constraints representing design requirements; generates a structured semantic feature set based on semantic entities, semantic constraints, and the relationship between them, and establishes an initial mapping relationship between the structured semantic feature set and geometric feature types. Based on the structured semantic feature set, semantic constraints are converted into executable parameter constraint relationships, and parameter variables used to control the size, position, or shape of geometric features are extracted; the parameter constraint relationships are associated with the corresponding parameter variables, geometric features, and semantic constraints respectively to construct a parameter constraint model; Based on the parameter constraint model, the driving and dependency relationships between each parameter constraint are analyzed, and the execution priority is determined by dividing the weights according to the semantic constraint type. Among them, the functional semantic constraint is the first priority, the size semantic constraint is the second priority, the position semantic constraint is the third priority, and the symmetry semantic constraint is the preceding sub-priority within the position semantic constraint. An adaptive parameter modeling sequence containing the geometric feature generation order and parameter constraint propagation path is generated. Based on the adaptive parametric modeling sequence, the corresponding geometric features are generated step by step in the CAD modeling environment by calling parameter variables, and constraint consistency is checked according to the parameter constraint relationship. When a constraint conflict is detected, the value range of the conflicting parameter variables, parameter constraint relationship or geometric feature generation order is adjusted according to the associated semantic constraints, and subsequent geometric feature generation continues to be executed to obtain the CAD drawing. The structural integrity and constraint satisfaction of CAD drawings are verified, and the structured semantic feature set and parameter constraint model are updated based on the verification results to generate optimized parameter modeling results. The process of generating adaptive parameter modeling sequences includes: Based on the parameter constraint model, the parameter constraint relationships are extracted, and the interaction relationships between each parameter constraint relationship are analyzed to generate driving parameter constraint relationships, driven parameter constraint relationships, and coupled parameter constraint relationships. Based on the driving parameter constraints, the baseline geometric feature is determined as the starting point for modeling; based on the driven parameter constraints, the subsequent geometric features generated based on the modeling results of the preceding geometric features are determined, where the baseline geometric feature is the first preceding geometric feature; based on the coupling parameter constraints, the associated geometric features that need to be updated synchronously are determined, and the geometric feature dependency results are generated. Based on the type of semantic constraints, the execution priority of each parameter constraint relationship is determined by the control order of functional semantic constraints, size semantic constraints, and positional semantic constraints, and the rule of symmetry semantic constraints as the pre-constraint of positional semantic constraints, and the parameter constraint priority result is generated. Based on the parameter constraint priority results, the preceding geometric features, subsequent geometric features, and associated geometric features in the geometric feature dependency results are sorted to generate the geometric feature generation order. Based on the generation order of geometric features, identify the parameter transfer relationship between the preceding and subsequent geometric features, and generate the parameter constraint propagation path based on the parameter transfer relationship; An adaptive parameter modeling sequence is generated based on the geometric feature generation order and parameter constraint propagation path.

2. The large-model-driven intelligent generation parametric modeling method for CAD drawings according to claim 1, characterized in that, Based on the structured semantic feature set, semantic constraints are converted into executable parametric constraint relationships, and parameter variables used to control the size, position, or shape of geometric features are extracted. The parametric constraint relationships are then associated with the corresponding parameter variables, geometric features, and semantic constraints to construct a parametric constraint model, including: The semantic constraints in the structured semantic feature set are parsed to generate size semantic constraints, positional semantic constraints and functional semantic constraints; Based on the semantic constraints of dimensions, the value restriction relationships of parameter variables are generated, and the parameter constraint relationships of dimension class are obtained; Based on positional semantic constraints, positional constraint relationships between geometric features are generated. These positional constraint relationships include symmetry relationships, alignment relationships, spacing relationships, or reference positioning relationships, resulting in positional parameter constraint relationships. Based on the functional semantic constraints, the functional constraint relationship of the geometric features is generated, and the functional class parameter constraint relationship is obtained. Based on the geometric feature type, extract the corresponding size parameter variables, position parameter variables, and shape parameter variables, establish the calling relationship between parameter variables and geometric features, and generate the corresponding results of parameter variables and geometric features; Based on the constraints of size parameters, position parameters, function parameters, and the correspondence between parameter variables and geometric features, a unique association identifier is generated for each parameter constraint relationship, and the unique association identifier is written into the association record of the corresponding semantic constraint, parameter variable, and geometric feature. Based on each associated record, semantic constraints, parameter constraint relationships, parameter variables, and geometric features are associated and organized to generate a parameter constraint model with bidirectional traceability paths.

3. The large-model-driven intelligent generation parametric modeling method for CAD drawings according to claim 1, characterized in that, Based on the adaptive parametric modeling sequence, corresponding geometric features are generated step-by-step by calling parameter variables in the CAD modeling environment, and constraint consistency checks are performed according to parameter constraint relationships. When a constraint conflict is detected, the value range of the conflicting parameter variables, parameter constraint relationships, or geometric feature generation order are adjusted according to the associated semantic constraints, and subsequent geometric feature generation continues to be executed to obtain CAD drawings, including: Based on the adaptive parameter modeling sequence, the geometric feature generation order and parameter constraint propagation path are extracted, and the corresponding parameter variables are called to perform geometric feature generation operations according to the geometric feature generation order to generate the stage geometric model; Based on the stage geometric model, extract the parameter constraint relationships associated with the current geometric features, and detect the consistency of size, position, topological connectivity, and feature dependency, generating constraint detection results; Based on the constraint detection results, determine whether there is a constraint conflict; when a constraint conflict exists, locate the corresponding semantic constraints, conflict parameter variables and conflict parameter constraint relationships based on the associated records in the parameter constraint model, and generate a conflict location result based on the conflict detection items. Based on the conflict location results, determine the conflict type; when the conflict type is size conflict, adjust the value range of the conflict parameter variables according to the corresponding semantic constraints; when the conflict type is topological conflict, adjust the conflict parameter constraint relationship according to the corresponding semantic constraints; when the conflict type is dependency conflict, adjust the geometric feature generation order according to the corresponding semantic constraints, and generate conflict correction results; among them, position consistency anomalies are classified into topological conflicts or dependency conflicts according to the cause of the anomaly. Based on the conflict correction results, continue with the remaining geometric feature generation operation to obtain the CAD drawing.

4. The large-model-driven intelligent generation parametric modeling method for CAD drawings according to claim 3, characterized in that, Based on the stage geometric model, parameter constraints associated with the current geometric features are extracted, and consistency in size, position, topological connectivity, and feature dependency is detected to generate constraint detection results, including: Based on the stage geometric model, the size parameters, position parameters, and shape parameters of the current geometric feature are identified, and the parameter constraint relationships associated with the current geometric feature are extracted to generate the current detection data; Based on the current detection data, the actual size status of the current geometric features is compared with the constraint relationship of the size parameters to identify whether there are size over-limits, size missing or size linkage failures, and generate size consistency detection results; Based on the current detection data, the actual position state of the current geometric features is compared with the position class parameter constraint relationship to identify whether there is symmetry failure, alignment deviation, spacing mismatch or reference offset, and generate position consistency detection results; Based on the current detection data, identify the connection status between the current geometric feature and its adjacent geometric features, match it with the connection requirements in the parameter constraint relationship, identify whether there are connection interruptions, cross interference or topological gaps, and generate topological connection consistency detection results; Based on the current detection data, identify the reference status of the current geometric feature to the preceding geometric feature and the call status of the subsequent geometric feature to the current geometric feature, and match them with the dependency relationships in the adaptive parameter modeling sequence to identify whether there are invalid references, circular dependencies or order misalignments, and generate feature dependency consistency detection results. Constraint detection results are generated based on the size consistency detection results, position consistency detection results, topology connection consistency detection results, and feature dependency consistency detection results.

5. The large-model-driven intelligent generation parametric modeling method for CAD drawings according to claim 3, characterized in that, Based on the constraint detection results, determine whether constraint conflicts exist. When a constraint conflict exists, locate the corresponding semantic constraints, conflicting parameter variables, and conflicting parameter constraint relationships based on the associated records in the parameter constraint model, and generate conflict localization results based on the conflict detection items, including: Based on the constraint detection results, identify abnormal size consistency detection results, position consistency detection results, topology connection consistency detection results, and feature dependency consistency detection results, and generate conflict detection items; Based on the conflict detection items, determine the current geometric feature where the conflict has occurred and the target parameter constraint relationship associated with the current geometric feature, and generate the constraint conflict object; Extract the unique association identifier corresponding to the target parameter constraint relationship based on the constraint conflict object, and retrieve the associated record in the parameter constraint model based on the unique association identifier to locate the corresponding semantic constraint, parameter variable and geometric feature, and generate the first location result; Based on the parameter variables in the first positioning result, identify the conflicting parameter variables that are directly related to the target parameter constraint relationship, and identify the related parameter variables affected by the transmission along the parameter constraint propagation path to generate the second positioning result; Based on the semantic constraints in the first localization result, the original design requirements, modeling objects and constraint types that caused the conflict are extracted to generate semantic source results; Based on the first location result, the second location result, and the semantic source result, a conflict location result is generated.

6. The large-model-driven intelligent generation parametric modeling method for CAD drawings according to claim 3, characterized in that, Based on the conflict location results, determine the conflict type; when the conflict type is a size conflict, adjust the value range of the conflict parameter variables according to the corresponding semantic constraints; when the conflict type is a topological conflict, adjust the conflict parameter constraint relationship according to the corresponding semantic constraints. When the conflict type is dependency conflict, the order of geometric feature generation is adjusted according to the corresponding semantic constraints to generate conflict correction results, including: Based on the conflict parameter variables, conflict parameter constraint relationships and corresponding semantic constraints in the conflict localization results, the conflict type corresponding to the constraint conflict is determined and the conflict classification result is generated. The conflict types include size conflict, topology conflict and dependency conflict. When the conflict classification result is a size conflict, extract the size limitation requirements and functional limitation requirements from the corresponding semantic constraints, keep the parameter constraint relationship corresponding to the functional limitation requirements unchanged, redetermine the allowable value range of the conflict parameter variable, write the redetermined allowable value range into the corresponding parameter variable, and generate the size conflict correction result. When the conflict classification result is a topological conflict, the positional and connectivity constraints in the corresponding semantic constraints are extracted, the connectivity, positioning or constraint objects in the target parameter constraint relationship are adjusted, and the geometric features that have met the functional constraints are kept unchanged, and the topological conflict correction result is generated. When the conflict classification result is dependency conflict, extract the functional requirements and modeling object relationships from the corresponding semantic constraints, identify the current geometric features, previous geometric features and subsequent geometric features affected by the conflict, perform local order rearrangement of the geometric features affected by the conflict, and keep the order of the geometric features not affected by the conflict unchanged, and generate dependency conflict correction results. Generate conflict correction results based on size conflict correction results, topology conflict correction results, or dependency conflict correction results.

7. A large-model-driven intelligent parametric modeling system for generating CAD drawings, characterized in that: The system, used in the method of any one of claims 1 to 6, comprises: The requirements parsing module is used to obtain design requirements information, parse the natural language descriptions, multi-source engineering data and historical design samples in the design requirements information, identify the semantic entities representing the objects to be modeled and the semantic constraints representing the design requirements; generate a structured semantic feature set based on the semantic entities, semantic constraints and the relationship between them, and establish an initial mapping relationship between the structured semantic feature set and the geometric feature types. The parameter modeling module is used to convert semantic constraints into executable parameter constraint relationships based on a set of structured semantic features, and extract parameter variables used to control the size, position, or shape of geometric features; the parameter constraint relationships are associated with the corresponding parameter variables, geometric features, and semantic constraints to construct a parameter constraint model. The sequence generation module is used to parse the driving and dependency relationships between various parameter constraints based on the parameter constraint model, and to determine the execution priority by dividing the weights according to the semantic constraint type. Among them, functional semantic constraints are the first priority, size semantic constraints are the second priority, position semantic constraints are the third priority, and symmetry semantic constraints are the preceding sub-priority within position semantic constraints. It generates an adaptive parameter modeling sequence that includes the geometric feature generation order and parameter constraint propagation path. The drawing generation module is used to generate corresponding geometric features step by step in the CAD modeling environment by calling parameter variables according to the adaptive parameter modeling sequence, and to perform constraint consistency detection according to parameter constraint relationships. When constraint conflicts are detected, the value range of conflicting parameter variables, parameter constraint relationships or geometric feature generation order are adjusted according to the associated semantic constraints, and subsequent geometric feature generation is continued to obtain CAD drawings. The structural optimization module is used to verify the structural integrity and constraint satisfaction of CAD drawings, and update the structured semantic feature set and parameter constraint model based on the verification results, generating optimized parameter modeling results.

8. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images