An automatic positioning method and system for workpieces of a numerical control machine tool

Abstract

The present disclosure provides an automatic positioning method and system for workpieces of a numerical control machine, relating to the technical field of automatic positioning, which comprises: performing surface scanning on a target workpiece through a laser radar device to obtain point cloud data; extracting model feature points from a CAD model of the target workpiece; performing registration matching on the point cloud data and the model feature points through a registration algorithm to obtain a registration result; obtaining a transformation matrix according to the registration result, determining a first position and a first attitude of the target workpiece in a laser radar coordinate system according to the transformation matrix, and obtaining a first coordinate system result; obtaining a second position and a second attitude of the laser radar coordinate system in a numerical control machine coordinate system, obtaining a second coordinate system result, calculating a target position and a target attitude of the target workpiece in the numerical control machine coordinate system, obtaining a target coordinate system result, and driving the numerical control machine to position the target workpiece, thereby improving the automation level of the numerical control machine and achieving intelligent positioning effect.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of automatic positioning, in particular to an automatic positioning method and system for workpieces of a numerical control machine tool. BACKGROUND

[0002] At present, it is difficult to accurately position fine parts, small parts and complex parts through manual positioning. If the numerical control machine tool is not accurately positioned, the material will be scrapped and cannot be used. Therefore, a method is needed to accurately position the components in the numerical control machine tool, and then the machine tool performs subsequent operations on the components. SUMMARY

[0003] The present disclosure provides an automatic positioning method and system for workpieces of a numerical control machine tool to solve the technical problem that fine parts are difficult to accurately position through manual positioning.

[0004] According to a first aspect of the present disclosure, an automatic positioning method for workpieces of a numerical control machine tool is provided, comprising: performing surface scanning on a target workpiece by a laser radar device to obtain point cloud data; extracting model feature points from a CAD model of the target workpiece; performing registration matching on the point cloud data and the model feature points by a registration algorithm to obtain a registration result; obtaining a transformation matrix according to the registration result, and determining a first position and a first attitude of the target workpiece in a laser radar coordinate system according to the transformation matrix; obtaining a first coordinate system result according to the first position and the first attitude; obtaining a second position and a second attitude of the laser radar coordinate system in a numerical control machine tool coordinate system to obtain a second coordinate system result; calculating a target position and a target attitude of the target workpiece in the numerical control machine tool coordinate system according to the first coordinate system result and the second coordinate system result to obtain a target coordinate system result; and driving the numerical control machine tool to position the target workpiece according to the target coordinate system result.

[0005] According to a second aspect of the present disclosure, an automatic positioning system for workpieces of a numerical control machine tool is provided, comprising:

[0006] The first obtaining module obtains point cloud data by surface scanning of the target workpiece through the laser radar device; the second obtaining module extracts model feature points from a CAD model of the target workpiece; the third obtaining module obtains a registration result by registration and matching of the point cloud data and the model feature points through a registration algorithm; the fourth obtaining module obtains a transformation matrix according to the registration result, and determines a first position and a first attitude of the target workpiece in the laser radar coordinate system according to the transformation matrix; the fifth obtaining module obtains a first coordinate system result according to the first position and the first attitude; the sixth obtaining module obtains a second position and a second attitude of the laser radar coordinate system in a numerical control machine tool coordinate system, and obtains a second coordinate system result; the seventh obtaining module calculates a target position and a target attitude of the target workpiece in the numerical control machine tool coordinate system according to the first coordinate system result and the second coordinate system result, and obtains a target coordinate system result; and the first processing module drives the numerical control machine tool to position the target workpiece according to the target coordinate system result.

[0007] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0008] According to the automatic positioning method and system of the workpiece of the numerical control machine tool, the surface of the target workpiece is scanned through the laser radar device to obtain point cloud data; model feature points are extracted from a CAD model of the target workpiece; registration and matching of the point cloud data and the model feature points are performed through a registration algorithm to obtain a registration result; a transformation matrix is obtained according to the registration result, a first position and a first attitude of the target workpiece in the laser radar coordinate system are determined according to the transformation matrix, and a first coordinate system result is obtained; a second position and a second attitude of the laser radar coordinate system in a numerical control machine tool coordinate system are obtained to obtain a second coordinate system result, a target position and a target attitude of the target workpiece in the numerical control machine tool coordinate system are calculated, a target coordinate system result is obtained, and the numerical control machine tool is driven to position the target workpiece, thereby improving the automation level of the numerical control machine tool and realizing intelligent positioning effect.

[0009] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the present disclosure or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and other drawings can be obtained by those skilled in the art without creating any creative labor on the basis of the provided drawings.

[0011] Figure 1 A flowchart of an automatic positioning method of a workpiece of a numerical control machine tool provided by an embodiment of the present disclosure is shown in the figure;

[0012] Figure 2 A flowchart of extracting model feature points from a CAD model of the target workpiece in an automatic positioning method of a workpiece of a numerical control machine tool provided by an embodiment of the present disclosure is shown in the figure;

[0013] Figure 3 A structural diagram of an automatic positioning system of a workpiece of a numerical control machine tool provided by an embodiment of the present disclosure is shown in the figure.

[0014] The reference signs are explained as follows: a first obtaining module 11, a second obtaining module 12, a third obtaining module 13, a fourth obtaining module 14, a fifth obtaining module 15, a sixth obtaining module 16, a seventh obtaining module 17, and a first processing module 18. DETAILED DESCRIPTION

[0015] Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, which include various details of the embodiments of the present disclosure to assist in understanding, and should be considered as merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Also, in order to be clear and concise, descriptions of well-known functions and structures are omitted in the following description.

[0016] Embodiment one

[0017] Figure 1 A figure provided by an embodiment of the present application is shown in the figure, and the method comprises: Figure 1

[0018] Step S100: performing surface scanning on a target workpiece by a laser radar device to obtain point cloud data;

[0019] Specifically, the laser radar calculates the precise distance to the observed object by emitting and receiving laser beams. The laser radar device is used to scan by a laser radar detector, detect objects with low radio wave reflectivity, or the distance and positional relationship with surrounding obstacles. The target workpiece is the workpiece to be scanned. The point cloud data is the data obtained by scanning the workpiece to be scanned based on the laser radar device. The point cloud data is a set of vectors in a three-dimensional coordinate system, and the scanning data is recorded in the form of points. Each point contains three-dimensional coordinates, which contain geometric positions, and some also contain color information. Further, the laser radar detector of the laser radar device scans the surface of the target workpiece to obtain the point cloud data of the target workpiece.

[0020] Step S200: extracting model feature points from a CAD model of the target workpiece;​

[0021] Specifically, the CAD model is used to describe a virtual model of an object, which can be used to simulate the geometry, shape and size of the object and the relative relationship therebetween. That is, the CAD model of the target workpiece can provide accurate data of the target workpiece. The feature points in the CAD model include end points, midpoints, circle centers, nodes, quadrant points, intersection points, extensions, insertion points, tangent points, appearance intersection points, parallelism, etc. Further, based on the CAD model of the target workpiece, the edges, corner and other feature points of the target workpiece can be extracted as the CAD model feature points of the target workpiece.

[0022] Step S300: registering and matching the point cloud data and the model feature points by a registration algorithm to obtain a registration result;

[0023] Specifically, the registration algorithm inputs two point clouds Ps(source) and Pt(target) and outputs a transformation T to make the coincidence degree of T(Ps) and Pt as high as possible. For example, for two coordinate systems under different perspectives, such as a world coordinate system and a camera coordinate system, a transformation T is needed to be obtained to make the two coordinate systems transformed to a unified perspective, wherein the transformation can include rotation and translation.

[0024] Further, the point cloud data and the model feature points are coarsely registered and finely registered by the registration algorithm to compare or fuse the point cloud data obtained under different collection conditions and the model feature points. For the point cloud data obtained by the laser radar device of the target workpiece and the model feature points obtained by the CAD model, the point cloud data is mapped to the model feature points or the model feature points are mapped to the point cloud data by spatial transformation, so that the points corresponding to the same position in space in the point cloud data and the model feature points are one-to-one corresponding, thereby achieving information fusion. For example, a tea table has multiple cups, the scene is scanned by a depth camera, the objects in the scene need to be scanned multiple times, and then all the objects in the scene are scanned. If the point cloud data under different scene angles is obtained, then the multiple point cloud data is fused to obtain a complete scene.

[0025] Step S400: obtaining a transformation matrix according to the registration result, and determining a first position and a first attitude of the target workpiece in the laser radar coordinate system according to the transformation matrix;

[0026] Step S500: obtaining a first coordinate system result according to the first position and the first attitude;

[0027] Step S600: obtaining a second position and a second attitude of the laser radar coordinate system in the numerical control machine tool coordinate system to obtain a second coordinate system result;

[0028] Specifically, the first position is a position of the target workpiece in the laser radar coordinate system. The first attitude is a posture of the target workpiece in the laser radar coordinate system. For example, the first attitude can be a vertical attitude, a horizontal attitude, or the like. According to the registration result of the point cloud data and the model feature points, a transformation matrix is obtained. For example, it can be a rotation or translation transformation matrix. The first position and the first attitude of the target workpiece in the laser radar coordinate system are determined based on the transformation matrix. Further, based on the first position of the target workpiece and the first attitude of the target workpiece, a result of the target workpiece in the laser radar coordinate system is obtained, and the result of the target workpiece in the laser radar coordinate system is extracted as the first coordinate system result.

[0029] Specifically, the second position is a position of the laser radar coordinate system in the numerical control machine tool coordinate system. The second attitude is a posture of the laser radar coordinate system in the numerical control machine tool coordinate system. The second position and the second attitude of the laser radar coordinate system in the numerical control machine tool coordinate system are obtained, and the second coordinate system result is obtained based on the result of the laser radar coordinate system in the numerical control machine tool coordinate system.

[0030] Step S700: according to the first coordinate system result and the second coordinate system result, calculating a target position and a target attitude of the target workpiece in the numerical control machine tool coordinate system, and obtaining a target coordinate system result;

[0031] The step S700 in the method provided by the embodiment of the application comprises:

[0032] S710: obtaining the first position and the first attitude of the target workpiece in the first coordinate system result, and constructing a first transformation matrix T1;

[0033] S720: obtaining the second position and the second attitude of the target workpiece in the second coordinate system result, and constructing a second transformation matrix T2;

[0034] S730: obtaining a target coordinate system result according to the first transformation matrix T1 and the second transformation matrix T2.

[0035] Specifically, the first position is a position of the target workpiece in the laser radar coordinate system. The first attitude is a posture of the target workpiece in the laser radar coordinate system. According to the first position and the first attitude of the target workpiece in the laser radar coordinate system, a result of the target workpiece in the laser radar coordinate system is obtained, and the result of the target workpiece in the laser radar coordinate system is extracted, and a first transformation matrix T1 is constructed.

[0036] Further, the second position is a position of the laser radar coordinate system in the numerical control machine tool coordinate system. The second attitude is an attitude of the laser radar coordinate system in the numerical control machine tool coordinate system. According to the second position and the second attitude of the laser radar coordinate system in the numerical control machine tool coordinate system, a result of the laser radar coordinate system in the numerical control machine tool coordinate system is obtained, the result of the laser radar coordinate system in the numerical control machine tool coordinate system is extracted, and a second transformation matrix T2 is constructed. Further, according to the first transformation matrix T1 and the second transformation matrix T2, a target transformation matrix of the target workpiece in the numerical control machine tool coordinate system is obtained through matrix multiplication, and a target coordinate system result is obtained.

[0037] Step S800: according to the target coordinate system result, driving the numerical control machine tool to position the target workpiece.

[0038] The step S800 in the method provided by the embodiment of the application comprises:

[0039] S810: extracting a target position of the target workpiece in the numerical control machine tool coordinate system in the target coordinate system result, and obtaining target position data;

[0040] S820: extracting a target attitude of the target workpiece in the numerical control machine tool coordinate system in the target coordinate system result, and obtaining target attitude data;

[0041] S830: inputting the target position data and the target attitude data into the automatic positioning system, and obtaining machine tool motion instructions;

[0042] S840: driving the machine tool motor to position the target workpiece according to the machine tool motion instructions.

[0043] Specifically, the target position of the target workpiece in the numerical control machine tool coordinate system in the target coordinate system result is extracted, that is, a coordinate system point, and target position data is obtained. The target attitude of the target workpiece in the numerical control machine tool coordinate system in the target coordinate system result is extracted, that is, a rotation matrix data, through which the specific attitude of the target workpiece in the coordinate system can be obtained, and target attitude data is further obtained. Wherein, the target position data and the target attitude data are input into the automatic positioning system to obtain machine tool motion instructions, that is, numerical control machine tool operation instructions. Further, the machine tool motor is driven to position the target workpiece according to the machine tool motion instructions.

[0044] Wherein, the point cloud registration and computer vision are used to realize the accurate positioning of the workpiece of the numerical control machine tool, realize the detection, recognition and positioning of the workpiece, improve the automation level of the numerical control machine tool, and realize the intelligent positioning effect.

[0045] The step S100 in the method provided by the embodiment of the application comprises:

[0046] S110: initializing a laser radar coordinate system of the laser radar device;

[0047] S120: planning a scanning path of the laser radar device according to the target workpiece;

[0048] S130: driving the laser radar device to continuously scan the target workpiece according to the scanning path, and obtaining scanning data;

[0049] S140: preprocessing the scanning data to obtain the point cloud data.

[0050] Specifically, the laser radar coordinate system is the data generated by the sensor based on the coordinate system of the sensor itself. The laser radar coordinate system is generally a right-handed coordinate system with the x-axis forward, the y-axis left, and the z-axis upward. The point coordinates measured by the laser radar coordinate system are three-dimensional coordinates in the laser radar coordinate system. Since the laser radar coordinate system needs to be installed in the numerical control machine tool coordinate system, the laser radar measurement points need to be converted from the relative coordinate system to the position points on the absolute coordinate system, so as to be applied to different systems. Further, the laser radar coordinate system of the laser radar device is initialized to enable the laser radar coordinate system to be applied to the numerical control machine tool coordinate system.

[0051] Further, the scanning path of the laser radar device is the path required by the laser radar device to scan the to-be-scanned object. The laser radar device has a transmitter and a receiver. When the laser radar device scans, the laser radar device transmitter emits at different angles to obtain omnidirectional data of the to-be-scanned workpiece. Therefore, the laser radar device can set scanning paths in multiple different directions. Based on the position and size of the target workpiece in the numerical control machine tool, the laser radar device performs omnidirectional scanning on the target workpiece. Therefore, the scanning path of the laser radar device on the target workpiece is planned to enable the laser radar device to perform omnidirectional scanning on the target workpiece in the numerical control machine tool. For example, the laser radar device performs vertical scanning and horizontal scanning on the target workpiece, and then the vertical scanning path and the horizontal scanning path can be planned and set.

[0052] Further, the laser radar generates line beam data during scanning. The more line beam data the laser radar has, the better the scanning effect on the to-be-scanned object. The transmitter of the laser radar device emits laser radar based on different angles. After one polling period, a frame of laser point cloud data is obtained. Since four point cloud data can form surface information, the to-be-scanned object needs to be continuously scanned to obtain multi-dimensional information of the to-be-scanned object. The laser radar device is driven to continuously scan the target workpiece according to the scanning path to obtain multi-dimensional scanning data of the target workpiece.

[0053] Further, the scanning data is preprocessed, and the preprocessing can include data cleaning, data integration, data reduction, data transformation, etc. Then, the point cloud data of the target workpiece is obtained.

[0054] The surface of the target workpiece is scanned by the laser radar device, the point cloud data of the target workpiece is obtained, and the basic information of the target workpiece is obtained to more efficiently perform automatic positioning.

[0055] The step S200 in the method provided by the embodiment of the application includes:

[0056] S210: Extracting high-curvature points of the CAD model to form a first feature point set;

[0057] S220: Extracting boundary points and corner points of the CAD model to form a second feature point set;

[0058] S230: Merging the first feature point set and the second feature point set to form a model feature point.

[0059] Specifically, the curvature is used to describe the bending degree of a curve. The smaller the curvature is, the greater the bending degree is, and the greater the curvature is, the smaller the bending degree is. The curvature point can describe the uneven degree of the measured object. If the curvature point is higher, the measured object is more uneven, and the measured object contains more obvious features. Optionally, the CAD model of the target workpiece has multiple unsorted curvature points. The multiple unsorted curvature points of the CAD model of the target workpiece can be arranged in descending order to obtain the multiple curvature points of the CAD model of the target workpiece in the arrangement position. The curvature of the curvature point arranged in front decreases from front to back, and the curvature point arranged in front is a high-curvature point. The multiple curvature points in the arrangement position are extracted in random number from front to back to form the first feature point set.

[0060] Further, the boundary point is the edge of the measured object. If the measured object is a regular object, the boundary point is the edge of the measured object. If the measured object is an irregular object, the boundary point is the edge of the measured object with high curvature. The corner point is the convex corner of the measured object. If the measured object is a regular object, the corner point is the convex corner of the measured object. If the measured object is an irregular object, the corner point is the convex corner of the measured object with high curvature. The boundary points and corner points of the CAD model of the target workpiece are extracted to form the second feature point set.

[0061] Further, the first feature point set and the second feature point set are merged, and the first feature point set and the second feature point set are clustered to form the CAD model feature point of the target workpiece.

[0062] The model feature point is extracted from the CAD model of the target workpiece to obtain accurate data of the target workpiece.

[0063] The step S230 in the method provided by the embodiment of the present application comprises:

[0064] S231: calculating the curvature values of the high-curvature points in the first feature point set, sorting the curvature values, and selecting N high-curvature points with the largest curvature values, wherein N is a predetermined value, and N>1;

[0065] S232: calculating the feature degrees of the boundary points and the corner points in the second feature point set, sorting the feature degrees, and selecting M boundary points and corner points with the largest feature degrees, wherein M is a predetermined value, and M>1;

[0066] S233: merging the N high-curvature points and the M boundary points and corner points to form a candidate feature point set;

[0067] S234: clustering the candidate feature point set by cluster analysis to obtain K cluster centers, forming model feature points.

[0068] Specifically, the curvature formula is k=1 / r, and the smaller the curvature radius is, the larger the curvature is. The curvature values of the high-curvature points in the first feature point set are calculated, wherein all the curvature radii are extracted, all the curvature radii are input into the curvature formula, the curvature is calculated, and a plurality of curvature values are obtained. Optionally, the curvature values can be sorted in descending order, and the curvature value with the first sorting position is the maximum curvature value, that is, the high-curvature point. The high-curvature points with a random number N are selected in order according to the descending order of the sorting positions of the curvature values, and N is a predetermined random selection number, and N>1.

[0069] Further, the edge radian of the boundary points and the angle of the corner points in the second feature point set are calculated. If the edge radian of the boundary points is larger, the feature degree of the boundary points is larger, and if the angle of the corner points is smaller, the feature degree of the corner points is larger. Optionally, all the feature degrees of the boundary points can be sorted in descending order, and the feature degree with the first sorting position is the maximum. The boundary points and the corner points with a random number M are selected in order according to the descending order of the feature degrees, and M is a predetermined random selection number, and M>1.

[0070] Further, the N high-curvature points and the M boundary points and corner points are merged to form a candidate feature point set. The candidate feature point set is clustered by cluster analysis to obtain K cluster centers, forming model feature points. For example, the boundary point cluster center, the corner point cluster center, etc. can be obtained to form the feature points of the target workpiece CAD model.

[0071] The first feature point set and the second feature point set are merged to form model feature points, so that the target workpiece CAD model can more accurately describe the feature points of the target workpiece.

[0072] The step S300 in the method provided by the embodiment of the present application comprises:

[0073] S310: extracting a point cloud feature point from the point cloud data;

[0074] S320: constructing a corresponding relationship between the point cloud feature point and the model feature point;

[0075] S330: constructing an initial transformation matrix according to the corresponding relationship;

[0076] S340: performing normalization processing on the point cloud data, the CAD model, and the initial transformation matrix to obtain a first input result;

[0077] S350: constructing a matrix optimization model through an IPC algorithm;

[0078] S360: inputting the first input result into the matrix optimization model to obtain a first transformation matrix;

[0079] S370: performing normalization processing on the point cloud data, the CAD model, and the first transformation matrix to obtain a second input result;

[0080] S380: inputting the second input result into the matrix optimization model to obtain a second transformation matrix, wherein the second transformation matrix is the registration result.

[0081] Specifically, based on the target workpiece point cloud data obtained by laser radar scanning, a point cloud feature point is extracted. A corresponding relationship between the point cloud feature point and the model feature point is constructed, wherein the point cloud feature point and the model feature point can be a mapping relationship. Based on the mapping of the point cloud feature point to the model feature point, correspondingly, based on the mapping of the model feature point to the point cloud feature point.

[0082] Further, according to the corresponding relationship, an initial transformation matrix is constructed, and the transformation matrix is an estimated value. The point cloud data, the CAD model, and the initial transformation matrix are normalized. The dimensional point cloud data, the CAD model, and the initial transformation matrix are transformed into dimensionless expressions and become scalars. Further, a first input result is obtained.

[0083] Further, the ICP algorithm is an iterative closest point algorithm, the ICP algorithm has a good initial position for two pieces of point clouds to be registered, that is, two pieces of point clouds are roughly aligned. The closest points in the two pieces of point clouds are selected as corresponding points, a rotation and translation transformation matrix is solved through all the corresponding points, and the error between the two pieces of point clouds is made smaller and smaller through continuous iteration until a preset threshold or the number of iterations is met. The matrix optimization model is constructed through the ICP algorithm, that is, the initial transformation matrix constructed according to the corresponding relationship between the feature points of the point cloud and the feature points of the model is input into the ICP algorithm for optimization to obtain the matrix optimization model. The first input result is input into the matrix optimization model to obtain the first change matrix, and the point cloud coarse registration algorithm is obtained.

[0084] Further, the point cloud data, the CAD model, and the first change matrix are normalized, the dimensional point cloud data, the CAD model, and the first change matrix are transformed into dimensionless expressions, and become scalars. Further, the second input result is obtained. The second input result is input into the matrix optimization model to obtain the second change matrix, wherein the second change matrix is a registration result. That is, the ICP algorithm is iterated twice to obtain the point cloud fine registration algorithm.

[0085] Wherein, the point cloud data and the model feature points are matched and registered by the registration algorithm, so that the point cloud data and the feature points are matched, and then the target element can be accurately positioned in the numerical control machine tool.

[0086] Embodiment two

[0087] Based on the same inventive concept as the automatic positioning method of a workpiece of a numerical control machine tool in the foregoing embodiment, as shown in Figure 3 The application also provides an automatic positioning system for a workpiece of a numerical control machine tool, the system comprising:

[0088] A first obtaining module 11 acquires point cloud data by surface scanning of a target workpiece through a laser radar device;

[0089] A second obtaining module 12 extracts model feature points from a CAD model of the target workpiece;

[0090] A third obtaining module 13 matches and registers the point cloud data and the model feature points through a registration algorithm to obtain a registration result;

[0091] A fourth obtaining module 14 acquires a transformation matrix according to the registration result and determines a first position and a first attitude of the target workpiece in the laser radar coordinate system according to the transformation matrix;

[0092] A fifth obtaining module 15 acquires a first coordinate system result according to the first position and the first attitude;

[0093] A sixth obtaining module 16 obtains a second position and a second pose of the laser radar coordinate system in the numerical control machine tool coordinate system, and obtains a second coordinate system result;

[0094] A seventh obtaining module 17 calculates a target position and a target pose of the target workpiece in the numerical control machine tool coordinate system according to the first coordinate system result and the second coordinate system result, and obtains a target coordinate system result;

[0095] A first processing module 18 drives the numerical control machine tool to position the target workpiece according to the target coordinate system result.

[0096] Further, the system further comprises:

[0097] A second processing module initializes a laser radar coordinate system of the laser radar device;

[0098] A third processing module plans a scanning path of the laser radar device according to the target workpiece;

[0099] An eighth obtaining module drives the laser radar device to continuously scan the target workpiece according to the scanning path, and obtains scanning data;

[0100] A ninth obtaining module pre-processes the scanning data, and obtains the point cloud data.

[0101] Further, the system further comprises:

[0102] A first constructing module extracts high-curvature points of the CAD model to form a first feature point set;

[0103] A second constructing module extracts boundary points and corner points of the CAD model to form a second feature point set;

[0104] A third constructing module merges the first feature point set and the second feature point set to form model feature points.

[0105] Further, the system further comprises:

[0106] A fourth processing module calculates curvature values of the high-curvature points in the first feature point set, sorts the curvature values, and selects N high-curvature points with the largest curvature values, where N is a predetermined value and N>1;

[0107] A fifth processing module calculates feature degrees of the boundary points and the corner points in the second feature point set, sorts the feature degrees, and selects M boundary points and corner points with the largest feature degrees, where M is a predetermined value and M>1;

[0108] A fourth construction module merges the N high-curvature points and the M boundary points and corner points to form a candidate feature point set;

[0109] A fifth construction module clusters the candidate feature point set by cluster analysis to obtain K cluster centers to form a model feature point.

[0110] Further, the system further comprises:

[0111] A sixth processing module extracts a point cloud feature point from the point cloud data;

[0112] A sixth construction module constructs a correspondence between the point cloud feature point and the model feature point;

[0113] A seventh construction module constructs an initial transformation matrix according to the correspondence;

[0114] A tenth obtaining module normalizes the point cloud data, the CAD model, and the initial transformation matrix to obtain a first input result;

[0115] An eleventh obtaining module constructs a matrix optimization model by an ICP algorithm;

[0116] A twelfth obtaining module inputs the first input result into the matrix optimization model to obtain a first transformation matrix;

[0117] A thirteenth obtaining module normalizes the point cloud data, the CAD model, and the first transformation matrix to obtain a second input result;

[0118] A fourteenth obtaining module inputs the second input result into the matrix optimization model to obtain a second transformation matrix, wherein the second transformation matrix is the registration result.

[0119] Further, the system further comprises:

[0120] A fifteenth obtaining module obtains the first position and the first pose of the target workpiece in the first coordinate system result to construct a first transformation matrix T1;

[0121] A sixteenth obtaining module obtains the second position and the second pose of the target workpiece in the second coordinate system result to construct a second transformation matrix T2;

[0122] A seventeenth obtaining module obtains a target coordinate system result according to the first transformation matrix T1 and the second transformation matrix T2.

[0123] Further, the system further comprises:

[0124] The eighteenth module extracts the target position of the target workpiece in the CNC machine tool coordinate system from the target coordinate system result and obtains the target position data;

[0125] The nineteenth module extracts the target posture of the target workpiece in the CNC machine tool coordinate system from the target coordinate system result and obtains the target posture data;

[0126] The twentieth acquisition module inputs the target position data and the target attitude data into the automatic positioning system to obtain machine tool motion commands;

[0127] The machine tool motor is driven to position the target workpiece according to the machine tool motion command.

[0128] The specific example of the automatic positioning method for CNC machine tool workpieces in Embodiment 1 described above is also applicable to the automatic positioning system for CNC machine tool workpieces in this embodiment. Through the foregoing detailed description of the automatic positioning method for CNC machine tool workpieces, those skilled in the art can clearly understand the automatic positioning system for CNC machine tool workpieces in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here. As for the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant details can be found in the method section.

[0129] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders.

[0130] This document does not impose any restrictions on the technical solutions disclosed herein, as long as the desired results can be achieved.

[0131] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method of automatic positioning of a workpiece in a numerically controlled machine tool, characterised in that, The method is applied to an automatic positioning system in communication connection with a laser radar device, and the method comprises: performing surface scanning on a target workpiece by the laser radar device to obtain point cloud data; extracting model feature points from a CAD model of the target workpiece; performing registration matching on the point cloud data and the model feature points by a registration algorithm to obtain a registration result; obtaining a transformation matrix according to the registration result, and determining a first position and a first attitude of the target workpiece in the laser radar coordinate system according to the transformation matrix; obtaining a first coordinate system result according to the first position and the first attitude; obtaining a second position and a second attitude of the laser radar coordinate system in a numerical control machine tool coordinate system to obtain a second coordinate system result; calculating a target position and a target attitude of the target workpiece in the numerical control machine tool coordinate system according to the first coordinate system result and the second coordinate system result to obtain a target coordinate system result; driving the numerical control machine tool to position the target workpiece according to the target coordinate system result; the extracting model feature points from the CAD model of the target workpiece comprises: extracting high-curvature points of the CAD model to form a first feature point set; extracting boundary points and corner points of the CAD model to form a second feature point set; merging the first feature point set and the second feature point set to form model feature points; the merging the first feature point set and the second feature point set to form model feature points comprises: calculating curvature values of the high-curvature points in the first feature point set, and sorting the curvature values to select N high-curvature points with the largest curvature values, wherein N is a predetermined value, and N>1; calculating feature degrees of the boundary points and the corner points in the second feature point set, and sorting the feature degrees to select M boundary points and corner points with the largest feature degrees, wherein M is a predetermined value, and M>1; merging the N high-curvature points and the M boundary points and corner points to form a candidate feature point set; obtaining K clustering centers by clustering analysis on the candidate feature point set to form model feature points; the performing registration matching on the point cloud data and the model feature points by a registration algorithm to obtain a registration result comprises: extracting point cloud feature points from the point cloud data; constructing a corresponding relationship between the point cloud feature points and the model feature points; constructing an initial transformation matrix according to the corresponding relationship; performing normalization processing on the point cloud data, the CAD model and the initial transformation matrix to obtain a first input result; constructing a matrix optimization model by an ICP algorithm; inputting the first input result into the matrix optimization model to obtain a first transformation matrix; performing normalization processing on the point cloud data, the CAD model and the first transformation matrix to obtain a second input result; inputting the second input result into the matrix optimization model to obtain a second transformation matrix, wherein the second transformation matrix is the registration result.

2. The method of claim 1, wherein, the performing surface scanning on a target workpiece by the laser radar device to obtain point cloud data comprises: Initialize a laser radar coordinate system of the laser radar device; Plan a scanning path of the laser radar device according to the target workpiece; Drive the laser radar device to continuously scan the target workpiece according to the scanning path, and obtain scanning data; Preprocess the scanning data to obtain the point cloud data.

3. The method of claim 1, wherein the target coordinate system result is obtained by calculating the position and attitude of the target workpiece in the numerical control machine tool coordinate system according to the first coordinate system result and the second coordinate system result, comprising: obtaining the first position and first attitude of the target workpiece in the first coordinate system result to construct a first transformation matrix T1; obtaining the second position and second attitude of the target workpiece in the second coordinate system result to construct a second transformation matrix T2; obtaining the target coordinate system result according to the first transformation matrix T1 and the second transformation matrix T2.

4. The method of claim 1, wherein the target workpiece is positioned by the numerical control machine tool according to the target coordinate system result, comprising: extracting the target position of the target workpiece in the numerical control machine tool coordinate system from the target coordinate system result to obtain target position data; extracting the target attitude of the target workpiece in the numerical control machine tool coordinate system from the target coordinate system result to obtain target attitude data; inputting the target position data and the target attitude data into the automatic positioning system to obtain machine tool motion instructions; positioning the target workpiece by driving the machine tool motor according to the machine tool motion instructions.

5. An automatic positioning system for workpieces of a numerically controlled machine, characterized in that, The system is used to execute the method of any one of claims 1-4, and is applied to an automatic positioning system and in communication with a laser radar device, and comprises: a first obtaining module that obtains point cloud data by surface scanning of a target workpiece by a laser radar device; a second obtaining module that extracts model feature points from a CAD model of the target workpiece; a third obtaining module that obtains a registration result by registration and matching of the point cloud data and the model feature points by a registration algorithm; a fourth obtaining module that obtains a transformation matrix according to the registration result, and determines a first position and a first attitude of the target workpiece in the laser radar coordinate system according to the transformation matrix; a fifth obtaining module that obtains a first coordinate system result according to the first position and the first attitude; a sixth obtaining module that obtains a second coordinate system result by obtaining a second position and a second attitude of the laser radar coordinate system in a numerical control machine tool coordinate system; a seventh obtaining module that obtains a target coordinate system result by calculating a target position and a target attitude of the target workpiece in the numerical control machine tool coordinate system according to the first coordinate system result and the second coordinate system result; a first processing module that positions the target workpiece by driving a numerical control machine tool according to the target coordinate system result.

Images