Method and system for detecting relative position of part under three-axis movement and medium
By using a three-axis motion-based method for detecting the relative position of parts, and optimizing the detection path using 3D point cloud data and feature recognition, the problems of insufficient accuracy and low efficiency in detecting the relative position of parts are solved, achieving high-precision and high-efficiency detection results.
Patent Information
- Application Number
- CN202511430974.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-09
- Publication Date
- 2025-12-19
- Estimated Expiration
- 2045-10-09
AI Technical Summary
Existing technologies suffer from insufficient accuracy and low efficiency in detecting the relative position of parts. Traditional detection methods rely on manual measurement or data acquisition in a single coordinate system, resulting in limited utilization of point cloud data, incomplete feature extraction, and complex coordinate mapping and position calculation processes. This leads to large measurement errors and lengthy detection steps, failing to meet the requirements for high-precision and high-efficiency detection.
A method for detecting the relative position of parts under three-axis movement is adopted. The detection path is planned by importing the three-dimensional model of the target part, and the three-axis movement control command is constructed to drive the three-coordinate device to scan and obtain three-dimensional point cloud data. Region segmentation and cluster analysis are performed to determine the feature space coordinates, a local coordinate system is constructed for position calculation, the feature relative position relationship diagram is traversed, an anomaly database is constructed, and deviation training is performed to optimize the detection path.
It improves the accuracy and efficiency of relative position detection of parts. Through 3D point cloud acquisition and feature recognition and deviation prediction optimization, it achieves higher accuracy and more efficient detection results.
Smart Images

Figure CN121170012A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of position detection, in particular to a part relative position detection method, system and medium under three-axis movement. BACKGROUND
[0002] In the process of machining and assembling, the relative position of the part needs to be detected. The traditional detection method relies on manual measurement or data acquisition in a single coordinate system, the utilization rate of point cloud data is limited, the feature extraction is not comprehensive, the coordinate mapping and position calculation process is complex, and measurement errors are easy to occur. The detection steps are lengthy, the data processing efficiency is low, and the high-precision and high-efficiency detection requirements cannot be met. SUMMARY
[0003] The present application provides a part relative position detection method, system and medium under three-axis movement, which is used to solve the technical problems of insufficient detection accuracy and low efficiency of part relative position detection in the prior art.
[0004] In view of the above problems, the present application provides a part relative position detection method, system and medium under three-axis movement.
[0005] The first aspect of the present application provides a part relative position detection method under three-axis movement, the method comprising: Importing a three-dimensional model of a target part to plan a detection path, constructing a three-axis movement control instruction to drive a three-coordinate device to move a detection head to scan the target part, obtaining three-dimensional point cloud data; based on the three-dimensional point cloud data, region segmentation and identification are performed on the three-dimensional model, clustering analysis is performed on the identification results, a plurality of feature point cloud clusters are obtained for global fitting, a plurality of feature space coordinates are determined; a part local coordinate system is constructed, the plurality of feature space coordinates are mapped to the part local coordinate system for position calculation, a plurality of feature relative position relationship graphs are obtained; deviation markers are added to the plurality of feature relative position relationship graphs, a plurality of out-of-tolerance items are obtained and bound to the target part for storage, and a position detection anomaly database is constructed; position deviation training is performed according to the position detection anomaly database, a position deviation prediction channel is constructed, and the detection path is feedback optimized.
[0006] The second aspect of the present application provides a part relative position detection system under three-axis movement, the system comprising: The scanning module is configured to import a three-dimensional model of a target part to plan a detection path, construct a three-axis movement control instruction to drive a three-coordinate device to move a detection head to scan the target part, and obtain three-dimensional point cloud data; the fitting module is configured to perform regional segmentation and identification on the three-dimensional model based on the three-dimensional point cloud data, perform cluster analysis on the identification result, obtain point cloud clusters of a plurality of feature points for global fitting, and determine a plurality of feature space coordinates; the calculation module is configured to construct a part local coordinate system, map the plurality of feature space coordinates to the part local coordinate system for position calculation, and obtain a plurality of feature relative position relationship graphs; the binding and storage module is configured to traverse the plurality of feature relative position relationship graphs to mark deviations, obtain a plurality of out-of-tolerance items, bind and store the target part, and construct a position detection anomaly database; and the feedback optimization module is configured to perform position deviation training based on the position detection anomaly database, construct a position deviation prediction channel, and perform feedback optimization on the detection path.
[0007] In a third aspect, the present application provides a computer readable storage medium storing a computer program, which, when executed by a processor, implements the part relative position detection method under three-axis movement.
[0008] The one or more technical solutions provided in the present application have at least the following technical effects or advantages: The scanning module is configured to import a three-dimensional model of a target part to plan a detection path, construct a three-axis movement control instruction to drive a three-coordinate device to move a detection head to scan the target part, and obtain three-dimensional point cloud data; the fitting module is configured to perform regional segmentation and identification on the three-dimensional model based on the three-dimensional point cloud data, perform cluster analysis on the identification result, obtain point cloud clusters of a plurality of feature points for global fitting, and determine a plurality of feature space coordinates; the calculation module is configured to construct a part local coordinate system, map the plurality of feature space coordinates to the part local coordinate system for position calculation, and obtain a plurality of feature relative position relationship graphs; the binding and storage module is configured to traverse the plurality of feature relative position relationship graphs to mark deviations, obtain a plurality of out-of-tolerance items, bind and store the target part, and construct a position detection anomaly database; and the feedback optimization module is configured to perform position deviation training based on the position detection anomaly database, construct a position deviation prediction channel, and perform feedback optimization on the detection path. The present application solves the technical problems of insufficient part relative position detection precision and low efficiency in the prior art, and achieves the technical effects of improving part relative position detection precision and detection efficiency through three-dimensional point cloud acquisition, feature identification, and deviation prediction optimization. BRIEF DESCRIPTION OF DRAWINGS
[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0010] Figure 1 This is a schematic flowchart of a method for detecting the relative position of a part under three-axis movement provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a part relative position detection system under three-axis movement provided in an embodiment of this application.
[0011] Figure labeling: Scanning module 11, Fitting module 12, Calculation module 13, Binding and storage module 14, Feedback optimization module 15. Detailed Implementation
[0012] This application provides a method, system, and medium for detecting the relative position of parts under three-axis movement. It addresses the technical problems of insufficient accuracy and low efficiency in the detection of the relative position of parts in the prior art. By optimizing the acquisition of three-dimensional point clouds, feature recognition, and deviation prediction, it achieves the technical effect of improving the accuracy and efficiency of the detection of the relative position of parts.
[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.
[0014] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.
[0015] Example 1, as Figure 1 As shown, this application provides a method for detecting the relative position of a part under three-axis movement, the method comprising: Step S100: Import the 3D model of the target part, plan the detection path, construct a three-axis motion control command to drive the three-coordinate device to move the probe head to scan the target part and obtain 3D point cloud data.
[0016] In this embodiment, pre-prepared 3D model data is first imported. This 3D model is generated and stored in the form of a CAD file and contains the geometric feature information and dimensional data of the target part. Then, a pre-planned detection path is invoked based on this 3D model. This detection path includes a scanning trajectory covering the key feature areas of the part. Subsequently, a CNC interpolation algorithm is used to convert the pre-planned detection path into three-axis motion parameters, forming three-axis movement control commands. These three-axis movement control commands contain displacement data and velocity information in the X, Y, and Z directions.
[0017] Then, the three-axis motion control command drives the three-coordinate device to move the probe head to scan the target part. The three-coordinate device moves the probe head to scan the surface of the part point by point at a fixed sampling frequency, collects spatial coordinate data in real time, and records and organizes the scanning results to finally obtain three-dimensional point cloud data.
[0018] Step S200: Based on the three-dimensional point cloud data, perform region segmentation and recognition on the three-dimensional model, perform cluster analysis on the recognition results, obtain point cloud clusters of multiple feature points, perform global fitting, and determine multiple feature space coordinates.
[0019] In this embodiment, when performing region segmentation and recognition on a 3D model based on 3D point cloud data, the 3D point cloud data is first mapped onto the 3D model to calculate normal vectors, constructing an initial point cloud normal vector field. An initial reference direction is then defined based on the initial point cloud normal vector field to segment the 3D model. Next, unsegmented point clouds are randomly selected as seed points based on the segmentation results. Neighboring points of the seed points are traversed and matched with the initial point cloud normal vector field. The normal vectors of the neighboring points are extracted, and the angle between the neighboring point normal vector and the seed point normal vector is calculated. When the angle is less than a preset angle threshold, the neighboring points are merged into the seed point, and the initial point cloud normal vector field is updated to obtain the point cloud normal vector field. Subsequently, based on the point cloud normal vector field and the segmentation results, the 3D model undergoes repeated region growth and merging until all unsegmented point clouds are segmented. Finally, multiple segmented regions are determined and geometric recognition is performed to obtain the recognition result.
[0020] Next, cluster analysis is performed on the recognition results to obtain point cloud clusters of multiple feature points for global fitting. In this process, the recognition results are first clustered using Euclidean distance to obtain point cloud clusters of multiple feature points. These clusters are then geometrically decomposed into aperture feature point cloud clusters, planar edge feature point cloud clusters, and curved surface feature point cloud clusters. Subsequently, for each type of point cloud cluster, cylindrical surface fitting, edge line fitting, and curved surface fitting are performed to obtain the corresponding aperture geometric feature parameters, planar edge geometric feature parameters, and curved surface geometric feature parameters. Finally, a global coordinate system is constructed based on the 3D model, and the aforementioned geometric feature parameters are transformed into this global coordinate system for identification, ultimately determining the coordinates of multiple feature spaces.
[0021] Further, the method provided by the application embodiment further comprises the following steps: mapping the three-dimensional point cloud data into the three-dimensional model to perform normal vector calculation, constructing an initial point cloud normal vector field, the initial point cloud normal vector field having a corresponding relationship with the three-dimensional point cloud data; segmenting the three-dimensional model according to an initial reference direction determined based on the initial point cloud normal vector field, randomly selecting an unsegmented point cloud as a seed point according to the segmentation result; matching a neighborhood point of the seed point with the initial point cloud normal vector field, extracting a neighborhood point normal vector, and calculating an included angle data between the neighborhood point normal vector and the seed point normal vector; when the included angle data is less than a preset angle threshold, merging the neighborhood point into the seed point, updating the initial point cloud normal vector field according to the neighborhood point normal vector and the seed point normal vector, and obtaining a point cloud normal vector field; and performing region growing and merging on the three-dimensional model based on the point cloud normal vector field and the segmentation result until all unsegmented point clouds are segmented, determining a plurality of segmentation regions for geometric recognition, and obtaining the recognition result.
[0022] In the application embodiment, when the three-dimensional point cloud data is mapped into the three-dimensional model to perform normal vector calculation, each point in the three-dimensional point cloud data is a point having a three-dimensional space coordinate, a neighborhood point set of each point is obtained through fixed-radius neighborhood search, a plane normal vector of the neighborhood point set is obtained through plane fitting of the neighborhood point set by using a least square plane, and an initial point cloud normal vector field is obtained through point-by-point calculation, wherein the initial point cloud normal vector field has a one-to-one corresponding relationship with each point in the three-dimensional point cloud data.
[0023] Next, when the three-dimensional model is segmented according to the initial reference direction determined based on the initial point cloud normal vector field, all normal vectors in the initial point cloud normal vector field are counted and directionally clustered, and the direction with the highest occurrence frequency is extracted as the initial reference direction. For each point in the three-dimensional point cloud data, an included angle between the normal vector of the point and the initial reference direction is calculated, the included angle is obtained through vector dot product divided by modulus product and taking inverse cosine, when the included angle is close to zero, it is considered that the normal vector of the point is consistent with the initial reference direction, and the point is divided into a segmented point set. When the included angle deviates greatly, it is considered that the normal vector of the point is inconsistent with the initial reference direction, and the point is divided into an unsegmented point set. In this way, the point cloud data is divided into the segmented point set and the unsegmented point set, and the segmented point set and the unsegmented point set are combined as a segmentation result. Then, an unsegmented point cloud is randomly selected as a seed point according to the segmentation result, and a point in the unsegmented point set is randomly selected as the seed point and as a starting point of subsequent region growing.
[0024] After the neighborhood points of the seed point are matched with the initial point cloud normal vector field, the neighborhood point normal vector is extracted by fixed k-neighbor search, the included angle data between the neighborhood point normal vector and the seed point normal vector is calculated, and the included angle data is obtained by vector dot product and inverse sine function. When the included angle data is less than a preset angle threshold, the neighborhood point is incorporated into the region where the seed point is located.
[0025] After the neighborhood point is incorporated, the initial point cloud normal vector field is updated according to the neighborhood point normal vector and the seed point normal vector, and the updating mode is to calculate the average normal vector and replace the normal vector of the corresponding point to obtain a new point cloud normal vector field.
[0026] Finally, based on the updated point cloud normal vector field, the region is repeatedly grown and incorporated according to the segmentation result, and the steps of neighborhood search, included angle data calculation, threshold judgment, neighborhood point incorporation and average normal vector updating are repeatedly executed until the unsegmented point set is empty. Through the process, a plurality of segmentation regions are finally obtained, and the geometric characteristics of the segmentation regions are identified by using the methods of plane fitting, cylindrical fitting or surface fitting, and a complete identification result is obtained.
[0027] Further, the method provided by the application embodiment includes clustering analysis of the identification result to obtain a plurality of point cloud clusters of feature points for global fitting to determine a plurality of feature space coordinates. The Euclidean distance is used to traverse the identification result for feature point clustering to obtain a plurality of point cloud clusters of feature points. The point cloud clusters of feature points are geometrically decomposed to obtain hole feature point cloud clusters, plane edge feature point cloud clusters and curved surface feature point cloud clusters. Cylinder surface fitting is performed based on the hole feature point cloud clusters to obtain hole geometric feature parameters. Edge line fitting is performed based on the plane edge feature point cloud clusters to obtain plane edge geometric feature parameters. Curved surface fitting is performed based on the curved surface feature point cloud clusters to obtain curved surface geometric feature parameters. A global coordinate system is constructed based on a three-dimensional model, and the hole geometric feature parameters, the plane edge geometric feature parameters and the curved surface geometric feature parameters are converted to the global coordinate system for identification to determine the plurality of feature space coordinates.
[0028] In the application embodiment, when the identification result is processed, the Euclidean distance is first used to traverse all points, the spatial distance relationship between points is judged by setting a distance threshold, when the distance between a certain unassigned point and a clustered point is less than the threshold, the point is divided into the same set, and the process is repeatedly executed until all points are assigned, thereby obtaining a plurality of point cloud clusters of feature points, and each point cloud cluster represents a point set that is close to each other in space and has similar features.
[0029] Next, the point cloud clusters of multiple feature points are geometrically decomposed, the point cloud clusters are fitted by using a random sample consensus method, the number of inliers in different geometric forms is calculated iteratively, the geometric form with the largest number of inliers is selected as the classification basis of the point cloud cluster, and finally the point cloud cluster is divided into a hole feature point cloud cluster, a plane edge feature point cloud cluster and a curved surface feature point cloud cluster.
[0030] Subsequently, for the hole feature point cloud cluster, a least square cylindrical surface fitting method is used to determine the hole geometric feature parameters, an axis vector is obtained through principal component analysis, an average radius is calculated on the axis vertical plane, and a center coordinate is obtained from the point cloud cluster centroid, and a cylindrical diameter is determined. Through this process, the hole geometric feature parameters including the center coordinate, the axis vector and the diameter are obtained.
[0031] For the plane edge feature point cloud cluster, a least square straight line fitting method is used to determine the plane edge geometric feature parameters. First, the first principal component is extracted as the direction vector of the edge straight line by using principal component analysis, the point cloud is projected onto the direction to form a one-dimensional coordinate sequence, the minimum value point and the maximum value point in the projection sequence are taken as the key corner point coordinates, and the straight line direction and the centroid point coordinates are combined to construct a straight line equation. Through this process, the plane edge geometric feature parameters are obtained, which include the edge equation parameters and the key corner point coordinates.
[0032] For the curved surface feature point cloud cluster, a B-spline surface least square fitting method is used to determine the curved surface geometric feature parameters. First, the node vector and control grid structure are set according to the point cloud range, the B-spline basis function matrix is established, the point cloud coordinates are substituted to form an overdetermined equation set, and the least square method is used to solve the control point coordinates. The principal curvature and Gaussian curvature of each point are calculated by fitting the curved surface. Through this process, the curved surface geometric feature parameters are obtained, which include the control point coordinates and the curvature information.
[0033] After obtaining the hole geometric feature parameters, the plane edge geometric feature parameters and the curved surface geometric feature parameters, a global coordinate system is constructed based on the three-dimensional model, and a homogeneous coordinate transformation method is used to uniformly process the parameters. The position parameters including the center coordinates, the key corner point coordinates and the control point coordinates are completed by the rotation matrix and the translation vector, and the direction parameters including the axis vector and the straight line direction are completed by the rotation matrix. Through this process, multiple feature space coordinates are finally obtained, and the identification is completed in the global coordinate system.
[0034] Step S300: constructing a part local coordinate system, mapping the multiple feature space coordinates to the part local coordinate system for position calculation, and obtaining multiple feature relative position relationship graphs.
[0035] In the embodiment of the present application, when constructing the part local coordinate system, first, the reference features are extracted in the three-dimensional model through reference analysis, and the reference features are mapped to the global coordinate system for identification to determine the reference plane, reference axis and reference point. Then, the reference plane is taken as the coordinate plane, the reference axis is taken as the coordinate axis, and the reference point is taken as the coordinate origin to complete the establishment of the part local coordinate system. Then, the coordinate transformation matrix is constructed, and the plurality of feature space coordinates are mapped to the part local coordinate system for position calculation to obtain the plurality of feature relative position relationship graphs.
[0036] Further, the method provided by the embodiment of the present application, in which the part local coordinate system is constructed, the plurality of feature space coordinates are mapped to the part local coordinate system for position calculation to obtain the plurality of feature relative position relationship graphs, further comprises: The reference features are obtained based on the three-dimensional model through reference analysis, the reference features are mapped to the global coordinate system for identification to determine the reference plane, reference axis and reference point, the reference plane is taken as the coordinate plane, the reference axis is taken as the coordinate axis, and the reference point is taken as the coordinate origin to construct the part local coordinate system, the coordinate transformation matrix is constructed, and the plurality of feature space coordinates are mapped to the part local coordinate system according to the coordinate transformation matrix for position calculation to obtain the plurality of feature relative position relationship graphs.
[0037] In the embodiment of the present application, when the reference features are obtained based on the three-dimensional model through reference analysis, first, the least square plane fitting method is used to process the three-dimensional point cloud data, the plane equation is calculated using the three-dimensional point cloud data to obtain the normal vector, and the reference plane is determined. Then, the least square line fitting method is used to process the three-dimensional point cloud data, the line equation is calculated using the three-dimensional point cloud data to obtain the direction vector, and the reference axis is determined. Then, the centroid calculation method is used to process the three-dimensional point cloud data to obtain the three-dimensional coordinate position, and the reference point is determined. The reference plane, reference axis and reference point are mapped to the global coordinate system for identification to obtain the parameter information of the reference features in the global coordinate system.
[0038] When the reference plane is taken as the coordinate plane, the reference axis is taken as the coordinate axis, and the reference point is taken as the coordinate origin to construct the part local coordinate system, the direction cosine method is used to establish the local three-dimensional orthogonal basis. Specifically, first, the local x-axis direction is aligned with the reference axis direction, then the local z-axis direction is aligned with the normal vector of the reference plane, and then the local y-axis direction is obtained through vector cross product to complete the three-axis definition of the part local coordinate system, and the reference point is taken as the coordinate origin to form the part local coordinate system.
[0039] Subsequently, when constructing the coordinate transformation matrix, firstly, the coordinate transformation analysis is performed based on the coordinate data of the reference feature in the global coordinate system to obtain the rotation matrix parameters and the translation vector parameters, and the direction cosine and the coordinate offset are calculated respectively based on the rotation matrix parameters and the translation vector parameters to form the rotation transformation matrix and the translation transformation vector. Then, the initial coordinate transformation matrix is generated by combining the rotation transformation matrix and the translation transformation vector, and the correctness of the initial coordinate transformation matrix is verified to obtain the coordinate transformation matrix.
[0040] After the construction of the part local coordinate system and the obtaining of the coordinate transformation matrix, the coordinate transformation matrix is used to map the multiple feature space coordinates from the global coordinate system to the part local coordinate system. For the position type feature space coordinates, the global coordinates are subtracted by the translation transformation vector and left multiplied by the transpose of the rotation transformation matrix to obtain the coordinate values in the part local coordinate system. For the direction type feature space coordinates, the global direction vector is left multiplied by the transpose of the rotation transformation matrix to obtain the direction parameters in the part local coordinate system. For the scalar type feature space coordinates, the coordinates remain unchanged, so that the complete feature geometric parameters in the part local coordinate system are obtained. After obtaining the mapped feature space coordinates, the position calculation is performed based on the part local coordinate system. The Euclidean distance formula is used to calculate the distance between the feature space coordinates, the direction vector dot product formula is used to calculate the included angle between the feature space coordinates, the parallel criterion and the orthogonal criterion are used to calculate the parallelism and the perpendicularity between the feature space coordinates respectively, and the coaxial criterion is used to calculate the coaxiality between the feature space coordinates, so that the geometric relationship between the multiple feature space coordinates is obtained.
[0041] After the position calculation is completed, the mapping results of the hole geometric feature parameters, the plane edge geometric feature parameters and the curved surface geometric feature parameters in the part local coordinate system are associated with the calculated geometric relationship, and are organized and visualized as multiple feature relative position relationship graphs. The multiple feature relative position relationship graphs take the feature space coordinates as nodes and take the relative distance, the relative included angle, the coaxiality, the parallelism and the perpendicularity and the like as edge attributes, so as to intuitively express the spatial distribution and the geometric constraint of the multiple features in the part local coordinate system.
[0042] Further, the method provided by the application embodiment further comprises: The coordinate data of the reference feature in the global coordinate system is extracted for coordinate transformation analysis to obtain coordinate transformation requirement parameters, the coordinate transformation requirement parameters including rotation matrix parameters and translation vector parameters; the coordinate axes of the local coordinate system are mapped to the global coordinate system according to the rotation matrix parameters to calculate direction cosines and construct a rotation transformation matrix; the coordinate origin of the local coordinate system is mapped to the global coordinate system according to the translation vector parameters to calculate an offset and construct a translation transformation vector; the rotation transformation matrix and the translation transformation vector are combined to construct a coordinate transformation initial matrix, and the coordinate transformation initial matrix is verified, and when the verification is passed, the coordinate transformation matrix is constructed.
[0043] In the embodiments of the present application, first, the coordinate data of the reference plane, the reference axis and the reference point in the global coordinate system is extracted, and the reference plane normal vector, the reference axis direction vector and the reference point space coordinate are analyzed. The normal vector components of the reference plane in the global coordinate system are calculated by vector algebra method, the direction vector components of the reference axis in the global coordinate system are calculated, and the three-dimensional coordinate position of the reference point in the global coordinate system is read. Through the process, the coordinate transformation requirement parameters describing the mapping relationship between the local coordinate system and the global coordinate system are obtained, and the coordinate transformation requirement parameters include rotation matrix parameters and translation vector parameters.
[0044] Next, the coordinate axes of the local coordinate system are mapped to the global coordinate system according to the rotation matrix parameters. The direction cosines are calculated by using the direction cosine method, and the direction cosines are used to describe the direction relationship between the coordinate axes of the local coordinate system and the three coordinate axes of the global coordinate system. Taking the X-axis of the local coordinate system as an example, the included angle cosine value is obtained by calculating the dot product of the local X-axis and the global X-axis and dividing by the modulus, and the dot product operations with the global Y-axis and the global Z-axis are repeated to obtain the three direction cosines of the local X-axis in the global coordinate system, and the three cosine values form the first column of the rotation transformation matrix. Similarly, the above calculation process is repeated for the Y-axis and the Z-axis of the local coordinate system to obtain the second column and the third column. Through the process, the rotation transformation matrix satisfying the orthogonal normalization condition is finally obtained.
[0045] Subsequently, the origin of the local coordinate system is mapped to the global coordinate system according to the translation vector parameters. By using the coordinate difference method, taking the three-dimensional coordinates of the reference point in the global coordinate system as a reference, the three components of the three-dimensional coordinates are extracted as the components of the translation vector in the X, Y and Z directions, respectively, to form the translation transformation vector.
[0046] Then, the rotation transformation matrix and the translation transformation vector are combined to form a 4×4 coordinate transformation initial matrix by using the homogeneous coordinate method. The rotation transformation matrix is filled into the upper left 3×3 sub-block of the 4×4 matrix, the translation transformation vector is filled into the upper right 3×1 sub-block, and the last row of the matrix is supplemented with 0, 0, 0 and 1 to form a homogeneous coordinate transformation initial matrix capable of processing position and direction simultaneously.
[0047] After the construction of the coordinate transformation initial matrix is completed, the coordinate transformation initial matrix is verified. In this process, first, a plurality of feature points are randomly selected as verification points based on the part local coordinate system, the verification points are mapped to the global coordinate system to obtain a plurality of initial coordinates, and the plurality of initial coordinates are calculated by coordinate transformation using the coordinate transformation initial matrix to obtain a plurality of coordinate values, which are compared with theoretical coordinate values for verification. The error distribution parameters are determined through error analysis, and the coordinate transformation initial matrix is finally iteratively corrected according to the error distribution parameters, so as to construct the coordinate transformation matrix.
[0048] Further, the method provided by the application embodiment includes the following steps when the coordinate transformation initial matrix is verified and the verification is passed: a plurality of feature points are randomly selected as a plurality of verification points based on the local coordinate system; the plurality of verification points are mapped to the global coordinate system to obtain a plurality of initial coordinates; the plurality of initial coordinates are calculated by coordinate transformation according to the coordinate transformation initial matrix to obtain a plurality of coordinate values; the plurality of coordinate values are compared with theoretical coordinate values for verification, and error distribution parameters are determined according to the verification result through error analysis; and the coordinate transformation initial matrix is iteratively corrected according to the error distribution parameters to construct the coordinate transformation matrix.
[0049] In the application embodiment, first, a plurality of feature points are randomly selected as a plurality of verification points from the part local coordinate system by using a uniform random sampling method, and the coordinates of these verification points in the part local coordinate system are recorded one by one, which are the theoretical coordinate values.
[0050] Next, when the plurality of verification points are mapped to the global coordinate system, a model mapping method is used to map the coordinates of the verification points in the part local coordinate system to the global coordinate system in proportion and direction through the corresponding relationship between the local coordinate system and the global coordinate system established in the three-dimensional model. Through this process, a plurality of initial coordinates corresponding to the plurality of verification points are obtained.
[0051] Subsequently, when the plurality of initial coordinates are calculated by coordinate transformation according to the coordinate transformation initial matrix, a homogeneous coordinate matrix multiplication method is used to expand each initial coordinate into a homogeneous coordinate form and perform matrix operation with the coordinate transformation initial matrix. Through this process, a plurality of coordinate values in the part local coordinate system are obtained.
[0052] After obtaining the plurality of coordinate values, the coordinate values are compared with theoretical coordinate values for verification. In the comparison process, a residual analysis method is used to calculate a residual vector of each verification point, the residual vector being obtained by subtracting the theoretical coordinate values from the coordinate values. By performing statistics on the residuals of all verification points, root mean square error, mean absolute error, and maximum absolute error are obtained, and the residual distribution is fitted to extract error distribution parameters, including a mean deviation vector and a residual covariance matrix.
[0053] Next, the initial coordinate transformation matrix is iteratively corrected according to the error distribution parameters. In the correction process, a least squares rigid registration method is used to update the rotation transformation matrix and the translation transformation vector. Specifically, the theoretical coordinate values and the coordinate values obtained by the initial coordinate transformation matrix are taken as corresponding point pairs, the centroids of the two groups of points are calculated and are subjected to a centering process, a cross-covariance matrix is constructed and is subjected to singular value decomposition, an updated rotation transformation matrix is extracted to ensure that it satisfies the conditions of orthogonality and determinant of 1, and then an updated translation vector is obtained according to the centroid offset. Subsequently, the updated rotation transformation matrix and the translation vector are combined to replace and update the initial coordinate transformation matrix, until the root mean square error converges to a preset threshold or the number of iterations reaches an upper limit. After the correction is completed, the final obtained matrix is the coordinate transformation matrix.
[0054] Step S400: traversing the plurality of feature relative position relationship graphs to mark deviations, obtaining a plurality of out-of-tolerance items and binding the target part for storage, and constructing a position detection anomaly database.
[0055] In the embodiments of the present application, in the position detection process, first, the feature relative position relationship graphs are traversed to analyze deviations, and the feature relative position relationships that exceed the tolerance range are extracted as out-of-tolerance items. Then, the target part is assigned a part identification code in combination with the multi-dimensional information of the part, and the corresponding three-dimensional point cloud data segments are extracted from the feature relative position relationship graphs of the plurality of out-of-tolerance items. Next, the plurality of out-of-tolerance items and the part identification code are associated and bound according to the plurality of three-dimensional point cloud data segments to form a binding data list. Finally, the binding data list is stored in a null structured database, thereby constructing a position detection anomaly database.
[0056] Further, in the method provided by the embodiments of the present application, traversing the plurality of feature relative position relationship graphs to mark deviations, obtaining a plurality of out-of-tolerance items and binding the target part for storage, and constructing a position detection anomaly database further comprises: The deviation analysis is performed on the plurality of feature relative position relationship graphs, and a feature relative position relationship exceeding a tolerance range is extracted as an out-of-tolerance item to determine a plurality of out-of-tolerance items; multi-dimensional analysis is performed on the target part, and a part identification code is assigned to the target part according to part multi-dimensional information; a plurality of three-dimensional point cloud data segments are extracted from the plurality of feature relative position relationship graphs based on the plurality of out-of-tolerance items; the plurality of out-of-tolerance items and the part identification code are associated and bound according to the plurality of three-dimensional point cloud data segments to construct a bound data list; and an empty structured database is constructed, the bound data list is stored in the database, and the position detection anomaly database is constructed.
[0057] In the embodiments of the present application, when performing deviation analysis on the plurality of feature relative position relationship graphs, the relative distance, relative angle, coaxiality, parallelism and perpendicularity of each feature pair are compared with the theoretical design parameters one by one, and a threshold comparison method is used to determine whether the deviation value is within the tolerance band. When the deviation value is greater than the upper limit of the tolerance band or less than the lower limit of the tolerance band, the feature relative position relationship is marked as an out-of-tolerance item, and finally a plurality of out-of-tolerance items are obtained.
[0058] When performing multi-dimensional analysis on the target part to assign a part identification code, multi-dimensional information such as geometric information, material information and processing information of the part is extracted, and a part identification code is generated by combining these information using a unique coding method, and the part identification code is assigned to the target part.
[0059] Next, when extracting a plurality of three-dimensional point cloud data segments from the plurality of feature relative position relationship graphs based on the plurality of out-of-tolerance items, the feature space coordinates corresponding to each out-of-tolerance item are taken as the center, a fixed search radius is set by using a neighborhood search method, neighborhood retrieval is performed on the three-dimensional point cloud data, and the local point cloud region related to the out-of-tolerance item is cropped and extracted, thereby obtaining a plurality of three-dimensional point cloud data segments containing deviation features.
[0060] Subsequently, the plurality of out-of-tolerance items and the part identification code are associated and bound according to the plurality of three-dimensional point cloud data segments, the part identification code is taken as an index, each out-of-tolerance item and the corresponding three-dimensional point cloud data segment are bound and stored, and a bound data list is formed. At the same time, the detection path parameters and the detection environment parameters are written in the bound data list, wherein the detection path parameters come from the control instruction data when the detection head performs a scanning task, and include the coordinate sequence of the scanning track point, the motion speed and the acceleration; the detection environment parameters come from the acquisition results of the environmental sensors connected to the detection equipment, and include the temperature, the humidity and the vibration amplitude. Through this process, the bound data list containing the out-of-tolerance item, the out-of-tolerance type, the out-of-tolerance value, the part identification code, the three-dimensional point cloud data segment, the detection path parameter and the detection environment parameter is obtained.
[0061] Finally, when constructing the null structured database, the relational database design method is used to establish the table structure and field constraints, the binding data list is written into the database table, and the index of the field is established to ensure the query efficiency and data consistency. Through the process, the construction of the position detection exception database is completed.
[0062] Step S500: According to the position detection exception database, the position deviation training is performed, and the position deviation prediction channel is constructed to feedback and optimize the detection path.
[0063] In the embodiments of the present application, when the position deviation training is performed according to the position detection exception database, the historical out-of-tolerance data is first extracted from the position detection exception database, and the historical out-of-tolerance data includes historical out-of-tolerance types, historical out-of-tolerance values, historical detection path parameters and historical detection environment parameters. Then, the historical detection path parameters and the historical detection environment parameters are used as input features, and the historical out-of-tolerance types and the historical out-of-tolerance values are used as prediction targets for training, so as to construct a position deviation prediction channel. In a new detection task, the detection path is synchronized to the position deviation prediction channel for prediction to obtain a position deviation prediction result. Finally, the position deviation prediction result is fed back to the detection path for optimization and adjustment to obtain a detection optimization path, which is used to drive the three-coordinate device to move the detection head to detect the relative position of the target part.
[0064] Further, the method provided by the embodiments of the present application further comprises: Based on the position detection exception database, historical out-of-tolerance data is extracted, and the historical out-of-tolerance data includes historical out-of-tolerance types, historical out-of-tolerance values, historical detection path parameters and historical detection environment parameters. The historical detection path parameters and the historical detection environment parameters are used as input features, and the historical out-of-tolerance types and the historical out-of-tolerance values are used as prediction targets for training to construct a position deviation prediction channel. The detection path is synchronized to the position deviation prediction channel for prediction to construct a position deviation prediction result. The position deviation prediction result is fed back to the detection path for optimization and adjustment to obtain a detection optimization path, which is used to drive the three-coordinate device to move the detection head to detect the relative position of the target part.
[0065] In the embodiments of the present application, the historical out-of-tolerance data is first extracted from the position detection exception database, and the historical out-of-tolerance data includes historical out-of-tolerance types, historical out-of-tolerance values, historical detection path parameters and historical detection environment parameters. At the same time, the Z-score standardization method is used to standardize the numerical fields, and the timestamps are aligned to ensure that the historical detection path parameters and the historical detection environment parameters correspond to the historical out-of-tolerance types and the historical out-of-tolerance values in the same time scale.
[0066] Next, the historical probe path parameters and the historical probe environment parameters are taken as input features, the historical deviation type and the historical deviation value are taken as prediction targets, and a gradient boosting tree method is used for modeling. In the training process, the training set and the validation set are divided according to a fixed ratio, a deviation type classifier is first trained to output the historical deviation type, and then a deviation value regressor is trained to output the historical deviation value. Key hyperparameters such as learning rate, maximum depth and number of weak learners are selected through the performance of the validation set. After the training converges, the classifier and the regressor are encapsulated as a position deviation prediction channel through a unified interface.
[0067] Next, the to-be-executed probe path is synchronized to the position deviation prediction channel for prediction. The position deviation prediction channel receives the probe path parameters and the current probe environment parameters as inputs, adopts a batch forward inference method for calculation, and generates a position deviation prediction result. The prediction result includes the deviation type and the deviation value corresponding to the trajectory segment or the sampling point. For example, under the conditions of a scanning speed of 200 mm / s, a sampling frequency of 100 Hz and an environment temperature of 35 DEG C, a coaxiality deviation of 0.03 mm is predicted for a certain trajectory segment, and the output is labeled, thereby explicitly indicating the potential high-risk trajectory segment.
[0068] Finally, the position deviation prediction result is fed back to the probe path for optimization and adjustment. In this process, a Bayesian optimization method is adopted to reduce the predicted deviation value as the target, while combining the maximum speed, maximum acceleration and stroke boundary of the three-coordinate device and other constraint conditions. The optimization variables include the trajectory spacing, the scanning speed, the sampling frequency and the probe head attitude angle. The path parameters are adjusted iteratively to gradually reduce the predicted deviation until the optimization target converges or the maximum number of iterations is reached. The probe path obtained after optimization is used to drive the three-coordinate device to move the probe head to detect the relative position of the target part, thereby improving the detection accuracy and efficiency.
[0069] In the embodiments of the present application, as described above, the embodiments of the present application have at least the following technical effects: The application imports a three-dimensional model of a target part to plan a detection path, constructs a three-axis movement control instruction to drive a three-coordinate device to move a detection head to scan the target part, and obtains three-dimensional point cloud data; based on the three-dimensional point cloud data, the three-dimensional model is regionally segmented and identified, the identification result is clustered and analyzed, a plurality of feature point point cloud clusters are globally fitted to determine a plurality of feature space coordinates; a part local coordinate system is constructed, the plurality of feature space coordinates are mapped to the part local coordinate system for position calculation to obtain a plurality of feature relative position relationship graphs; deviation marks are marked by traversing the plurality of feature relative position relationship graphs, a plurality of out-of-tolerance items are bound and stored with the target part to construct a position detection anomaly database; position deviation training is performed according to the position detection anomaly database to construct a position deviation prediction channel to feed back and optimize the detection path. The application solves the technical problems of insufficient part relative position detection accuracy and low efficiency in the prior art, and achieves the technical effects of improving part relative position detection accuracy and detection efficiency through three-dimensional point cloud acquisition, feature identification and deviation prediction optimization.
[0070] In the second embodiment, based on the same inventive concept as the part relative position detection method under three-axis movement in the foregoing embodiments, as shown in the following table, the application provides a part relative position detection system under three-axis movement, and the system and method embodiments in the application are based on the same inventive concept. The system includes: Figure 2 The scanning module 11 is configured to import a three-dimensional model of a target part to plan a detection path, construct a three-axis movement control instruction to drive a three-coordinate device to move a detection head to scan the target part, and obtain three-dimensional point cloud data; the fitting module 12 is configured to perform regional segmentation and identification on the three-dimensional model based on the three-dimensional point cloud data, perform clustering analysis on the identification result, globally fit a plurality of feature point point cloud clusters, and determine a plurality of feature space coordinates; the calculation module 13 is configured to construct a part local coordinate system, map the plurality of feature space coordinates to the part local coordinate system for position calculation, and obtain a plurality of feature relative position relationship graphs; the binding and storage module 14 is configured to mark deviations by traversing the plurality of feature relative position relationship graphs, obtain a plurality of out-of-tolerance items, and bind and store the out-of-tolerance items with the target part to construct a position detection anomaly database; and the feedback and optimization module 15 is configured to perform position deviation training according to the position detection anomaly database, construct a position deviation prediction channel, and feed back and optimize the detection path.
[0071] Further, the system is further configured to implement the following functions: mapping the three-dimensional point cloud data into a three-dimensional model for normal vector calculation, constructing an initial point cloud normal vector field, the initial point cloud normal vector field having a corresponding relationship with the three-dimensional point cloud data; segmenting the three-dimensional model according to an initial reference direction demarcated by the initial point cloud normal vector field, and randomly selecting an unsegmented point cloud as a seed point according to a segmentation result; matching a neighborhood point of the seed point with the initial point cloud normal vector field, extracting a neighborhood point normal vector, and calculating an included angle data of the neighborhood point normal vector and a seed point normal vector; when the included angle data is less than a preset angle threshold, incorporating the neighborhood point into the seed point, updating the initial point cloud normal vector field according to the neighborhood point normal vector and the seed point normal vector, and obtaining a point cloud normal vector field; performing region growing and incorporation on the three-dimensional model based on the point cloud normal vector field and the segmentation result until all unsegmented point clouds are segmented, determining a plurality of segmentation regions for geometric recognition, and obtaining the recognition result.
[0072] Further, the system is further configured to implement the following functions: The system is further configured to implement the following functions: The system is further configured to implement the following functions:
[0073] The system is further configured to implement the following functions: The system is further configured to implement the following functions:
[0074] Further, the system is further configured to implement the following functions: The coordinate data of the reference feature in the global coordinate system is extracted for coordinate transformation analysis to obtain coordinate transformation requirement parameters, the coordinate transformation requirement parameters including rotation matrix parameters and translation vector parameters; the coordinate axes of the local coordinate system are mapped to the global coordinate system according to the rotation matrix parameters to calculate direction cosines and construct a rotation transformation matrix; the coordinate origin of the local coordinate system is mapped to the global coordinate system according to the translation vector parameters to calculate an offset and construct a translation transformation vector; the rotation transformation matrix and the translation transformation vector are combined to construct a coordinate transformation initial matrix, and the coordinate transformation initial matrix is verified, and when the verification is passed, the coordinate transformation matrix is constructed.
[0075] Further, the system is also used to implement the following functions: A plurality of feature points are randomly selected as a plurality of verification points based on the local coordinate system; the plurality of verification points are mapped to a global coordinate system to obtain a plurality of initial coordinates; the plurality of initial coordinates are subjected to coordinate conversion calculation according to the coordinate transformation initial matrix to obtain a plurality of coordinate values; the plurality of coordinate values are compared with theoretical coordinate values for verification, error analysis is performed according to the verification result, and error distribution parameters are determined; the coordinate transformation initial matrix is iteratively corrected according to the error distribution parameters to construct the coordinate transformation matrix.
[0076] Further, the system is also used to implement the following functions: Deviation analysis is performed on the plurality of feature relative position relationship graphs, feature relative position relationships exceeding a tolerance range are extracted as out-of-tolerance items, and a plurality of out-of-tolerance items are determined; multi-dimensional analysis is performed on the target part, and a part identification code is assigned to the target part according to part multi-dimensional information; a plurality of three-dimensional point cloud data segments are extracted based on the plurality of out-of-tolerance items and the plurality of feature relative position relationship graphs; the plurality of out-of-tolerance items and the part identification code are associated and bound according to the plurality of three-dimensional point cloud data segments to construct a bound data list; a null structured database is constructed, the bound data list is stored in the database, and the position detection abnormality database is constructed.
[0077] Further, the system is also used to implement the following functions: Extract historical over-limit data based on the position detection anomaly database, the historical over-limit data including historical over-limit type, historical over-limit value, historical detection path parameter, historical detection environment parameter; take the historical detection path parameter and the historical detection environment parameter as input features, take the historical over-limit type and the historical over-limit value as prediction targets for training, and construct a position deviation prediction channel; synchronize the detection path to the position deviation prediction channel for prediction, and construct a position deviation prediction result; feed back the position deviation prediction result to the detection path for optimization adjustment, and obtain a detection optimization path to drive a three-coordinate device to move a detection head to detect the relative position of the target part.
[0078] In the third embodiment, based on the method for detecting the relative position of a part under three-axis movement in the foregoing embodiments, the same inventive concept is provided, and the application also provides a computer-readable storage medium, which stores a computer program. The computer program, when executed, implements the steps of the method in any one of the first embodiment.
[0079] It should be noted that the above sequence of the embodiments of the application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.
[0080] The above only describes the preferred embodiments of the application and does not limit the application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application should be included in the protection scope of the application.
[0081] The specification and drawings are merely exemplary of the application, and any and all modifications, variations, combinations or equivalents that are within the scope of the application should be considered. Obviously, those skilled in the art can make various modifications and variations to the application without departing from the scope of the application. Thus, if these modifications and variations of the application belong to the scope of the application and its equivalents, the application is intended to include these modifications and variations.
Claims
1. A method for detecting the relative position of a part under three-axis movement, characterized in that, The method includes: Import the 3D model of the target part to plan the detection path, construct a three-axis motion control command to drive the three-coordinate device to move the probe head to scan the target part and obtain 3D point cloud data; Based on the three-dimensional point cloud data, the three-dimensional model is segmented and identified. The identification results are then subjected to cluster analysis to obtain point cloud clusters of multiple feature points. Global fitting is then performed to determine the spatial coordinates of multiple features. Construct a local coordinate system for the part, map the spatial coordinates of the multiple features to the local coordinate system for position calculation, and obtain a relative position relationship diagram of the multiple features. By traversing the relative position relationship diagrams of the multiple features, deviation marking is performed, and multiple out-of-tolerance items are obtained and bound to the target part for storage, thus constructing a position detection anomaly database; The location deviation is trained based on the location detection anomaly database, and a location deviation prediction channel is constructed to provide feedback optimization for the detection path.
2. The method for detecting the relative position of a part under three-axis movement as described in claim 1, characterized in that, The method for performing region segmentation and recognition of a 3D model based on the aforementioned 3D point cloud data includes: The three-dimensional point cloud data is mapped to a three-dimensional model to calculate the normal vectors, and an initial point cloud normal vector field is constructed. The initial point cloud normal vector field has a corresponding relationship with the three-dimensional point cloud data. The three-dimensional model is segmented according to the initial reference direction defined by the initial point cloud normal vector field, and the unsegmented point cloud is randomly selected as seed point based on the segmentation result. The neighboring points of the seed point are traversed and matched with the initial point cloud normal vector field to extract the normal vectors of the neighboring points and calculate the angle data between the normal vectors of the neighboring points and the normal vector of the seed point. When the included angle data is less than a preset angle threshold, the neighboring points are merged into the seed point, and the initial point cloud normal vector field is updated according to the normal vectors of the neighboring points and the normal vectors of the seed point to obtain the point cloud normal vector field. Based on the point cloud normal vector field and the segmentation results, the 3D model is repeatedly grown and incorporated into the region until all unsegmented point clouds are segmented. Multiple segmented regions are then identified for geometric recognition to obtain the recognition results.
3. The method for detecting the relative position of a part under three-axis movement as described in claim 1, characterized in that, The recognition results are subjected to cluster analysis to obtain point cloud clusters of multiple feature points. Global fitting is then performed to determine the coordinates of multiple feature spaces. The methods include: The Euclidean distance traversal recognition results are used to cluster feature points to obtain point cloud clusters of multiple feature points; Geometrically decompose point cloud clusters with multiple feature points to obtain point cloud clusters with hole features, point cloud clusters with planar edge features, and point cloud clusters with curved surface features. Based on the hole feature point cloud cluster, a cylindrical surface is fitted to obtain the hole geometric feature parameters; Edge line fitting is performed based on the planar edge feature point cloud cluster to obtain the planar edge geometric feature parameters; Surface fitting is performed based on the surface feature point cloud cluster to obtain surface geometric feature parameters; A global coordinate system is constructed based on the 3D model. The geometric feature parameters of the hole, the geometric feature parameters of the plane edge, and the geometric feature parameters of the curved surface are transformed into the global coordinate system for identification, thereby determining the coordinates of the multiple feature spaces.
4. The method for detecting the relative position of a part under three-axis movement as described in claim 3, characterized in that, Constructing a local coordinate system for the part, mapping the spatial coordinates of the multiple features to the local coordinate system for position calculation, and obtaining a relative positional relationship diagram of the multiple features, the method includes: Based on the 3D model, benchmark analysis is performed to obtain benchmark features. The benchmark features are then mapped to the global coordinate system for identification, thereby determining the benchmark plane, benchmark axis, and benchmark point. A local coordinate system for the part is constructed by using the reference plane as the coordinate plane, the reference axis as the coordinate axis, and the reference point as the origin. Construct a coordinate transformation matrix, and map the spatial coordinates of the multiple features to the local coordinate system of the part according to the coordinate transformation matrix to calculate the position, thereby obtaining a relative position relationship diagram of the multiple features.
5. The method for detecting the relative position of a part under three-axis movement as described in claim 4, characterized in that, The process of constructing the coordinate transformation matrix includes the following methods: The coordinate data of the reference feature in the global coordinate system are extracted and subjected to coordinate transformation analysis to obtain the coordinate transformation requirement parameters, which include rotation matrix parameters and translation vector parameters. The coordinate axes of the local coordinate system are mapped to the global coordinate system according to the rotation matrix parameters, and the direction cosines are calculated to construct the rotation transformation matrix. The local coordinate system origin is mapped to the global coordinate system according to the translation vector parameters to calculate the offset and construct the translation transformation vector. The rotation transformation matrix and the translation transformation vector are combined to construct an initial coordinate transformation matrix. The initial coordinate transformation matrix is verified. When the verification is successful, the coordinate transformation matrix is constructed.
6. The method for detecting the relative position of a part under three-axis movement as described in claim 5, characterized in that, Verify the initial coordinate transformation matrix. If the verification passes, construct the coordinate transformation matrix. The method includes: Multiple feature points are randomly selected as multiple verification points based on the local coordinate system; The multiple verification points are mapped to the global coordinate system to obtain multiple initial coordinates; The multiple initial coordinates are transformed and calculated according to the initial coordinate transformation matrix to obtain multiple coordinate values; The multiple coordinate values are compared and verified with the theoretical coordinate values. Based on the verification results, error analysis is performed to determine the error distribution parameters. The coordinate transformation matrix is constructed by iteratively correcting the initial coordinate transformation matrix according to the error distribution parameters.
7. The method for detecting the relative position of a part under three-axis movement as described in claim 1, characterized in that, The method involves iterating through the relative positional relationship diagrams of multiple features, marking deviations, obtaining multiple out-of-tolerance items, binding and storing them with the target part, and constructing a position detection anomaly database. The method includes: By traversing the multiple feature relative position relationship diagrams, deviation analysis is performed, and feature relative position relationships that exceed the tolerance range are extracted as out-of-tolerance items, thus identifying multiple out-of-tolerance items. Perform multidimensional analysis on the target part, and determine the part identification code to be assigned to the target part based on the multidimensional information of the part; Based on the multiple out-of-range terms, multiple 3D point cloud data fragments are extracted by traversing the multiple feature relative position relationship graphs; The multiple out-of-tolerance items and the part identification code are associated and bound according to the multiple three-dimensional point cloud data fragments to construct a bound data list; Construct a null-structured database, store the bound data list in the database, and build the location detection anomaly database.
8. The method for detecting the relative position of a part under three-axis movement as described in claim 1, characterized in that, The method includes training on position deviation based on the aforementioned position detection anomaly database, constructing a position deviation prediction channel to optimize the detection path, and the following steps: Historical out-of-tolerance data is extracted based on the location detection anomaly database. The historical out-of-tolerance data includes historical out-of-tolerance type, historical out-of-tolerance value, historical detection path parameters, and historical detection environment parameters. The historical detection path parameters and historical detection environment parameters are used as input features, and the historical deviation types and historical deviation values are used as prediction targets for training to construct a position deviation prediction channel. The detection path is synchronized to the position deviation prediction channel for prediction, and a position deviation prediction result is constructed. The position deviation prediction result is fed back to the detection path for optimization and adjustment. The optimized detection path drives the coordinate measuring machine to move the probe head to detect the relative position of the target part.
9. A system for detecting the relative position of a part under three-axis movement, characterized in that, The system is used to perform the part relative position detection method under three-axis movement as described in any one of claims 1-8, and the system includes: The scanning module is used to import the 3D model of the target part, plan the detection path, and construct three-axis motion control commands to drive the three-coordinate device to move the probe head to scan the target part and obtain 3D point cloud data. The fitting module is used to perform region segmentation and recognition on the 3D model based on the 3D point cloud data, perform cluster analysis on the recognition results, obtain point cloud clusters of multiple feature points for global fitting, and determine multiple feature space coordinates. The calculation module is used to construct a local coordinate system for the part, map the spatial coordinates of the multiple features to the local coordinate system for position calculation, and obtain a relative position relationship diagram of the multiple features. The binding and storage module is used to traverse the relative position relationship diagram of the multiple features, mark the deviation, obtain multiple out-of-tolerance items and bind and store them with the target part, and build a position detection anomaly database. The feedback optimization module is used to train the position deviation based on the position detection anomaly database and construct a position deviation prediction channel to perform feedback optimization on the detection path.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the method for detecting the relative position of a part under three-axis movement as described in any one of claims 1-8.
Citation Information
Patent Citations
Part digital twin geometric modeling and error evaluation method and system based on point cloud
CN116402866A
Three-dimensional point cloud segmentation method for production line part identification
CN118351127A
Intelligent manufacturing-oriented aviation CAD part three-dimensional detection method
CN118799365A
Multi-station foundation laser scanning point cloud registration method and system suitable for forest environment
CN119648758A
Smart Design and Error Validation System for 3D Modeling
KR102834265B1
Cited By
Digital supervision method and system for power construction project
CN121639807A
