Intelligent identification method and system for boundary characteristics of closed cavity based on geometric topology information

Abstract

The invention discloses an intelligent identification method and system for boundary features of a closed cavity based on geometric topological information, and the method comprises the steps: constructing a full-process automatic method of boundary part identification, inner and outer surface identification, conduction surface identification and gap identification. Then, a B-Rep space scattering recognition algorithm and a graph theory search algorithm are adopted to complete inner surface recognition and outer surface recognition respectively; then accurate recognition of the conducting surface is realized based on contact area detection and curvature verification; and finally, gap identification is completed through surface pair screening, parallelism verification and non-contact area detection. The problems that traditional manual detection is low in efficiency and prone to missing detection are solved, automatic and high-precision recognition of geometric features is achieved through a standardized algorithm, the method is suitable for structural safety and sealing performance verification of closed cavities in the fields of aerospace and the like, and the detection efficiency and accuracy can be greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of computer-aided design and manufacturing technology, specifically involving intelligent recognition technology for the boundary features of sealed cavities based on geometric topological information. Background Technology

[0002] In the aerospace field, the sealing and structural integrity of sealed cavities such as aircraft fuel tanks are directly related to the operational safety of equipment. Abnormal contact at conductive surfaces can lead to media leakage, and gaps in boundary components can easily induce stress concentration and fatigue damage, thereby threatening the overall reliability of the aircraft. Such cavity assemblies are composed of multiple boundary components working together. Traditional inspection methods rely on engineers manually checking geometric features in CATIA (Computer-Aided Tri-Dimensional Interactive Application), which is not only inefficient (inspecting a single cavity can take 2-3 working days), but also prone to missed inspections and misjudgments due to subjective human judgment bias, making it difficult to meet the stringent requirements of modern aerospace manufacturing for inspection accuracy and efficiency.

[0003] The current technology faces three major bottlenecks: First, the native functions of CAD (Computer Aided Design) lack the ability to process geometric data of multi-boundary parts in batches, and the unification of the assembly coordinate system and the cavity modeling coordinate system is difficult, which directly affects the reliability of core judgment links such as distance detection. Second, the identification of internal and external surfaces relies on the subjective judgment of engineers based on experience, and lacks quantitative judgment standards, which can easily lead to identification confusion when faced with complex curved surface structures. Third, there is no dedicated automated tool to support the verification of curvature matching of conductive surfaces and the accurate identification of notch contours. In particular, the detection coverage of hidden notches inside the assembly is not comprehensive, and there is a significant risk of missed detection.

[0004] Therefore, there is an urgent need for a fully automated closed cavity structure inspection solution that can fundamentally solve the efficiency bottlenecks and accuracy defects of traditional manual operation and significantly improve the efficiency of closed cavity inspection. Summary of the Invention

[0005] Purpose of the Invention: This invention aims to provide an intelligent identification method and system for the boundary features of sealed cavities based on geometric topological information, addressing the industry pain points of low efficiency and difficulty in controlling errors under traditional manual inspection modes. By integrating interferometric detection technology, geometric topological feature rules, CAA secondary development tools, and rule-based feature matching algorithms, it achieves fully automated identification of boundary parts, boundary surfaces, and gaps in sealed cavities, significantly improving detection accuracy and execution efficiency. It can accurately adapt to the complex assembly scenarios in the aerospace manufacturing field that require few samples and high precision.

[0006] Technical solution: The intelligent recognition method for the boundary features of a closed cavity based on geometric topological information according to the present invention includes the following steps:

[0007] S1: Boundary Part Recognition: Obtain the assembly coordinate system of the sealed cavity assembly and the modeling coordinate system of the imaginary cavity, and transform the geometric data of all parts in the assembly from the assembly coordinate system to the modeling coordinate system; extract the geometric data of all parts in the assembly except the imaginary cavity, calculate the distance between the geometric data of the parts and the geometric data of the imaginary cavity, and filter the parts whose distance is within the preset threshold range and determine them as boundary parts.

[0008] S2: Perform internal and external surface identification on boundary parts: When identifying the internal surface, sample the ideal cavity surface row by row to obtain discrete sampling points, and then perform ray emission and contact detection operations; when identifying the external surface, construct a surface map model and recursively verify the constraint conditions of adjacent triangular mesh patches using a depth-first search method.

[0009] S3: Conductive Surface Recognition: Identify the local contact surface between the boundary part and the imaginary cavity, verify whether the local contact surface meets the constraints, and mark the local contact surface that meets the constraints as a conductive surface;

[0010] S4: Notch Identification: Filter two boundary parts that are in contact, detect whether there is a surface pair between the two boundary parts that are not in contact but are within the contact range, verify whether the surface pair meets the constraint of being aligned and parallel, and mark the surface pair that meets the constraint as a notch.

[0011] Furthermore, the distance between the part geometry data and the hypothetical cavity geometry data in step S1 includes point-to-point distance, point-to-line distance, point-to-plane distance, line-to-line distance, line-to-plane distance, and plane-to-plane distance; during the calculation process, a bounding box pre-detection method is used to exclude non-candidate geometric element pairs to improve the efficiency of distance calculation.

[0012] Furthermore, the inner surface identification process described in step S2 is as follows:

[0013] The CATBRepAccess interface is called to obtain the B-Rep model of the hypothetical cavity. The surface of the ideal cavity is sampled row by row along the grid array to generate a uniformly distributed set of discrete sampling points. Each sampling point contains three-dimensional coordinates and cavity surface normal vector information;

[0014] With each sampling point A fixed-length ray is emitted from the origin, along the opposite direction of the ideal cavity normal, that is, in the direction pointing into the assembly. The ray length is set to half the typical wall thickness of the cavity; ray emission is controlled via the CATRAY interface, and the CATIntersection interface is called to detect the contact status between the ray and the triangular mesh facets of the boundary parts in real time.

[0015] If the ray With a certain triangular mesh surface If contact occurs, and the angle between the ray at the contact point and the normal vector of the surface is less than a preset value, then the surface is marked. For the inner surface;

[0016] Traverse all sampling points and rays to form a set of inner surface patches. .

[0017] Furthermore, the outer surface identification process described in step S2 is as follows:

[0018] Treat all triangular mesh faces of the boundary part as undirected graphs. vertex If two faces are adjacent, then an edge is created between their corresponding vertices. This forms a patch adjacency graph; the CATTopologicalQuery interface is called to parse the topological relationships between patches, ensuring the accuracy of the adjacency graph.

[0019] Inner surface patch assembly Starting with a vertex in the given area, the Depth-First Search (DFS) algorithm is initiated to recursively search for adjacent vertices. For each candidate adjacent vertex... Verify the three constraints:

[0020] Patch spacing constraints: The minimum distance from the initial inner surface patch is less than or equal to 0.1 mm;

[0021] Normal parallel constraint: The angle between the normal vector of the surface and the normal vector of the initial inner surface patch is less than or equal to 5°.

[0022] Same component constraints: The ID of the part to which it belongs is consistent with the ID of the part to which the initial inner surface patch belongs;

[0023] Candidate facets that simultaneously satisfy all three constraints Marked as the outer surface, it eventually forms a set of outer surface patches. .

[0024] Furthermore, step S3 is implemented as follows:

[0025] Contact surface positioning: Based on the set of inner surfaces of boundary parts By comparing the surface of the hypothetical cavity with that of the B-Rep model, "local contact areas" are selected—that is, areas where the inner surface patch overlaps with the cavity surface, or where the distance between them is less than or equal to the contact threshold. The inner surface patch corresponding to this area is then used as the set of candidate conductive surface patches. ;

[0026] Curvature constraint verification: Calculate the candidate patch set by calling the CATCurvature interface. The curvature value of each facet is calculated and compared with a preset curvature range:

[0027] If the curvature value of a candidate surface is within a preset range, and the surface is indeed generated by the contact between the boundary component and the imaginary cavity, then the surface is marked as a conductive surface; ultimately, a set of conductive surfaces is formed. This provides a core testing object for verifying the sealing performance of the cavity.

[0028] Furthermore, step S4 is implemented as follows:

[0029] From the set of boundary parts, select part pairs that have a contact relationship, i.e., the surface distance between the two boundary parts is less than or equal to the contact determination threshold, or there is local surface overlap, and record them as candidate part pairs. ;

[0030] For each candidate part The CATEdge interface is called to extract the boundary curves of both surfaces, and the existence of surface pairs that are not in contact but are within the contact range is analyzed: from Extracting surfaces from inner / outer surfaces ,from Extracting facets from the inner or outer surface ,like and If the spacing is between 0.05-0.5mm (within the contact range but not actually in contact), then... As candidate surfaces for gaps;

[0031] For candidate surface pairs Verify two key constraints:

[0032] Position alignment constraint: geometric centroid and The line connecting the geometric centroids deviates from the preset reference direction by less than or equal to 3°;

[0033] Parallel constraints: normal vector and The angle between the normal vectors is less than or equal to 2°;

[0034] Candidate surface pairs that simultaneously satisfy both constraints Mark it as "gap" and record the location coordinates, size, and part pair information of the gap.

[0035] The present invention discloses an intelligent recognition system for the boundary features of a sealed cavity based on geometric topological information, comprising:

[0036] Data preprocessing module: used to obtain the assembly coordinate system of the sealed cavity assembly and the modeling coordinate system of the imaginary cavity, realize the unification of the assembly coordinate system and the modeling coordinate system, and at the same time complete the extraction and conversion of the geometric data of all parts in the assembly.

[0037] Boundary part recognition module: Calls the CATMathTransformation interface of CATIA CAA to realize coordinate transformation of the geometric data of the part, calls the CATGeoFactory interface to extract the point, line and surface geometric elements of the part, calculates the distance between the part and the imaginary cavity based on the Euclidean distance algorithm and projection method, and filters out boundary parts by combining bounding box pre-detection.

[0038] Inner and outer surface recognition module: The CATBRepAccess interface is called to obtain the ideal cavity BRep model, and the inner surface is recognized by the BRep spatial scattering algorithm; the CATTopologicalQuery interface is used to construct a patch adjacency graph, and the depth-first search algorithm is used to recognize the outer surface.

[0039] Conductive surface recognition module: Calls the CATCurvature interface to calculate the curvature of the contact surface, verifies whether the contact surface meets the constraint of "contact generation and curvature within a preset range", and marks the surface that meets the constraint as a conductive surface;

[0040] Notch Recognition Module: Calls the CATEdge interface to extract the boundary curve of the part, verifies the constraint of "two parts in contact, surfaces not in contact but parallel and aligned", and marks the surface pairs that meet the constraint as notches.

[0041] Furthermore, the system also includes a user interaction module integrated into the CATIA interface, including a dedicated workbench and command menu, supporting the setting of distance threshold and curvature range parameters, while highlighting the recognition results and exporting the data.

[0042] Furthermore, the exported data includes part ID, surface type, and identification status.

[0043] Furthermore, the exported data supports multiple formats, including Excel and TXT, which facilitates engineers in analyzing and archiving the recognition results.

[0044] Beneficial Effects: Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs a fully automated detection method that integrates "boundary part identification - inner and outer surface determination - conductive surface matching - notch feature extraction," which can significantly improve the detection efficiency of sealed cavities and fundamentally solve the efficiency bottleneck and accuracy defects of traditional manual operations. By solidifying the detection logic through standardized algorithms, the consistency and accuracy of the identification results are ensured, providing reliable technical support for the structural safety verification and sealing assurance of sealed cavities. At the same time, the solution design is highly compatible with subsequent technical modules such as the CAA (Component Application Architecture) development environment, boundary part distance detection algorithm, and inner and outer surface quantitative determination model, laying a solid foundation for rapid functional implementation. Attached Figure Description

[0045] Figure 1 This is a flowchart of intelligent recognition of the boundary features of a sealed cavity;

[0046] Figure 2 This is a flowchart for identifying internal and external surfaces;

[0047] Figure 3 This is a flowchart of the conductive surface identification process;

[0048] Figure 4 This is a flowchart of the gap identification process;

[0049] Figure 5 This is a diagram showing the results of boundary part identification;

[0050] Figure 6 These are images showing the results of the internal and external surface identification.

[0051] Figure 7 This is an overview of the results of the conduction surface identification;

[0052] Figure 8 This is a detailed image of the identification results for the conductive surface portion;

[0053] Figure 9 This is an overview of the gap identification results;

[0054] Figure 10 This is a detailed image showing the results of the gap identification. Detailed Implementation

[0055] The present invention will now be described in further detail with reference to the accompanying drawings.

[0056] like Figure 1 As shown, this invention proposes an intelligent recognition method for the boundary features of a closed cavity based on geometric topological information, specifically including the following steps:

[0057] Step 1: Boundary Part Identification: Based on the core logic of "coordinate unification - multi-dimensional distance calculation - threshold filtering", ensure that the parts that constitute the boundary of the sealed cavity are accurately separated from the assembly.

[0058] The assembly coordinate system (actual assembly position coordinates of parts) of the sealed cavity assembly and the modeling coordinate system (cavity design reference coordinates) of the hypothetical cavity are obtained. The CATMathTransformation interface is called to transform the geometric data (point, line, and surface elements) of all parts from the assembly coordinate system to the modeling coordinate system, eliminating the problem of incomparable geometric data caused by different coordinate systems. At the same time, the CATGeoFactory interface is used to extract the geometric elements of all parts in the assembly except the hypothetical cavity, filter out redundant data such as duplicate points and invalid triangular patches, and uniformly format the data into a "point coordinate set - line equation parameters - surface mesh patch" structure to provide standardized input for distance calculation.

[0059] For the preprocessed part geometry data and the hypothetical cavity geometry data, perform full-dimensional distance calculations, covering 6 types of geometric element combinations, and optimize calculation efficiency by combining bounding box pre-detection:

[0060] Point-to-point distance: The Euclidean distance algorithm is used to calculate the straight-line distance between discrete points of the part and discrete points of the imaginary cavity. The formula is as follows:

[0061]

[0062] in, Here are the coordinates of the part points. These are the coordinates of a point in the cavity.

[0063] Point-to-line / point-to-plane distance: Using the projection method, the point is projected onto the perpendicular plane of the line / plane, and the perpendicular distance between the projected point and the origin is calculated.

[0064] Line-to-line / line-to-plane / plane-to-plane distance: First, construct the minimum bounding box (axis-aligned bounding box AABB) of the geometric elements. By comparing the positional relationship of the bounding boxes, exclude non-candidate element pairs whose distances are far beyond the preset threshold. Only calculate the precise distance for candidate pairs whose bounding boxes overlap or are adjacent.

[0065] Based on cavity specifications and engineering experience, a preset distance threshold is established (which can be flexibly adjusted via the user interaction module, such as setting the threshold to 0.1-0.3mm for an aircraft fuel tank scenario). Parts whose distance from the hypothetical cavity falls within this threshold range are selected and marked as "boundary parts," forming a set of boundary parts. This set defines the analysis range for subsequent internal and external surface identification. The identification results of boundary parts will be highlighted in the model diagram, such as... Figure 5 As shown.

[0066] Step 2: Identify the inner and outer surfaces, such as Figure 2As shown, the two-step method of "inner surface positioning (spatial scattering) - outer surface derivation (graph theory search)" is used to achieve accurate differentiation of the surface properties of boundary parts.

[0067] (2.1) Inner surface identification: Based on "ideal cavity sampling-ray detection", locate the surface of the boundary part facing the inner side of the cavity.

[0068] Ideal cavity surface sampling: The CATBRepAccess interface is used to obtain the B-Rep (boundary representation) model of the hypothetical cavity. The surface of the ideal cavity is sampled row by row along the grid array (the grid density is configured according to the accuracy requirements of the cavity, such as sampling one point per 10mm×10mm area), generating a uniformly distributed set of discrete sampling points. Each sampling point contains three-dimensional coordinates and cavity surface normal vector information.

[0069] X-ray emission and contact detection: at each sampling point A fixed-length ray is emitted from the origin along the opposite direction of the ideal cavity normal (i.e., pointing towards the interior of the assembly). The ray length is set to "1 / 2 of the typical cavity wall thickness" (e.g., when the wall thickness of an aircraft fuel tank is 5mm, the ray length is set to 2.5mm) to avoid ray penetration of boundary parts and subsequent misjudgment. Ray emission is controlled via the CATRAY interface, and the CATIntersection interface is called to detect the contact status between the ray and the triangular mesh facets of the boundary parts in real time.

[0070] If the ray With a certain triangular mesh surface If contact occurs, and the angle between the ray at the contact point and the normal vector of the surface is less than 15° (to avoid misjudging oblique contact), then mark the surface. For the "inner surface"; traverse all sampling points and rays to form a set of inner surface patches. .

[0071] (2.2) External surface identification: Based on the marked internal surface, the external surface of the boundary part is derived by graph theory search.

[0072] Surface graph model construction: Treat all triangular mesh faces of the boundary part as undirected graphs. vertex (Each vertex corresponds to a face). If two faces are adjacent (sharing an edge or two or more vertices), then an edge is created between the corresponding vertices. This forms a "patch adjacency graph"; the CATTopologicalQuery interface is called to parse the topological relationships between patches, ensuring the accuracy of the adjacency graph.

[0073] Depth-First Search (DFS) Verification: Using the set of inner surface patches Starting with a vertex in the given area, the Depth-First Search (DFS) algorithm is initiated to recursively search for adjacent vertices. For each candidate adjacent vertex... Verify the three constraints:

[0074] Patch spacing constraints: Minimum distance from the initial inner surface patch ≤ 0.1mm (to avoid misjudgment across parts or non-associated patches); Normal parallel constraint: The angle between the normal vector and the normal vector of the initial inner surface patch is ≤5° (to ensure consistent surface orientation); same as part constraints: The ID of the part to which the candidate surface belongs is consistent with the ID of the part to which the initial inner surface patch belongs (excluding patch interference from other boundary parts); candidate surfaces that simultaneously satisfy all three constraints Labeled as "outer surface", it eventually forms a set of outer surface patches. The identification results of the inner and outer surfaces are as follows: Figure 6 As shown, all of them are facets from the boundary parts combined into a whole.

[0075] Step 3: Conductor surface identification: such as Figure 3 As shown, the contact surfaces between the focus boundary parts and the imaginary cavity are selected through the dual constraints of "contact positioning-curvature verification" to identify key surfaces with sealing and conduction functions.

[0076] Contact surface positioning: Based on the set of inner surfaces of boundary parts By comparing the surface of the hypothetical cavity with the B-Rep model, "local contact areas" are selected—that is, areas where the inner surface patch overlaps with the cavity surface, or where the distance between them is ≤0.05mm (contact threshold). The inner surface patch corresponding to this area is used as the set of candidate conductive surface patches. .

[0077] Curvature constraint verification: Calculate the candidate patch set by calling the CATCurvature interface. The curvature values ​​(including principal curvature and Gaussian curvature) of each surface are calculated and compared with a preset curvature range (configured according to the cavity material and sealing requirements, such as 0.01-10mm⁻¹ for aircraft fuel tank scenarios). If the curvature value of a candidate surface is within the preset range, and the surface is indeed generated by the contact between the boundary part and the hypothetical cavity (not an isolated surface), then the surface is marked as a conductive surface; finally, a set of conductive surfaces is formed. This provides the core testing object for cavity sealing verification. All results of conductive surface identification are displayed in the same assembly model, such as... Figure 7 As shown, details of some conductive surfaces are as follows: Figure 8 As shown, all facets are contact surfaces that meet the requirements.

[0078] Step 4: Gap Identification: For structural defects of "incomplete contact" between boundary parts, identification is performed through methods such as... Figure 4 The three-step method shown, namely "candidate part screening - surface pairing inspection - constraint verification", accurately locates the gap area that is prone to stress concentration.

[0079] Candidate part selection: From the set of boundary parts, select "part pairs with contact relationship"—that is, the surface distance between two boundary parts is ≤0.5mm (contact determination threshold), or there is local surface overlap, and these are recorded as candidate part pairs. .

[0080] Surface pair detection: for each candidate part The CATEdge interface is called to extract the boundary curves of both surfaces and analyze whether there are "surface pairs that are not in contact but are within the contact range":

[0081] from Extracting surfaces from inner / outer surfaces ,from Extracting patches from the inner / outer surface ,like and If the spacing is between 0.05-0.5mm (within the contact range but not actually in contact), then... As a candidate surface pair for the gap.

[0082] Notch constraint verification: for candidate surface pairs Verify two key constraints: Position alignment constraint: geometric centroid and The deviation of the line connecting the geometric centroids from the preset reference direction (such as the cavity axis direction) is ≤3°; parallel constraint: normal vector and The angle between the normal vectors is ≤2°; candidate surface pairs that simultaneously satisfy both constraints The notch is labeled as a "gap" and its location coordinates, dimensions, and associated part pair information are recorded to provide a basis for subsequent structural optimization. All the final notch identification results are displayed in the same assembly model, such as... Figure 9 As shown, details of some of the gaps are as follows: Figure 10 As shown, all the gaps are geometric gaps formed between multiple parts.

[0083] This invention also proposes an intelligent recognition system for the boundary features of a sealed cavity based on geometric topological information, comprising:

[0084] The data preprocessing module is used to obtain the assembly coordinate system of the sealed cavity assembly and the modeling coordinate system of the imaginary cavity, achieving unification between the assembly and modeling coordinate systems. It also extracts and transforms the geometric data of all parts in the assembly. After extracting the part geometric data, the data is cleaned and formatted to remove redundant data, ensuring the accuracy of subsequent distance calculations and feature recognition.

[0085] Boundary part recognition module: It calls the CATMathTransformation interface of CATIA CAA to realize the coordinate transformation of the geometric data of the part, calls the CATGeoFactory interface to extract the point, line and surface geometric elements of the part, calculates the distance between the part and the imaginary cavity based on the Euclidean distance algorithm and projection method, and filters out boundary parts by combining bounding box pre-detection.

[0086] Inner and outer surface recognition module: The CATBRepAccess interface is called to obtain the ideal cavity BRep model, and the inner surface is recognized by the BRep spatial scattering algorithm; the CATTopologicalQuery interface is used to construct a patch adjacency graph, and the depth-first search algorithm is used to recognize the outer surface.

[0087] Conductive surface identification module: Calls the CATCurvature interface to calculate the curvature of the contact surface, verifies whether the contact surface meets the constraint of "contact generation and curvature within a preset range", and marks the surface that meets the constraint as a conductive surface.

[0088] Notch Recognition Module: Calls the CATEdge interface to extract the boundary curve of the part, verifies the constraint of "two parts in contact, surfaces not in contact but parallel and aligned", and marks the surface pairs that meet the constraint as notches.

[0089] User interaction module: Integrated into the CATIA interface, including a dedicated workbench and command menu, supporting the setting of parameters such as distance threshold and curvature range. It also supports highlighting of recognition results (red for inner surfaces, blue for outer surfaces, green for conductive surfaces, and yellow for notches) and data export (including part ID, surface type, and recognition status). Data supports multiple formats, including Excel and TXT, facilitating subsequent analysis and archiving of recognition results by engineers.

[0090] This invention is based on the CATIACAA (Component Application Architecture) secondary development architecture. It constructs a fully automated implementation system for the identification of boundary parts, inner and outer surfaces, conductive surfaces and gaps in sealed cavities (such as aircraft fuel tanks and engine hydraulic cavities). This system covers the entire process of “development environment setup - boundary part identification - inner and outer surface identification - conductive surface identification - gap identification”, which solves the problems of low efficiency and poor accuracy of traditional manual identification.

[0091] This specification has described the technical solution and core features of the present invention through specific embodiments, with the aim of enabling those skilled in the art to understand and implement the present invention. It should be noted that the above descriptions are merely illustrative of the present invention and are not intended to limit the scope of protection of the present invention. All equivalent changes or modifications made based on the spirit and essence of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent recognition of boundary features of a sealed cavity based on geometric topological information, characterized in that, Includes the following steps: S1: Boundary Part Recognition: Obtain the assembly coordinate system of the sealed cavity assembly and the modeling coordinate system of the imaginary cavity, and transform the geometric data of all parts in the assembly from the assembly coordinate system to the modeling coordinate system; extract the geometric data of all parts in the assembly except the imaginary cavity, calculate the distance between the geometric data of the parts and the geometric data of the imaginary cavity, and filter the parts whose distance is within the preset threshold range and determine them as boundary parts. S2: Perform internal and external surface identification on boundary parts: When identifying the internal surface, sample the ideal cavity surface row by row to obtain discrete sampling points, and then perform ray emission and contact detection operations; when identifying the external surface, construct a surface map model and recursively verify the constraint conditions of adjacent triangular mesh patches using a depth-first search method. S3: Conductive Surface Recognition: Identify the local contact surface between the boundary part and the imaginary cavity, verify whether the local contact surface meets the constraints, and mark the local contact surface that meets the constraints as a conductive surface; S4: Notch Identification: Filter two boundary parts that are in contact, detect whether there is a surface pair between the two boundary parts that are not in contact but are within the contact range, verify whether the surface pair meets the constraint of being aligned and parallel, and mark the surface pair that meets the constraint as a notch.

2. The intelligent recognition method for the boundary features of a sealed cavity based on geometric topological information according to claim 1, characterized in that, The distance between the part geometry data and the hypothetical cavity geometry data mentioned in step S1 includes point-to-point distance, point-to-line distance, point-to-surface distance, line-to-line distance, line-to-surface distance, and surface-to-surface distance. During the calculation process, a bounding box pre-detection method is used to exclude non-candidate geometric element pairs to improve the efficiency of distance calculation.

3. The intelligent recognition method for the boundary features of a sealed cavity based on geometric topological information according to claim 1, characterized in that, The process of inner surface identification in step S2 is as follows: The CATBRepAccess interface is called to obtain the B-Rep model of the hypothetical cavity. The surface of the ideal cavity is sampled row by row along the grid array to generate a uniformly distributed set of discrete sampling points. Each sampling point contains three-dimensional coordinates and cavity surface normal vector information; With each sampling point A fixed-length ray is emitted from the origin, along the opposite direction of the ideal cavity normal, that is, in the direction pointing into the assembly. The ray length is set to half the typical wall thickness of the cavity; ray emission is controlled via the CATRAY interface, and the CATIntersection interface is called to detect the contact status between the ray and the triangular mesh facets of the boundary parts in real time. If the ray With a certain triangular mesh surface If contact occurs, and the angle between the ray at the contact point and the normal vector of the surface is less than a preset value, then the surface is marked. For the inner surface; Traverse all sampling points and rays to form a set of inner surface patches. .

4. The intelligent recognition method for the boundary features of a sealed cavity based on geometric topological information according to claim 1, characterized in that, The outer surface identification process described in step S2 is as follows: Treat all triangular mesh faces of the boundary part as undirected graphs. vertex If two faces are adjacent, then an edge is created between their corresponding vertices. This forms a patch adjacency graph; the CATTopologicalQuery interface is called to parse the topological relationships between patches, ensuring the accuracy of the adjacency graph. Inner surface patch assembly Starting with a vertex in the given area, the Depth-First Search (DFS) algorithm is initiated to recursively search for adjacent vertices. For each candidate adjacent vertex... Verify the three constraints: Patch spacing constraints: The minimum distance from the initial inner surface patch is less than or equal to 0.1 mm; Normal parallel constraint: The angle between the normal vector of the surface and the normal vector of the initial inner surface patch is less than or equal to 5°. Same component constraints: The ID of the part to which it belongs is consistent with the ID of the part to which the initial inner surface patch belongs; Candidate facets that simultaneously satisfy all three constraints Marked as the outer surface, it eventually forms a set of outer surface patches. .

5. The intelligent recognition method for the boundary features of a sealed cavity based on geometric topological information according to claim 1, characterized in that, The implementation process of step S3 is as follows: Contact surface positioning: Based on the set of inner surfaces of boundary parts By comparing the surface of the hypothetical cavity with that of the B-Rep model, "local contact areas" are selected—that is, areas where the inner surface patch overlaps with the cavity surface, or where the distance between them is less than or equal to the contact threshold. The inner surface patch corresponding to this area is then used as the set of candidate conductive surface patches. ; Curvature constraint verification: Calculate the candidate patch set by calling the CATCurvature interface. The curvature value of each facet is calculated and compared with a preset curvature range: If the curvature value of a candidate surface is within a preset range, and the surface is indeed generated by the contact between the boundary component and the imaginary cavity, then the surface is marked as a conductive surface; ultimately, a set of conductive surfaces is formed. This provides a core testing object for verifying the sealing performance of the cavity.

6. The intelligent recognition method for the boundary features of a sealed cavity based on geometric topological information according to claim 1, characterized in that, The implementation process of step S4 is as follows: From the set of boundary parts, select part pairs that have a contact relationship, i.e., the surface distance between the two boundary parts is less than or equal to the contact determination threshold, or there is local surface overlap, and record them as candidate part pairs. ; For each candidate part The CATEdge interface is called to extract the boundary curves of both surfaces, and the existence of surface pairs that are not in contact but are within the contact range is analyzed: from Extracting surfaces from inner / outer surfaces ,from Extracting facets from the inner or outer surface ,like and If the spacing is between 0.05-0.5mm (within the contact range but not actually in contact), then... As candidate surface pairs for gaps; For candidate surface pairs Verify two key constraints: Position alignment constraint: geometric centroid and The line connecting the geometric centroids deviates from the preset reference direction by less than or equal to 3°; Parallel constraints: normal vector and The angle between the normal vectors is less than or equal to 2°; Candidate surface pairs that simultaneously satisfy both constraints Mark it as "gap" and record the location coordinates, size, and part pair information of the gap.

7. A smart recognition system for the boundary features of a sealed cavity based on geometric topological information, employing the method described in any one of claims 1 to 6, characterized in that, include: Data preprocessing module: used to obtain the assembly coordinate system of the sealed cavity assembly and the modeling coordinate system of the imaginary cavity, realize the unification of the assembly coordinate system and the modeling coordinate system, and at the same time complete the extraction and conversion of the geometric data of all parts in the assembly. Boundary part recognition module: Calls the CATMathTransformation interface of CATIA CAA to realize coordinate transformation of the geometric data of the part, calls the CATGeoFactory interface to extract the point, line and surface geometric elements of the part, calculates the distance between the part and the imaginary cavity based on the Euclidean distance algorithm and projection method, and filters out boundary parts by combining bounding box pre-detection. Inner and outer surface recognition module: The CATBRepAccess interface is called to obtain the ideal cavity BRep model, and the inner surface is recognized by the BRep spatial scattering algorithm; the CATTopologicalQuery interface is used to construct a patch adjacency graph, and the depth-first search algorithm is used to recognize the outer surface. Conductive surface recognition module: Calls the CATCurvature interface to calculate the curvature of the contact surface, verifies whether the contact surface meets the constraint of "contact generation and curvature within a preset range", and marks the surface that meets the constraint as a conductive surface; Notch Recognition Module: Calls the CATEdge interface to extract the boundary curve of the part, verifies the constraint of "two parts in contact, surfaces not in contact but parallel and aligned", and marks the surface pairs that meet the constraint as notches.

8. The intelligent recognition system for the boundary features of a sealed cavity based on geometric topological information according to claim 1, characterized in that, The system also includes a user interaction module integrated into the CATIA interface, which includes a dedicated workbench and command menu. It supports setting distance threshold and curvature range parameters, while highlighting the recognition results and exporting the data.

9. The intelligent recognition system for the boundary features of a sealed cavity based on geometric topological information according to claim 8, characterized in that, The exported data includes part ID, surface type, and identification status.

10. The intelligent recognition system for the boundary features of a sealed cavity based on geometric topological information according to claim 8, characterized in that, The exported data supports multiple formats, including Excel and TXT, which facilitates engineers in analyzing and archiving the recognition results.

Images