A one-click drilling method and system for parts within a unit

Abstract

This invention proposes a one-click hole-drilling method and system for parts within a unit. The software identifies the input digital model, converts it into a point cloud, performs preprocessing on the point cloud data including denoising and enhancement, and acquires existing hole information within the point cloud range. A hole database is built, and based on the existing hole information and the information stored in the database, pre-drilling is performed on the components with holes. The pre-drilled holes are checked; if the check fails, the process returns; if it passes, the process proceeds to the next step. Finally, it is determined whether the hole on the component is a through hole or a blind hole, and the drilling operation is executed. This invention automates the hole-drilling operation during the design of parts within a unit through software, making the overall hole-drilling and checking operation more standardized, and saving time and personnel costs.

Description

Technical Field

[0001] This invention belongs to the field of automotive automation design, and in particular relates to a one-click drilling method and system for parts within a unit. Background Technology

[0002] In automobiles, intra-unit parts refer to the basic components that make up a specific functional unit of the car body. Multiple intra-unit parts are assembled to form a fully functional unit. One common method for assembling intra-unit parts is through drilling.

[0003] Drilling holes in components within a unit typically involves creating openings on the adjacent component (the assembly that fits with the perforated component) based on the existing holes in the perforated component. The traditional drilling method uses CATIA's built-in drilling function, following step-by-step prompts: selecting the component to be drilled, manually selecting the drilling function, selecting the object on the component according to software prompts, defining the hole (hole type and location) after drilling, and, if it's a group of holes, drilling is performed hole by hole according to requirements. These steps are complex and repetitive, requiring manual drilling. Only single-hole design is possible; array or combination holes cannot be designed directly. Different hole types on different components can only be designed individually. It cannot provide reasonable solutions based on specific locations and drilling areas, nor can it output the specific hole positions and operations for adjustment. Summary of the Invention

[0004] This invention proposes a one-click drilling method and system for parts within a unit. The software automates the drilling operation during the design of parts within the unit, making the overall operation of drilling and verification more standardized and saving time and personnel costs.

[0005] To achieve the above objectives, the technical solution of the present invention is implemented as follows: A one-click drilling method for parts within a unit includes: S1. The software identifies the input digital model, converts the digital model into a point cloud, performs preprocessing on the point cloud data including denoising and enhancement, and obtains existing hole information within the point cloud range. S2. Establish a hole database, which stores hole rules, including part type, hole group arrangement, common hole diameter and hole spacing, hole mating relationship, priority rules, accuracy standards, and technical specifications; the part type includes hole-bearing parts and mating parts. S3. Based on the existing hole information and the information stored in the hole database, perform pre-hole processing on the handpiece of the hole-filled part. S4. Verify the pre-drilled hole. If the verification fails, return to step S3; if the verification passes, proceed to the next step. The verification method includes: using the center point of the pre-drilled hole as the center point of the curved surface, and using the normal direction of the intersection surface of the holed part and the counterpart as the axis direction to generate the central axis, generating an axial surface with the same radius as the pre-drilled hole. Select four lines evenly distributed in the same direction as the central axis on the axial surface. The angle formed by each pair of the four lines and the axis is 90 degrees. Check whether the interference amount between the four lines and the counterpart is the same. If they are the same, the verification passes; if they are not the same, the verification fails. S5. Determine whether the opening of the opposing component is a through hole or a blind hole, and perform the opening operation.

[0006] Furthermore, the enhancement of point cloud data in step S1 includes multi-data fusion, generating multi-dimensional point clouds of different precision for the same model using point cloud data, and fusing the point clouds generated multiple times.

[0007] Furthermore, the acquisition of existing hole information in step S1 employs a method combining normal vector analysis, region clustering, and non-convex robust circle fitting, including: Local normal vector and geometric feature calculation: Construct a neighborhood for the point cloud data, and use principal component analysis (PCA) to calculate the normal vector and curvature of each point; Based on region clustering of normal differences, candidate point sets for holes are extracted: using region growing clustering, the point sets of regions with abrupt changes in normal are automatically extracted as candidate point sets for hole boundaries; Establish a 3D hole model and set engineering constraints: Construct a 3D hole model based on the candidate point set of the hole boundary, and set upper and lower limits of the radius in combination with the hole diameter range of the hole database; GNC-TLS Non-convex Robust Circle Fitting: By optimizing GNC and global least squares TLS through generalized non-convexity optimization, the weights of outliers are automatically reduced by progressively increasing non-convexity, thus achieving deterministic fitting without random sampling. Output hole features: Output hole center coordinates, radius, and direction vector, which serve as the basis for subsequent hole design of hand parts.

[0008] Furthermore, in step S1, the acquisition of existing hole information involves obtaining information about circles within the point cloud range through a random sample consensus algorithm. This includes fitting a 3D circle using the random sample consensus algorithm, inputting the direction vector value of the circle, and constraining the radius range based on commonly used data to obtain information about circles within this radius range in the point cloud. The circle is then identified as an existing hole, serving as the basis for subsequent hole design of the handpiece.

[0009] Furthermore, in step S5, the method for determining whether an opening is a through hole or a blind hole includes: Select the axis intersection line of the existing hole and the pre-drilled hole. If the axis intersection line of the pre-drilled hole side is shorter, then open the hole. If the axis intersection line of the pre-drilled hole side is longer, set the through hole threshold. If the axis intersection line of the pre-drilled hole side is less than or equal to the through hole threshold, then open the hole. If it is greater than the through hole threshold, then open the blind hole.

[0010] In another aspect, the present invention also proposes a one-click drilling system for parts within a unit, comprising: Recognition and Conversion Module: The software recognizes the input digital model, converts the digital model into point cloud form, performs preprocessing on the point cloud data including denoising and enhancement, and obtains existing hole information within the point cloud range; Database module: This module establishes a hole database, which stores hole rules, including part type, hole group arrangement, common hole diameters and spacing, hole mating relationships, priority rules, precision standards, and technical specifications. The part types include perforated parts and mating parts. Pre-drilling module: Based on the existing hole information and the information stored in the hole database, pre-drilling is performed on the hand parts of the perforated parts; Verification Module: Verifies the pre-drilled hole. If the verification fails, it returns to the pre-drilled hole module; if the verification passes, it enters the hole-drilling module. The verification includes: using the center point of the pre-drilled hole as the center point of the curved surface, and using the normal direction of the intersection surface of the holed part and the counterpart as the axis direction to create a central axis, generating an axial surface with the same radius as the pre-drilled hole. Selecting four lines evenly distributed in the same direction as the central axis on the axial surface, with each pair of lines forming an angle of 90 degrees with the axis, and checking whether the interference amount between the four lines and the counterpart is the same. If they are the same, the verification passes; if they are not the same, the verification fails. Hole-opening module: Determines whether the hole in the device is a through hole or a blind hole, and performs the hole-opening operation.

[0011] Furthermore, the enhancement of point cloud data in the recognition and transformation module includes multi-data fusion, which generates multi-dimensional point clouds of different precision for the same model from point cloud data, and performs a fusion operation on the point clouds generated multiple times.

[0012] Furthermore, the identification and conversion module includes: The computational unit is used to calculate local normal vectors and geometric features: it constructs a neighborhood for point cloud data and uses principal component analysis (PCA) to calculate the normal vector and curvature of each point. Candidate point set unit: Based on region clustering of normal differences, the candidate point set of the hole is extracted: Using region growing clustering, the point set of the normal abrupt region is automatically extracted as the candidate point set of the hole boundary; Model unit: Build a 3D hole model and set engineering constraints: Construct a 3D hole model based on the candidate point set of the hole boundary, and set upper and lower limits of radius in combination with the hole diameter range of the hole database; Fitting unit for GNC-TLS non-convex robust circle fitting: By optimizing GNC and global least squares TLS through generalized non-convexity optimization, the weight of outliers is automatically reduced by progressively increasing non-convexity, thus achieving deterministic fitting without random sampling. Output unit, output hole features: output hole center coordinates, radius, and direction vector, which serve as the basis for subsequent hole design of hand parts.

[0013] Furthermore, the identification and conversion module includes: acquiring existing hole information, which involves obtaining information about circles within the point cloud range through a random sample consensus algorithm, including fitting a 3D circle using the random sample consensus algorithm, inputting the direction vector value of the circle, and constraining the radius range according to common data to obtain information about circles within this radius range in the point cloud; identifying the circle as an existing hole, which serves as the basis for subsequent hole design of the handpiece.

[0014] Furthermore, the hole-opening module includes a judgment unit, in which the axis intersection line of the existing hole and the pre-opened hole is selected. If the axis intersection line of the pre-opened hole is shorter, a through hole is opened; if the axis intersection line of the pre-opened hole is longer, a through hole threshold is set. If the axis intersection line of the pre-opened hole is less than or equal to the through hole threshold, a through hole is opened; if it is greater than the through hole threshold, a blind hole is opened.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention improves the ease of use of the identification operation and reduces user costs by identifying the parts within the unit and performing one-click drilling based on the correlation between the parts.

[0016] 2. This invention provides design specifications for array holes and combination holes by building a hole database. It can provide reasonable solutions based on specific locations and drilling areas, and realize automated drilling operations for different hole types on different parts.

[0017] 3. This invention provides a one-click drilling logic for parts within a unit. By using software recognition and judgment, the steps of manually drilling parts within a unit are simplified to software operation. After inputting the corresponding logic rules, one-click automated drilling can be achieved, making the drilling and verification operation more standardized, the overall operation more efficient and accurate, reducing the professional requirements for drilling parts within a unit, and saving time and personnel costs. Attached Figure Description

[0018] Figure 1 This is a flowchart of Embodiment 1 of the present invention.

[0019] Figure 2 This is a schematic diagram of the point cloud of the digital-to-analog conversion in Embodiment 1 of the present invention.

[0020] Figure 3This is a schematic diagram of the part bonding surface in Embodiment 1 of the present invention.

[0021] Figure 4 This is a schematic diagram of the part types in Embodiment 1 of the present invention.

[0022] Figure 5 This is a schematic diagram of the pre-drilled hole in Embodiment 1 of the present invention.

[0023] Figure 6 This is a schematic diagram of the pre-drilling verification in Embodiment 1 of the present invention.

[0024] Figure 7 This is a schematic diagram of the central axis of Embodiment 1 of the present invention.

[0025] Figure 8 This is a schematic diagram of a through hole / blind hole according to Embodiment 1 of the present invention.

[0026] Figure 9 This is a schematic diagram of the system structure of Embodiment 2 of the present invention. Detailed Implementation

[0027] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0028] To make the purpose and features of this invention patent clearer and easier to understand, the specific embodiments of this invention patent will be further described below with reference to the accompanying drawings.

[0029] Example 1: The one-click drilling method for parts within a unit proposed in this embodiment, such as Figure 1 As shown, it includes: S1. The software identifies the input digital model, converts it into a point cloud, performs preprocessing on the point cloud data including denoising and enhancement, and obtains existing hole information within the point cloud range.

[0030] In this step, the software interface allows users to select and input the digital model file of the part to be drilled. The software uses API functions to call 3D design software to read the digital model file. In this embodiment, the 3D design software is CATIA, which reads the model under the SM node of CATIA and then allows the user to click and select.

[0031] After inputting the digital model file, perform data conversion and structure recognition operations on the digital model file: The software recognizes the input digital model and converts it into a point cloud. This conversion can be achieved by calling CATIA's built-in function modules via API functions, such as the Digitized Shape Editor module. Using the point cloud generation tool within this module, parameters such as point density (spacing, quantity) and distribution (uniform, random) can be set to convert the digital model into point cloud data. Alternatively, point cloud data can be obtained through laser scanning or by using other digital model design software for point cloud data conversion.

[0032] like Figure 2 As shown, the left image is the digital model file, and the right image is the converted point cloud.

[0033] After being converted into point cloud data, the point cloud data needs to be preprocessed. The main preprocessing method is to denoise and enhance the point cloud data.

[0034] For example, the point cloud denoising method used includes: calculating the average point distance of the point cloud (such as the rasterized bounding box method), using Euclidean clustering, and separating noise points based on local density differences with a tolerance of 0.2 mm.

[0035] For example, point cloud enhancement adopts a multi-data fusion mode, which generates multiple point clouds from the same data and then merges the generated point clouds.

[0036] The point cloud is generated multiple times from the same data, which means that the point cloud is converted multiple times using the digital model file. Since CATIA converts the digital model into a point cloud based on the geometry and structure, there may be slight differences in the fitting and discretization of the surface in each conversion process. The internal algorithm and calculation process may have a certain degree of randomness or uncertainty, which leads to certain differences in the point cloud data of each conversion. The point cloud enhancement uses the point cloud data generated multiple times to perform multi-data fusion.

[0037] During the fusion process, it is necessary to ensure that the point cloud data generated multiple times are in the same coordinate system. The method is as follows: find corresponding point pairs (Pi, Pj) to align the point clouds generated in different times, ensuring that they are in the same coordinate system. A corresponding point pair (Pi, Pj) refers to the same feature point, which is Pi in the first input point cloud and Pj in the subsequently transformed point cloud. Through this set of corresponding points, the positional deviation between the two point clouds can be calculated, thereby unifying all point clouds to the same coordinate system.

[0038] For example, a method to find corresponding point pairs (Pi, Pj) can be to identify feature points in a point cloud using an algorithm, such as sharp corners, inflection points of curved surfaces, and the centers of regular holes. The positions of these feature points are relatively fixed in different point clouds. The software can automatically determine that the feature point Pi in the initially input point cloud and the feature point Pj in the subsequently transformed point cloud share the same feature, thus automatically generating corresponding point pairs.

[0039] After generating pairs of points with the same name, it is necessary to ensure that the transformation from one coordinate system to another can be completed through the same transformation relation (R, m, T), that is: ; ; Where X represents the position of the point in the horizontal direction; Y represents the position of the point in the vertical direction; Z represents the height of the point in the horizontal direction; x, y, z represent the horizontal, vertical, and height description parameters in the vehicle coordinate system; m represents the scale factor; R and T represent the rotation matrix and translation matrix, respectively; (x0, y0, z0) is defined as the translation vector, used to describe the change of the origin position when translating from coordinate system O-XYZ to another coordinate system O'-X'Y'Z'; (α, β, γ) are the rotation parameters.

[0040] By using the optimal values ​​of the rotation matrix R and the translation matrix T, the point cloud data generated multiple times are unified into the same coordinate system for subsequent fusion, thereby enhancing the point cloud data.

[0041] Methods for obtaining the optimal values ​​of the rotation matrix R and the translation matrix T, for example, include: For each pair of points with the same name, construct its coordinate residual equation, v = R × Pi + T - Pj, where v is the residual vector; Based on the residual equations of all pairs of points with the same name, the normal equations are constructed according to the least squares criterion. Solving the normal equations yields the optimal values ​​of the rotation matrix R and the translation matrix T.

[0042] Then, existing holes in the part can be identified based on the denoised and enhanced point cloud data. The method for identifying existing holes is as follows: Further noise reduction is achieved by employing indirect adjustment, and a recognition method combining normal vector analysis, region clustering, and non-convex robust circle fitting (GNC-TLS) is used to ensure stable recognition of hole information even in situations with high noise, incomplete hole edges, and the presence of chamfers.

[0043] Specifically, it includes: (1) The indirect adjustment method is adopted. The point cloud data (coordinates and angles) is further optimized by the least squares principle to eliminate random errors and obtain the point cloud data that is closest to the real value.

[0044] (2) Calculation of local normal vector and geometric features: Construct a neighborhood for the optimal cloud data, and use principal component analysis (PCA) to calculate the normal vector and curvature of each point.

[0045] (3) Extracting candidate point sets for holes based on region clustering of normal differences: Since the normal of the hole boundary points changes continuously along the circumference and differs significantly from the normal of the surrounding surface, region growing clustering can be used to automatically extract the point sets of the regions with abrupt changes in normal as candidate point sets for the hole boundaries. Region clustering can also be replaced by DBSCAN, KNN clustering, etc.

[0046] (4) Establish a three-dimensional hole model and set engineering constraints: construct a three-dimensional hole model based on the candidate point set of the hole boundary, and set the upper and lower limits of the radius in combination with the hole diameter range of the hole database.

[0047] (5) GNC-TLS Non-convex Robust Circle Fitting: GNC stands for Generalized Non-convex Optimization, which is used to solve non-convex problems such as incomplete data, the presence of outliers, and non-unimodal objective functions. It skips local optima and finds the global optimum that best matches the actual hole shape. TLS stands for Global Least Squares, which is more suitable for fitting 3D point cloud data.

[0048] This step increases non-convexity step by step, causing the weight of outliers to decrease automatically, thus achieving deterministic fitting without random sampling. The fitting results are stable and free from random fluctuations; it is more robust to noise and missing point clouds; and it is suitable for complex structures and chamfered holes.

[0049] In addition, GNC-TLS can be replaced by other robust fitting methods (such as IRLS, LMedS, etc.).

[0050] (6) Output hole features: Output hole center coordinates, radius, and direction vector, which serve as the basis for subsequent hole design of hand parts.

[0051] In addition to the methods mentioned above, existing holes can also be identified using the Random Sample Consensus (RANSAC) algorithm. The RANSAC algorithm, specifically `SACMODEL_CIRCLE3D`, is used to fit a 3D circle. The input is the direction vector value of the circle, and the radius is constrained according to common data. Subsequently, information about the circle within this radius range in the point cloud data can be obtained (coordinates, radius, vector direction, etc.). For example, the specific process includes: (1) Create a SACMODEL_CIRCLE3D sample consistency model in PCL. Input the normal vector direction of the plane containing the circle using the setAxis method to constrain the search space. Use setDistanceThreshold to set the distance threshold between points in the model, for example, 0.01 meters. Use the setRadiusLimits method to constrain the radius range based on common data, for example, set the minimum and maximum radii setRadiusLimits(0.05,0.2).

[0052] (2) Perform RANSAC fitting: Call the computeModel method to perform RANSAC iterative calculation. The algorithm randomly selects the minimum sample set (3 points for a 3D circle) to initialize the model, and iteratively optimizes by evaluating the consistency set (interior points) to finally find the best-fit model.

[0053] (3) Obtaining circle information: After fitting, obtain the model parameter coefficient vector using getModelCoefficients. This vector typically contains, in order: the center coordinates (x, y, z), the radius (r), and the normal vector of the plane containing the circle (nx, ny, nz). The indices of all interior points identified as belonging to the 3D circle can be obtained using getInliers.

[0054] (4) The circle can be identified as an existing hole in a part with a hole. The circle information is the information of the existing hole. Subsequently, the part with a hole is drilled into the part with a hole based on the information of the existing hole.

[0055] S2. Establish a hole database. The hole database stores hole rules, including part type, hole group arrangement, common hole diameter and hole spacing, hole fit relationship, priority rules, accuracy standards, technical specifications, etc.

[0056] The hole database stores hole information for parts. Taking a fixture as an example, the data can cover the following: (1) For workpieces with dimensional tolerance requirements, the corresponding dimensional tolerance of the fixture parts should be 1 / 5 to 1 / 2 of the workpiece tolerance; (2) For straight dimensions of workpieces without tolerance requirements, the corresponding dimensional tolerance of fixture parts can be taken as ±0.1. (3) For angle dimensions of workpieces without angle tolerance requirements, the corresponding angle tolerance of fixture parts can be taken as ±10 points; (4) Tolerance for the center distance L of holes for fasteners. When L is less than 150mm, it can be ±0.1mm; when L is greater than 150mm, it can be ±0.15mm. (5) Ensure that the perpendicularity and parallelism between the surface on which other parts (especially positioning elements) are installed on the fixture and the main base surface is less than 0.01 mm; (6) The parallelism, perpendicularity and coaxiality between planes, between planes and holes, and between holes of specific components such as the Base model, corner brackets, and support clamping blocks should be 1 / 3 to 1 / 2 of the corresponding tolerance of the workpiece. (7) Surface roughness of fixture parts. The surface roughness value of the workpiece surface of the fixture positioning element should be 1-3 numerical segments lower than the surface roughness value of the workpiece positioning reference surface; (8) Priority rule: accuracy requirement > efficiency > cost (e.g., H7 grade holes must use guide sleeves for guidance); (9) Accuracy standards: positioning surface roughness Ra 0.8μm, guide sleeve and drill bit clearance ≤ 0.1mm; (10) Hole arrangement: pin-hole-pin, pin-hole-pin-hole, hole-pin-hole-pin, etc.

[0057] S3. Based on the existing hole information and the information stored in the hole database, perform pre-drilling processing on the handpiece of the hole-bearing part.

[0058] Specifically, it includes: (1) Perform a mating surface analysis on the perforated part to obtain the size of the mating surface between the perforated part and the counterpart part, as well as the direction of the normal to the mating surface; for example Figure 3 The diagram shown is a schematic of the bonding surface.

[0059] The mating surface analysis can be performed by calling CATIA's functional modules through API functions, executing collision detection commands, confirming the mating surface range between the part with the hole and the counterpart, then extracting the mating surface and measuring its dimensions, then selecting the center point of the mating surface, and taking the normal vector at that point as the normal direction of the mating surface.

[0060] (2) Obtain the part type of the mating part (the pair of parts).

[0061] like Figure 4 As shown, the product type can be read from the product attributes of the parts. In this embodiment, the two parts that fit together are both self-made parts.

[0062] (3) Based on the part type, the size of the mating surface, and the direction of the normal to the mating surface, and in conjunction with the opening rules in the opening database, design the pre-opening of the hand part. For example... Figure 5 The diagram shown is a schematic of the pre-drilled hole design for the hand component.

[0063] S4. Verify the pre-drilled hole. If the verification fails, return to step S3. If the verification passes, proceed to the next step.

[0064] The method for verifying the pre-drilled hole includes: Use the center point of the pre-drilled hole as the center point of the curved surface; Using the normal direction of the mating surface between the perforated part and the mating part as the axial direction, a central axis is drawn to generate an axial surface with the same radius as the pre-drilled hole; the central axis is as follows: Figure 6 The light yellow shaft on the opponent's component is shown in the image.

[0065] Choose four lines evenly distributed on the axial plane in the same direction as the central axis, where the angle between any two lines and the axis is 90 degrees. For example... Figure 7 As shown. Then check whether the interference of the four lines with the counterpart is the same. If they are the same, the check is passed and proceed to the next step. If they are not the same, the check is failed and return to step S3 to pre-drill the hole again.

[0066] If it is a hole group design, then each hole in the hole group needs to be checked as described above.

[0067] The above-mentioned pre-drilled hole verification can also employ more measurement line or surface interference algorithms.

[0068] S5. Determine whether the opening of the opposing component is a through hole or a blind hole, and perform the opening operation.

[0069] Select the axis intersection line of the existing hole and the pre-drilled hole. The axis intersection line refers to the line segment formed by the intersection of the axes representing the depths of the existing hole and the pre-drilled hole.

[0070] If the line of intersection of the axes on the side to be drilled is shorter, then a through hole is drilled; If the axis intersection line on one side of the pre-drilled hole is longer, then according to the set through-hole threshold, if the axis intersection line on the pre-drilled hole side is less than or equal to the through-hole threshold, a through hole is drilled; if it is greater than the through-hole threshold, a blind hole is drilled. For example... Figure 8 As shown, the 16mm hole in the handpiece is an open hole.

[0071] In addition to the methods mentioned above, through-hole / blind-hole determination can also be achieved through geometric Boolean operations.

[0072] The one-click drilling method for parts within a unit proposed in this embodiment requires manual operation for the input of the digital model and the selection of nodes. The subsequent digital model conversion and recognition, digital model hole analysis and operation are all automatically executed by the software system, realizing a true "one-click drilling" in vehicle body design. This makes the hole drilling and verification operation more standardized, the overall operation more efficient and accurate, reduces the professional requirements for drilling parts within a unit, and saves time and personnel costs.

[0073] Example 2: This embodiment proposes a one-click drilling system for parts within a unit, such as... Figure 9 As shown, it includes: Recognition and Conversion Module: The software recognizes the input digital model, converts the digital model into point cloud form, performs preprocessing on the point cloud data including denoising and enhancement, and obtains existing hole information within the point cloud range; Database module: This module establishes a hole database, which stores hole rules, including part type, hole group arrangement, common hole diameters and spacing, hole mating relationships, priority rules, precision standards, and technical specifications. The part types include perforated parts and mating parts. Pre-drilling module: Based on the existing hole information and the information stored in the hole database, pre-drilling is performed on the hand parts of the perforated parts; Verification Module: Verifies the pre-drilled hole. If the verification fails, it returns to the pre-drilled hole module; if the verification passes, it enters the hole-drilling module. The verification includes: using the center point of the pre-drilled hole as the center point of the curved surface, and using the normal direction of the intersection surface of the holed part and the counterpart as the axis direction to create a central axis, generating an axial surface with the same radius as the pre-drilled hole. Selecting four lines evenly distributed in the same direction as the central axis on the axial surface, with each pair of lines forming an angle of 90 degrees with the axis, and checking whether the interference amount between the four lines and the counterpart is the same. If they are the same, the verification passes; if they are not the same, the verification fails. Hole-opening module: Determines whether the hole in the device is a through hole or a blind hole, and performs the hole-opening operation.

[0074] The enhancement of point cloud data in the recognition and transformation module includes multi-data fusion, which generates point clouds of different dimensions and precision for the same model from point cloud data, and then merges the point clouds generated multiple times.

[0075] The identification and conversion module includes: The computational unit is used to calculate local normal vectors and geometric features: it constructs a neighborhood for point cloud data and uses principal component analysis (PCA) to calculate the normal vector and curvature of each point. Candidate point set unit: Based on region clustering of normal differences, the candidate point set of the hole is extracted: Using region growing clustering, the point set of the normal abrupt region is automatically extracted as the candidate point set of the hole boundary; Model unit: Build a 3D hole model and set engineering constraints: Construct a 3D hole model based on the candidate point set of the hole boundary, and set upper and lower limits of radius in combination with the hole diameter range of the hole database; Fitting unit for GNC-TLS non-convex robust circle fitting: By optimizing GNC and global least squares TLS through generalized non-convexity optimization, the weight of outliers is automatically reduced by progressively increasing non-convexity, thus achieving deterministic fitting without random sampling. Output unit, output hole features: output hole center coordinates, radius, and direction vector, which serve as the basis for subsequent hole design of hand parts.

[0076] Another approach involves an identification and conversion module that includes: acquiring existing hole information by obtaining information about circles within the point cloud range through a random sample consensus algorithm. This includes fitting a 3D circle using the random sample consensus algorithm, inputting the direction vector value of the circle, and constraining the radius range based on common data to obtain information about circles within this radius range in the point cloud; identifying the circle as an existing hole as the basis for subsequent hole design of the component.

[0077] The hole-opening module includes a judgment unit. In the judgment unit, the axis intersection line of the existing hole and the pre-opened hole is selected. If the axis intersection line of the pre-opened hole is shorter, a through hole is opened. If the axis intersection line of the pre-opened hole is longer, a through hole threshold is set. If the axis intersection line of the pre-opened hole is less than or equal to the through hole threshold, a through hole is opened. If it is greater than the through hole threshold, a blind hole is opened.

[0078] The one-click drilling system for parts within a unit proposed in this embodiment can achieve the one-click drilling method for parts within a unit described in Embodiment 1, and has the same technical effect as Embodiment 1.

[0079] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A one-click drilling method for parts within a unit, characterized in that, include: S1. The software identifies the input digital model, converts the digital model into a point cloud, performs preprocessing on the point cloud data including denoising and enhancement, and obtains existing hole information within the point cloud range. S2. Establish a hole database, which stores hole rules, including part type, hole group arrangement, common hole diameter and hole spacing, hole mating relationship, priority rules, accuracy standards, and technical specifications; the part type includes hole-bearing parts and mating parts. S3. Based on the existing hole information and the information stored in the hole database, perform pre-hole processing on the handpiece of the hole-filled part. S4. Verify the pre-drilled hole. If the verification fails, return to step S3. If the verification passes, proceed to the next step. The verification method includes: taking the center point of the pre-drilled hole as the center point of the curved surface, taking the normal direction of the intersection surface of the hole part and the counterpart part as the axis direction to generate the central axis, generating an axial surface with the same radius as the pre-drilled hole, selecting four lines evenly distributed in the same direction as the central axis on the axial surface, with each pair of the four lines forming an angle of 90 degrees with the axis, and checking whether the interference amount between the four lines and the counterpart part is the same. If they are the same, the verification passes; if they are not the same, the verification fails. S5. Determine whether the opening of the opposing component is a through hole or a blind hole, and perform the opening operation.

2. The one-click drilling method for parts within a unit according to claim 1, characterized in that, The enhancement of point cloud data in step S1 includes multi-data fusion, which generates multi-dimensional point clouds of different precision for the same model using point cloud data, and then merges the point clouds generated multiple times.

3. The one-click drilling method for parts within a unit according to claim 1, characterized in that, The acquisition of existing hole information in step S1 employs a method combining normal vector analysis, region clustering, and non-convex robust circle fitting, including: Local normal vector and geometric feature calculation: Construct a neighborhood for the point cloud data, and use principal component analysis (PCA) to calculate the normal vector and curvature of each point; Based on region clustering of normal differences, candidate point sets for holes are extracted: using region growing clustering, the point sets of regions with abrupt changes in normal are automatically extracted as candidate point sets for hole boundaries; Establish a 3D hole model and set engineering constraints: Construct a 3D hole model based on the candidate point set of the hole boundary, and set upper and lower limits of the radius in combination with the hole diameter range of the hole database; GNC-TLS Non-convex Robust Circle Fitting: By optimizing GNC and global least squares TLS through generalized non-convexity optimization, the weights of outliers are automatically reduced by progressively increasing non-convexity, thus achieving deterministic fitting without random sampling. Output hole features: Output hole center coordinates, radius, and direction vector, which serve as the basis for subsequent hole design of hand parts.

4. The one-click drilling method for parts within a unit according to claim 1, characterized in that, In step S1, the acquisition of existing hole information is achieved by obtaining information about circles within the point cloud range through a random sample consensus algorithm. This includes fitting a 3D circle using the random sample consensus algorithm, inputting the direction vector value of the circle, and constraining the radius range according to common data to obtain information about circles within this radius range in the point cloud. The circle is then identified as an existing hole, serving as the basis for subsequent hole design of the handpiece.

5. The one-click drilling method for parts within a unit according to claim 1, characterized in that, In step S5, the methods for determining whether an opening is a through hole or a blind hole include: Select the axis intersection line of the existing hole and the pre-drilled hole. If the axis intersection line of the pre-drilled hole side is shorter, then open the hole. If the axis intersection line of the pre-drilled hole side is longer, set the through hole threshold. If the axis intersection line of the pre-drilled hole side is less than or equal to the through hole threshold, then open the hole. If it is greater than the through hole threshold, then open the blind hole.

6. A one-click drilling system for parts within a unit, characterized in that, include: Recognition and Conversion Module: The software recognizes the input digital model, converts the digital model into point cloud form, performs preprocessing on the point cloud data including denoising and enhancement, and obtains existing hole information within the point cloud range; Database module: This module establishes a hole database, which stores hole rules, including part type, hole group arrangement, common hole diameters and spacing, hole mating relationships, priority rules, precision standards, and technical specifications. The part types include perforated parts and mating parts. Pre-drilling module: Based on the existing hole information and the information stored in the hole database, pre-drilling is performed on the hand parts of the perforated parts; Verification Module: Verifies the pre-drilled hole. If the verification fails, it returns to the pre-drilled hole module; if the verification passes, it enters the hole-drilling module. The verification includes: using the center point of the pre-drilled hole as the center point of the curved surface, and using the normal direction of the intersection surface of the holed part and the counterpart as the axis direction to create a central axis, generating an axial surface with the same radius as the pre-drilled hole. Selecting four lines evenly distributed in the same direction as the central axis on the axial surface, with each pair of lines forming an angle of 90 degrees with the axis, and checking whether the interference amount between the four lines and the counterpart is the same. If they are the same, the verification passes; if they are not the same, the verification fails. Hole-opening module: Determines whether the hole in the device is a through hole or a blind hole, and performs the hole-opening operation.

7. The one-click drilling system for parts within a unit according to claim 6, characterized in that, The enhancement of point cloud data in the recognition and transformation module includes multi-data fusion, which generates multi-dimensional point clouds of different precision for the same model from point cloud data, and then merges the point clouds generated multiple times.

8. The one-click drilling system for parts within a unit according to claim 6, characterized in that, The identification and conversion module includes: The computational unit is used to calculate local normal vectors and geometric features: it constructs a neighborhood for point cloud data and uses principal component analysis (PCA) to calculate the normal vector and curvature of each point. Candidate point set unit: Based on region clustering of normal differences, the candidate point set of the hole is extracted: Using region growing clustering, the point set of the normal abrupt region is automatically extracted as the candidate point set of the hole boundary; Model unit: Build a 3D hole model and set engineering constraints: Construct a 3D hole model based on the candidate point set of the hole boundary, and set upper and lower limits of radius in combination with the hole diameter range of the hole database; Fitting unit for GNC-TLS non-convex robust circle fitting: By optimizing GNC and global least squares TLS through generalized non-convexity optimization, the weight of outliers is automatically reduced by progressively increasing non-convexity, thus achieving deterministic fitting without random sampling. Output unit, output hole features: output hole center coordinates, radius, and direction vector, which serve as the basis for subsequent hole design of hand parts.

9. The one-click drilling system for parts within a unit according to claim 6, characterized in that, The identification and conversion module includes: acquiring existing hole information, which involves obtaining information about circles within the point cloud range through a random sample consensus algorithm, including fitting a 3D circle using the random sample consensus algorithm, inputting the direction vector value of the circle, and constraining the radius range according to common data to obtain information about circles within this radius range in the point cloud; identifying the circle as an existing hole, which serves as the basis for subsequent hole design of the handpiece.

10. The one-click drilling system for parts within a unit according to claim 6, characterized in that, The hole-opening module includes a judgment unit, in which the intersection line of the axis of the existing hole and the pre-opened hole is selected. If the intersection line of the axis of the pre-opened hole is shorter, the hole is opened. If the axis intersection line of the pre-drilled hole is longer, a through hole threshold is set. If the axis intersection line of the pre-drilled hole is less than or equal to the through hole threshold, a through hole is opened; if it is greater than the through hole threshold, a blind hole is opened.

Images