Self-adaptive generation method for hole group detection data of large-scale structural member of airplane

Abstract

The invention relates to the technical field of data detection, in particular to a self-adaptive generation method for hole group detection data of a large-scale aircraft structural member, which comprises the following steps of: extracting the geometry of all cylindrical surfaces on the contour surface of a three-dimensional geometric model of the aircraft structural member, and grouping the geometry of the cylindrical surfaces based on a'surface-surface 'topological adjacency relation and radius parameters of the cylindrical surfaces; the axis direction consistency, the curved surface concavity and convexity and the radial contour closure of each group of cylindrical surfaces are judged respectively, cylindrical surface groups forming hole characteristics are identified, the holes are grouped and sorted based on hole characteristic parameters, the feeding direction of the hole characteristics along the hole axis measurement movement is adaptively planned through the pose of a three-dimensional geometric model in an observation view angle, and the measurement precision of the hole characteristics is improved. And hole feature measurement data of different hole diameters, hole depths and hole positions are created in batches and in groups. According to the method, the algorithm complexity and the time complexity are greatly reduced, and the characteristics of the airplane structural part holes are efficiently and accurately identified.

Description

An adaptive generation method for hole group detection data of large aircraft structural components Technical Field

[0001] This invention relates to the field of digital inspection technology, and in particular to an adaptive generation method for inspection data of hole groups in large aircraft structural components. Background Technology

[0002] The design of inspection data for hole groups in aircraft structural components requires the accurate identification of all hole features on the parts. Currently, in actual manufacturing processes, manual and semi-automatic design methods are used, which suffer from problems such as susceptibility to errors and omissions, low efficiency, and high labor intensity.

[0003] In the prior art, a Chinese invention patent document with publication number CN117113455A and publication date November 24, 2023, is disclosed. The technical solution disclosed in this patent document is as follows: A method for automatic recognition of three-dimensional geometric features based on topological rules, which includes the steps of: based on B The Rep model is used to obtain a list of faces and edges of the part; based on the face and edge lists and the model topology information, attribute description data of the part is constructed; the face list is iterated through; based on the attribute description data, the traversal results of the face list are matched with the feature rule base to identify the three-dimensional geometric features; and the basic dimensions of the three-dimensional geometric features are calculated.

[0004] In existing technologies, Shu Min and Yang Tao proposed a part model feature recognition method combining attribute adjacency graphs and point clouds in 2024. Wang Min and Yang Tao proposed an injection molded part through-hole feature recognition algorithm integrating OpenCASCADE and OpenCV in 2024. Zhang Hang, Zhang Shusheng, and Yang Lei proposed a hole feature manufacturability analysis method based on deep learning in 2021. Shen Dawei, Xiang Hua, Zhuang Xincun, and Zhao Zhen proposed a three-dimensional countersunk hole feature recognition and key parameter extraction based on convolutional neural networks in 2021.

[0005] In the aforementioned existing technologies, feature recognition methods based on attribute adjacency graphs and traces cannot meet the requirements for accurate feature recognition of complex and non-fixed topology aircraft structural components, and cannot automatically plan the measurement direction of complex aircraft structural components based on limited geometric information; feature recognition methods based on artificial intelligence algorithms such as convolutional neural networks and deep learning cannot meet the requirement of 100% accuracy of aerospace structural component inspection data. Summary of the Invention

[0006] To address the aforementioned technical issues, this invention proposes an adaptive generation method for detecting hole groups in large aircraft structural components. This method avoids the construction of an overall adjacency attribute graph of the structural component and the identification of multiple types of non-fixed topological hole feature subgraphs, significantly reducing algorithm complexity and time complexity. It directly extracts key geometric information and, based on geometric attribute information constraints, employs an elimination method to efficiently and accurately identify hole features in aircraft structural components. Addressing the adaptive planning requirements for the measurement direction of complex aircraft structural component hole groups, this invention proposes an approach that adaptively plans the measurement direction from the observer's perspective.

[0007] This invention is achieved by adopting the following technical solution: an adaptive generation method for hole group detection data of large aircraft structural components, comprising the following steps: Step S1. Extracting the geometry of all cylindrical surfaces on the contour surface of the three-dimensional geometric model of the structural component, forming a set of cylindrical surfaces S. c Step S2. Based on the "face-to-face" topological adjacency relationship and the cylindrical surface radius parameter, set the cylindrical surfaces S... c Divided into multiple subsets S c (i) (i=1,2,…); Step S3. Based on the consistency of the cylindrical surface axial direction, the concavity and convexity of the cylindrical surface, and the radial closure of the cylindrical surface in each subset, from set S c Delete subsets that do not meet the requirements; set S c The cylindrical surfaces within the remaining subset form hole features, completing the identification of hole feature geometry; Step S4. Identify hole feature parameters and group and sort the hole features based on the hole feature parameters; Step S5. Adaptively plan the feed direction of the hole feature measurement motion along the hole axis using the pose of the 3D geometric model in the observation view; Step S6. Based on the hole feature measurement direction, parameters, and measurement coordinate system, batch group and create path-optimized measurement data.

[0008] The division in step S2 is based on the following: all cylindrical surfaces that are directly adjacent in topology and have the same radius parameter are divided into the same subset.

[0009] Step S2 specifically includes the following steps: Step S 21 Obtain the unique identifiers for each cylindrical surface on the 3D geometric model, forming a set S of cylindrical surface identifiers. t Step S 22 Obtain the radius parameter values ​​for each cylindrical surface; Step S 23 Establish an extensible identifier matrix T and a set S. t The first element is used as the first element T of the identifier matrix. （1，1） Empty elements in matrix T are filled with 0, and each row of matrix T is an identifier for a cylinder with the same radius; Step S 24 Traverse the set S of cylindrical surface identifiers. tFor all elements, in order, perform step S on the cylindrical surface f1 corresponding to each identifier. 25 Step S 25 Obtain all topologically adjacent surfaces of cylindrical surface f1, and determine whether each topologically adjacent surface is a cylindrical surface; if any topologically adjacent surface f2 is a cylindrical surface, then proceed to step S. 26 After traversal ends, jump to step S. 24 Step S 26 Determine if the radius parameters of the adjacent surface f2 and the cylindrical surface f1 are the same. If the radii are the same, then group the adjacent cylindrical surface f2 into the same group as the cylindrical surface f1, add the identifier corresponding to the cylindrical surface f2 to the row containing the identifier corresponding to the cylindrical surface f1 in matrix T, and then add the identifier from the identifier set S. t Delete the identifier corresponding to the adjacent surface f2; if the radii are not the same, the subsequent steps create a new group for the adjacent cylindrical surface f2 and add the identifier corresponding to the cylindrical surface f2 to the new row of matrix T; step S 27 Treat the adjacent cylindrical surface f2 as the cylindrical surface f1, and iteratively execute step S. 25 Step S 28 Complete the set of cylindrical surface identifiers S. t By traversing and judging all elements, we obtain the identifier matrix T(i,j), where each identifier corresponds to a cylindrical surface, i is the number of cylindrical surface groups, j is the number of cylindrical surfaces in each group, and the radius parameter of each group of cylindrical surfaces is the same.

[0010] Based on the consistency of the cylindrical axis orientation in each subset, from set S c Deleting subsets that do not meet the requirements specifically means: traversing all subsets, if any two randomly selected cylinders in any subset have axes with different directions, then removing them from set S. c Delete that subset.

[0011] Based on the concavity and convexity of the cylindrical surfaces of each subset, from set S c Deleting subsets that do not meet the requirements specifically includes the following steps: Step S 31 Obtain the normal vector V1 at any point P on the cylindrical surface; step S 32 Obtain the material side direction at point P. If the normal vector V1 is opposite to the material side direction, reverse the normal vector V1. The material side direction is the direction from any point on the surface of the part geometry to the geometric entity. Step S 33 Construct a vector V2 pointing from point P to the axis of the cylindrical surface; step S 34 Determine the angle α between normal vectors V1 and V2. If the angle α = 0°, the cylindrical surface is a convex cylindrical surface; otherwise, it is a concave cylindrical surface. (Step S) 35 From set S c Remove the subset containing convex cylindrical surfaces.

[0012] Based on the radial closure of the cylindrical surfaces of each subset, from set S c Deleting subsets that do not meet the requirements specifically refers to deleting subsets that have radially unclosed cylindrical surfaces.

[0013] The method for determining whether a cylindrical surface is radially closed includes the following steps: Step 1. Extract the edges of each cylindrical surface within the subset; Step 2. Using the unique identifier of each edge, obtain all common edges between any two cylindrical surfaces, forming a set S. e Step 3. If two cylindrical surfaces share multiple edges, and among these shared edges there exists a pair of consecutive first-order shared edges, retain the shared edge with the longest projected length on the axis of the cylindrical surface. If the projected lengths of all shared edges on the axis of the cylindrical surface are zero, then retain the shared edge with the longest length. From set S... e By deleting all shared edges except the original shared edges from each pair of consecutive first-order shared edges, we obtain the set S of shared edges. e Step 4. Set S using cylindrical surfaces and shared edges. e 'Construct a face-edge graph G of the cylindrical surfaces and their shared edges within the subset, where faces are graph nodes and shared edges of the cylindrical surfaces are edges in the graph; Step 5. Extract the closed loop L of the face nodes in the face-edge graph G.' p Step 6. If a closed loop exists, starting from any node of the closed loop, traverse each face node in the loop sequentially along the closed loop path. If the sum of the angles between any two center normal vectors of each face node is 360°, then the cylindrical surface within the subset is radially closed.

[0014] Step S5 adaptively plans the feed direction of the hole feature along the hole axis by using the pose of the three-dimensional geometric model in the observation view. Specifically, it includes the following steps: Step S 51 Obtain the central axis of the hole by using the axes of the cylindrical surfaces that make up the hole; Step S 52 Adjust the pose of the structural component so that the feature of the hole to be measured is unobstructed and visible in the direction of measurement movement; Step S 53 Obtain the direction vector V3 from the observation point to the center of the hole; Step S 54 Determine the angle β between vector V3 and vector V4 in a certain direction of the hole axis. If the angle β > 90°, then vector V4 is the feed direction of the hole feature moving along the hole axis into the hole. Otherwise, the feed direction of the hole feature moving along the hole axis is the vector opposite to the direction of vector V4.

[0015] The step between steps S4 and S5 includes identifying stepped combination holes in the hole features through "face-to-face" topological adjacency relationships.

[0016] The method for identifying stepped combination holes specifically includes the following steps: obtaining groups of hole features with the same axis among all hole features; determining whether any two hole features in the group are adjacent to the same plane or the same conical surface according to the "face-to-face" topological adjacency relationship; if not, there are no stepped combination holes in the group of hole features; if so, and the normal of the adjacent plane is parallel to the hole axis, and the axis of the conical surface is coaxial with the hole axis, then these two hole features form a stepped combination hole; if so, and multiple hole features are sequentially adjacent to the same plane or conical surface, and the normal of the adjacent plane is parallel to the hole axis, and the axis of the conical surface is coaxial with the hole axis, then a multi-layered stepped combination hole is formed.

[0017] The measurement direction of the stepped combination hole along the hole axis is from the end with the larger hole diameter to the end with the smaller hole diameter.

[0018] Compared with existing technologies, the beneficial effects of this invention are as follows: 1. The generation method of this invention is based on the contour surface geometric characteristics and local topological relationships of the three-dimensional geometric model of the structural component, meeting the requirements for efficient and accurate identification of hole features in complex, non-fixed topological aircraft structural components. Compared with feature recognition methods that rely on the global topological structure of the three-dimensional model or are based on neural networks and deep learning, this invention avoids the construction of complex adjacency attribute graphs and the interference of redundant geometric nodes on hole feature extraction, significantly reducing the time complexity and computational overhead of the algorithm and improving recognition efficiency.

[0019] Specifically, this method first extracts the cylindrical geometric elements of all contour surfaces in the 3D model of the structural component. Then, based on the "face-to-face" topological adjacency relationship and the cylindrical surface radius parameter, the cylindrical surfaces are grouped. Next, by judging geometric constraints such as the consistency of the axial direction, the concavity and convexity of the surfaces, the sum of the central angles, and the radial contour closure of each group of cylindrical surfaces, the effective combinations of cylindrical surfaces constituting the hole features are identified. Based on this, the identified hole features are classified and sorted according to geometric attributes such as hole diameter, hole depth, and hole position.

[0020] To address the shortcomings of traditional methods where the measurement feed path for hole features along the axial direction requires manual setting and lacks versatility, this invention proposes an adaptive measurement direction planning method based on the observer's viewpoint pose. This method automatically determines the optimal measurement feed direction for each hole feature based on the spatial pose of the 3D model at a specified viewpoint, thereby achieving intelligent and batch generation of hole group detection paths.

[0021] In summary, this invention focuses on local geometric features and topological relationships to directly extract key identification information. While avoiding the complexity of global topological modeling, it uses a geometric constraint-guided elimination method to achieve accurate identification of hole features. At the same time, by combining the pose information from the observer's perspective, it achieves adaptive planning of the measurement direction, which significantly improves the automation, versatility, and engineering feasibility of hole group detection data generation. It can be widely applied to the automatic generation of batch data for the inspection of large and complex aircraft structural components.

[0022] 2. This invention automatically clusters related cylindrical surfaces based on "face-to-face" topological adjacency and radius consistency, avoiding global topological analysis, significantly improving processing efficiency, and eliminating interference from non-hole features. It uses matrix-based storage of identifiers, resulting in a clear structure that facilitates subsequent traversal and iterative processing.

[0023] 3. This invention uses the correlation analysis between the normal vector and the material lateral direction to accurately determine the concavity and convexity of the cylindrical surface based on the geometric vector relationship, effectively distinguishing between hole features (concave surfaces) and non-hole structures such as bosses.

[0024] 4. This invention extracts closed loops using a "face-edge" graph and verifies that the sum of the normal vectors' wrapping angles is 360°, achieving robust identification of complex hole structures (such as countersunk holes and through holes). Furthermore, it introduces a mechanism for optimizing the projection length of shared edges to eliminate redundant edge interference and improve the stability and accuracy of graph structure construction.

[0025] 5. The present invention also includes the identification of stepped combination holes, which improves the completeness of hole feature classification and provides a more accurate geometric basis for subsequent measurement path planning, avoiding detection errors caused by misjudgment.

[0026] 6. Based on the spatial relationship between the observation angle and the hole axis, this invention can adaptively plan the feed direction, realizing the automatic connection of the entire process from geometric recognition to measurement direction planning. Attached Figure Description

[0027] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments, wherein: Figure 1 is a schematic flowchart of the present invention. Detailed Implementation

[0028] Example 1 serves as a basic implementation of the present invention. The present invention includes an adaptive generation method for hole group detection data of large aircraft structural components, comprising the following steps: Step S1. Extracting the geometry of all cylindrical surfaces on the contour surface of the three-dimensional geometric model of the structural component to form a set of cylindrical surfaces S. c .

[0029] Step S2. Based on the "face-to-face" topological adjacency relationship and the cylinder radius parameter, set the cylindrical surfaces S. c Divided into multiple subsets S c(i) (i=1,2,…).

[0030] Step S3. Based on the consistency of the cylindrical surface axial direction, the concavity and convexity of the cylindrical surface, and the radial closure of the cylindrical surface in each subset, from set S c Remove subsets that do not meet the requirements. Set S c The cylindrical surfaces within the remaining subset form hole features, thus completing the geometric identification of the hole features.

[0031] Step S4. Identify the hole feature parameters and group and sort the hole features based on the hole feature parameters.

[0032] Step S5. Based on the pose of the three-dimensional geometric model in the observation view, adaptively plan the feed direction of the hole feature along the hole axis measurement motion.

[0033] Step S6. Based on the hole feature measurement direction, parameters, and measurement coordinate system, batch group and create path optimization measurement data.

[0034] Example 2 is a preferred embodiment of the present invention. The present invention includes an adaptive generation method for hole group detection data of large aircraft structural components, comprising the following steps: Step S1. Extract all cylindrical surface geometry of the three-dimensional geometric model contour surface of the structural component to form a set of cylindrical surfaces S. c .

[0035] Step S2. Based on the "face-to-face" topological adjacency relationship and the cylinder radius parameter, set the cylindrical surfaces S. c Divided into multiple subsets S c (i) (i=1,2,…). Specifically, all cylindrical surfaces that are directly adjacent in topology and have the same radius parameter are grouped into the same subset.

[0036] Step S3. Based on the consistency of the cylindrical surface axial direction, the concavity and convexity of the cylindrical surface, and the radial closure of the cylindrical surface in each subset, from set S c Remove subsets that do not meet the requirements. Set S c The cylindrical surfaces within the remaining subset form hole features, thus completing the geometric identification of the hole features.

[0037] Specifically, determine the consistency of the axial directions of the cylindrical surfaces within each subset Sc(i) (i=1,2,…). Iterate through all subsets; for any subset, select any two cylindrical surfaces. If their axial directions differ, remove the subset from Sc. Check the concavity / convexity of the cylindrical surfaces and remove subsets containing convex cylindrical surfaces from Sc. Check the radial closure of the cylindrical surfaces within a subset and remove subsets containing topologically unclosed cylindrical surfaces from Sc.

[0038] Step S4. Identify hole feature parameters and group and sort the hole features based on these parameters. These parameters include, but are not limited to: hole diameter, hole depth, hole position, and hole axis direction.

[0039] Step S5. Based on the pose of the three-dimensional geometric model in the observation view, adaptively plan the feed direction of the hole feature along the hole axis measurement motion.

[0040] Step S6. Based on the hole feature measurement direction, parameters, and measurement coordinate system, batch group and create path optimization measurement data.

[0041] Example 3, as another preferred embodiment of the present invention, includes an adaptive generation method for hole group detection data of large aircraft structural components, comprising the following steps: Step S1. Extracting the geometry of all cylindrical surfaces on the contour surface of the three-dimensional geometric model of the structural component to form a set of cylindrical surfaces S. c .

[0042] Step S2. Based on the "face-to-face" topological adjacency relationship and the cylinder radius parameter, set the cylindrical surfaces S. c Divided into multiple subsets S c (i) (i=1,2,…).

[0043] Step S3. Based on the consistency of the cylindrical surface axial direction, the concavity and convexity of the cylindrical surface, and the radial closure of the cylindrical surface in each subset, from set S c Remove subsets that do not meet the requirements. Set S c The cylindrical surfaces within the remaining subset form hole features, thus completing the geometric identification of the hole features.

[0044] Step S4. Identify the hole feature parameters and group and sort the hole features based on the hole feature parameters.

[0045] Step S5. Identify stepped combination holes in the hole features through "face-to-face" topological adjacency relationships. This specifically includes the following steps: Step S 51 Obtain all hole feature groups with the same centerline; Step S 52 Based on the "face-to-face" topological adjacency relationship, determine whether any two hole features within the group are adjacent to the same plane or the same cone surface. If so, proceed to step S. 53 Otherwise, there are no stepped combination holes in this group of hole features; Step S 53 If the normal of the adjacent plane is parallel to the hole axis and the axis of the cone is coaxial with the hole axis, then these two hole features form a stepped combination hole; if multiple hole features are successively adjacent to the same plane or cone, and the normal of the adjacent plane is parallel to the hole axis and the axis of the cone is coaxial with the hole axis, then they form a multi-layered stepped combination hole.

[0046] Step S6. Adaptively plan the feed direction of the hole feature along the hole axis measurement using the pose of the 3D geometric model in the viewing view. This specifically includes the following steps: Step S 61 Obtain the central axis of the hole by using the axes of the cylindrical surfaces that make up the hole; Step S 62 Adjust the pose of the structural component so that the feature of the hole to be measured is visible without obstruction in the direction of measurement movement; Step S 63 Obtain the direction vector V3 from the observation point to the center of the hole; Step S 64 Determine the angle β between vector V3 and vector V4 in a certain direction of the hole axis. If the angle β > 90°, then vector V4 is the feed direction of the hole feature moving along the hole axis into the hole. Otherwise, the feed direction of the hole feature moving along the hole axis is the vector opposite to the direction of vector V4.

[0047] Step S7. Based on the hole feature measurement direction, parameters, and measurement coordinate system, batch group and create path optimization measurement data.

[0048] Example 4, as another preferred embodiment of the present invention, includes an adaptive generation method for hole group detection data of large aircraft structural components. Referring to Figure 1 in the specification, it includes the following steps: Step S1. Extract all cylindrical surface geometry of the three-dimensional geometric model contour surface of the structural component to form a set of cylindrical surfaces S. c .

[0049] Step S2. Assemble the set S of cylindrical surfaces according to the "face-to-face" topological adjacency relationship and the cylindrical surface radius parameter. c Divided into multiple subsets S c (i) (i=1,2,…). Specifically, all cylindrical surfaces that are directly adjacent in topology and have the same radius parameter are divided into the same subset, which includes the following steps: Step S 21 Obtain the unique identifiers of each cylindrical surface on the 3D geometric model in the 3D computer-aided design software (3D CAD software), forming a set S of cylindrical surface identifiers. t Step S 22 Obtain the radius parameter values ​​for each cylindrical surface; Step S 23 Establish an extensible identifier matrix T and a set S. t The first element is used as the first element T of the identifier matrix. （1，1） Empty elements in matrix T are filled with 0, and each row of matrix T is an identifier for a cylinder with the same radius; Step S 24 Traverse the set S of cylindrical surface identifiers. t For all elements, in order, perform step S on the cylindrical surface f1 corresponding to each identifier. 25 Step S 25Obtain all topologically adjacent surfaces of cylindrical surface f1, and determine whether each topologically adjacent surface is a cylindrical surface; if any topologically adjacent surface f2 is a cylindrical surface, then proceed to step S. 26 After traversal ends, jump to step S. 24 Step S 26 Determine if the radius parameters of the adjacent surface f2 and the cylindrical surface f1 are the same. If the radii are the same, then group the adjacent cylindrical surface f2 into the same group as the cylindrical surface f1, add the identifier corresponding to the cylindrical surface f2 to the row containing the identifier corresponding to the cylindrical surface f1 in matrix T, and then add the identifier from the identifier set S. t Delete the identifier corresponding to the adjacent surface f2; if the radii are not the same, the subsequent steps create a new group for the adjacent cylindrical surface f2 and add the identifier corresponding to the cylindrical surface f2 to the new row of matrix T; step S 27 Treat the adjacent cylindrical surface f2 as the cylindrical surface f1, and iteratively execute step S. 25 Step S 28 Complete the set of cylindrical surface identifiers S. t By traversing and judging all elements, we obtain the identifier matrix T(i,j), where each identifier corresponds to a cylindrical surface, i is the number of cylindrical surface groups, j is the number of cylindrical surfaces in each group, and the radius parameter of each group of cylindrical surfaces is the same.

[0050] Step S3. Based on the consistency of the cylindrical surface axial direction, the concavity and convexity of the cylindrical surface, and the radial closure of the cylindrical surface in each subset, from set S c Delete subsets that do not meet the requirements; set S c The cylindrical surfaces within the remaining subset form hole features, thus completing the geometric identification of the hole features.

[0051] Based on the consistency of the cylindrical axis orientation in each subset, from set S c Deleting subsets that do not meet the requirements specifically means: traversing all subsets, if any two randomly selected cylinders in any subset have axes with different directions, then removing them from set S. c Delete that subset.

[0052] Based on the concavity and convexity of the cylindrical surfaces of each subset, from set S c Deleting subsets that do not meet the requirements includes the following steps: Step S 31 Obtain the normal vector V1 at any point P on the cylindrical surface; step S 32 Obtain the material side direction at point P. If the normal vector V1 is opposite to the material side direction, reverse the normal vector V1. Here, the material side direction is defined as the direction from any point on the surface of the part geometry to the geometric entity.

[0053] Step S 33 Construct a vector V2 pointing from point P to the axis of the cylindrical surface; step S34 Determine the angle α between normal vectors V1 and V2. If the angle α = 0°, the cylindrical surface is a convex cylindrical surface; otherwise, it is a concave cylindrical surface. (Step S) 35 From set S c Remove the subset containing convex cylindrical surfaces.

[0054] Based on the radial closure of the cylindrical surfaces of each subset, from set S c Deleting subsets that do not meet the requirements specifically refers to deleting subsets containing radially unclosed cylindrical surfaces. The method for determining whether a cylindrical surface is radially closed includes the following steps: Step 1. Extract the edges of each cylindrical surface within the subset; Step 2. Using the unique identifier of each edge in the 3D computer-aided design software (3D CAD software), obtain all common edges between any two cylindrical surfaces, forming set S. e Step 3. If two cylindrical surfaces share multiple edges, and among these shared edges there exists a pair of consecutive first-order shared edges, retain the shared edge with the longest projected length on the axis of the cylindrical surface. If the projected lengths of all shared edges on the axis of the cylindrical surface are zero, then retain the shared edge with the longest length. From set S... e By deleting all shared edges except the original shared edges from each pair of consecutive first-order shared edges, we obtain the set S of shared edges. e Step 4. Set S using cylindrical surfaces and shared edges. e 'Construct a face-edge graph G of the cylindrical surfaces and their shared edges within the subset, where faces are graph nodes and shared edges of the cylindrical surfaces are edges in the graph; Step 5. Extract the closed loop L of the face nodes in the face-edge graph G.' p Step 6. If a closed loop exists, starting from any node of the closed loop, traverse each face node in the loop sequentially along the closed loop path. If the sum of the angles between any two center normal vectors of each face node is 360°, then the cylindrical surface within the subset is radially closed.

[0055] Step S4. Identify pore feature parameters and group and sort the pore features based on these parameters. Specifically, sort them in ascending or descending order according to the pore diameter parameters to facilitate grouping of the detection data.

[0056] Step S5. Identify stepped combination holes in the hole features through "face-to-face" topological adjacency relationships. This specifically includes the following steps: Step S 51 Obtain all hole feature groups with the same centerline; Step S 52 Based on the "face-to-face" topological adjacency relationship, determine whether any two hole features within the group are adjacent to the same plane or the same cone surface. If so, proceed to step S. 53 Otherwise, there are no stepped combination holes in this group of hole features; Step S 53If the normal of the adjacent plane is parallel to the hole axis and the axis of the cone is coaxial with the hole axis, then these two hole features form a stepped combination hole; if multiple hole features are successively adjacent to the same plane or cone, and the normal of the adjacent plane is parallel to the hole axis and the axis of the cone is coaxial with the hole axis, then they form a multi-layered stepped combination hole.

[0057] Step S6. Based on the pose of the 3D geometric model in the observation viewpoint, adaptively plan the feed direction of the hole feature measurement motion along the hole axis, and plan from which hole opening the probe enters the hole for measurement. Specifically, this includes the following steps: Step S 61 Obtain the central axis of the hole by using the axes of the cylindrical surfaces that make up the hole; Step S 62 Adjust the pose of the structural component so that the feature of the hole to be measured is visible without obstruction in the direction of measurement movement; Step S 63 Obtain the direction vector V3 from the observation point to the center of the hole; Step S 64 Determine the angle β between vector V3 and vector V4 in a certain direction of the hole axis. If the angle β > 90°, then vector V4 is the feed direction of the hole feature moving along the hole axis into the hole. Otherwise, the feed direction of the hole feature moving along the hole axis is the vector opposite to the direction of vector V4.

[0058] Among them, the measurement direction of the movement of the stepped combination hole along the hole axis is from the end with the larger hole diameter to the end with the smaller hole diameter.

[0059] Step S7. Based on the hole feature measurement direction, parameters, and measurement coordinate system, batch group and create path-optimized measurement data to obtain hole inspection data according to the hole feature parameters (hole diameter, type). Here, the measurement direction refers to the feed direction of the measurement motion along the hole axis, and the measurement coordinate system refers to the workpiece coordinate system when the measuring machine executes the measurement program. The measurement data refers to the hole diameter, hole position, and normal vector, i.e., point position and vector; path optimization involves arranging these points according to the optimized motion path sequence.

[0060] Since the tolerance requirements for holes of different diameters are different, manufacturing accuracy evaluations need to be performed separately. Therefore, this embodiment can group holes based on their diameters. Because the distribution of holes on parts is not regular, and parts are generally not regular either, optimizing the path can reduce repetitive motion trajectories and redundant motion time.

[0061] In summary, any other corresponding modifications made by those skilled in the art after reading this invention document, without requiring creative mental effort, based on the technical solutions and concepts of this invention, are all within the scope of protection of this invention.

Claims

1. An adaptive generation method for hole group detection data of large aircraft structural components, characterized in that: The steps include: Step S1. Extracting the geometry of all cylindrical surfaces on the contour surface of the 3D geometric model of the structural component, forming a set of cylindrical surfaces S. c Step S2. Based on the "face-to-face" topological adjacency relationship and the cylindrical surface radius parameter, set the cylindrical surfaces S... c Divided into multiple subsets S c (i) (i=1,2,…); Step S3. Based on the consistency of the cylindrical surface axial direction, the concavity and convexity of the cylindrical surface, and the radial closure of the cylindrical surface in each subset, from set S c Delete subsets that do not meet the requirements; set S c The cylindrical surfaces within the remaining subset form hole features, completing the identification of hole feature geometry; Step S4. Identify hole feature parameters and group and sort the hole features based on the hole feature parameters; Step S5. Adaptively plan the feed direction of the hole feature measurement motion along the hole axis using the pose of the 3D geometric model in the observation view; Step S6. Based on the hole feature measurement direction, parameters, and measurement coordinate system, batch group and create path-optimized measurement data.

2. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 1, characterized in that: The division in step S2 is based on the following: all cylindrical surfaces that are directly adjacent in topology and have the same radius parameter are divided into the same subset.

3. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 2, characterized in that: Step S2 specifically includes the following steps: Step S 21 Obtain the unique identifiers for each cylindrical surface on the 3D geometric model, forming a set S of cylindrical surface identifiers. t Step S 22 Obtain the radius parameter values ​​for each cylindrical surface; Step S 23 Establish an extensible identifier matrix T and a set S. t The first element is used as the first element T of the identifier matrix. （1，1） Empty elements in matrix T are filled with 0, and each row of matrix T is an identifier for a cylinder with the same radius; Step S 24 Traverse the set S of cylindrical surface identifiers. t For all elements, in order, perform step S on the cylindrical surface f1 corresponding to each identifier. 25 Step S 25 Obtain all topologically adjacent surfaces of cylindrical surface f1, and determine whether each topologically adjacent surface is a cylindrical surface; if any topologically adjacent surface f2 is a cylindrical surface, then proceed to step S. 26 After traversal ends, jump to step S. 24 Step S 26 Determine if the radius parameters of the adjacent surface f2 and the cylindrical surface f1 are the same. If the radii are the same, then group the adjacent cylindrical surface f2 into the same group as the cylindrical surface f1, add the identifier corresponding to the cylindrical surface f2 to the row containing the identifier corresponding to the cylindrical surface f1 in matrix T, and then add the identifier from the identifier set S. t Delete the identifier corresponding to the adjacent surface f2; if the radii are not the same, the subsequent steps create a new group for the adjacent cylindrical surface f2 and add the identifier corresponding to the cylindrical surface f2 to the new row of matrix T; step S 27 Treat the adjacent cylindrical surface f2 as the cylindrical surface f1, and iteratively execute step S. 25 Step S 28 Complete the set of cylindrical surface identifiers S. t By traversing and judging all elements, we obtain the identifier matrix T(i,j), where each identifier corresponds to a cylindrical surface, i is the number of cylindrical surface groups, j is the number of cylindrical surfaces in each group, and the radius parameter of each group of cylindrical surfaces is the same.

4. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 1, characterized in that: Based on the consistency of the cylindrical axis orientation in each subset, from set S c Deleting subsets that do not meet the requirements specifically means: traversing all subsets, if any two randomly selected cylinders in any subset have axes with different directions, then remove them from set S. c Delete that subset.

5. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 1, characterized in that: Based on the concavity and convexity of the cylindrical surfaces of each subset, from set S c Deleting non-compliant subsets specifically includes the following steps: Step S 31 Obtain the normal vector V1 at any point P on the cylindrical surface; step S 32 Obtain the material side direction at point P. If the normal vector V1 is opposite to the material side direction, reverse the normal vector V1. The material side direction is the direction from any point on the surface of the part geometry to the geometric entity. Step S 33 Construct a vector V2 pointing from point P to the axis of the cylindrical surface; step S 34 Determine the angle α between normal vectors V1 and V2. If the angle α = 0°, the cylindrical surface is a convex cylindrical surface; otherwise, it is a concave cylindrical surface. (Step S) 35 From set S c Remove the subset containing convex cylindrical surfaces.

6. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 1, characterized in that: Based on the radial closure of the cylindrical surfaces of each subset, from set S c Deleting subsets that do not meet the requirements specifically refers to deleting subsets that have radially unclosed cylindrical surfaces.

7. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 6, characterized in that: The method for determining whether a cylindrical surface is radially closed includes the following steps: Step 1. Extract the edges of each cylindrical surface within the subset; Step 2. Using the unique identifier of each edge, obtain all common edges between any two cylindrical surfaces, forming a set S. e Step 3. If two cylindrical surfaces share multiple edges, and among these shared edges there exists a pair of consecutive first-order shared edges, retain the shared edge with the longest projected length on the axis of the cylindrical surface. If the projected lengths of all shared edges on the axis of the cylindrical surface are zero, then retain the shared edge with the longest length. From set S... e By deleting all shared edges except the original shared edges from each pair of consecutive first-order shared edges, we obtain the set S of shared edges. e Step 4. Set S using cylindrical surfaces and shared edges. e 'Construct a "face-edge" graph G of the cylindrical surfaces and their shared edges within the subset, where faces are graph nodes and the shared edges of the cylindrical surfaces are edges in the graph; Step 5. Extract the closed loop L of the face nodes in the "face-edge" graph G. p Step 6. If a closed loop exists, starting from any node of the closed loop, traverse each face node in the loop sequentially along the closed loop path. If the sum of the angles between any two center normal vectors of each face node is 360°, then the cylindrical surface within the subset is radially closed.

8. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 1, characterized in that: Step S5 specifically includes the following steps: Step S 51 Obtain the central axis of the hole by using the axes of the cylindrical surfaces that make up the hole; Step S 52 Adjust the pose of the structural component so that the feature of the hole to be measured is visible without obstruction in the direction of measurement movement; Step S 53 Obtain the direction vector V3 from the observation point to the center of the hole; Step S 54 Determine the angle β between vector V3 and vector V4 in a certain direction of the hole axis. If the angle β > 90°, then vector V4 is the feed direction of the hole feature moving along the hole axis into the hole. Otherwise, the feed direction of the hole feature moving along the hole axis is the vector opposite to the direction of vector V4.

9. An adaptive generation method for hole group detection data of large aircraft structural components according to any one of claims 1 to 8, characterized in that: Between steps S4 and S5, the method further includes: identifying stepped combination holes in the hole features through "face-to-face" topological adjacency relationships.

10. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 9, characterized in that: The method for identifying stepped combination holes specifically includes the following steps: obtaining groups of hole features with the same axis among all hole features; determining whether any two hole features in the group are adjacent to the same plane or the same conical surface according to the "face-to-face" topological adjacency relationship; if not, there are no stepped combination holes in the group of hole features; if yes, and the normal of the adjacent plane is parallel to the hole axis, and the axis of the conical surface is coaxial with the hole axis, then these two hole features form a stepped combination hole; if yes, and multiple hole features are sequentially adjacent to the same plane or conical surface, and the normal of the adjacent plane is parallel to the hole axis, and the axis of the conical surface is coaxial with the hole axis, then a multi-layered stepped combination hole is formed.

11. The adaptive generation method for hole group detection data of large aircraft structural components according to claim 10, characterized in that: The measurement direction of the stepped combination hole along the hole axis is from the end with the larger hole diameter to the end with the smaller hole diameter.

Images