A digital measurement method for assembly gap of a large aspect ratio component

Abstract

The present application relates to a kind of digital measurement methods of large deep-width ratio assembly gap, comprising the following steps, scanning obtains TR component assembly area cold plate and component point cloud data P, point cloud data is initially split into TR component point cloud P by clustering z And cold plate point cloud P l , respectively, the installation hole feature and the positioning plane feature of point cloud P z And P l According to the hole feature and the positioning plane feature extracted, corresponding TR component and cold plate theoretical model are aligned to point cloud data, the gap distribution between transceiver component and cold plate is calculated according to the relative pose of theoretical model, the gap distribution is converted into scalar and is shown on model in the form of color difference chart, in short, the present application is used for the recovery of theoretical model to assembly gap by efficient feature extraction algorithm and registration method, to realize the calculation of complete assembly gap in invisible area, substantially improve the gap calculation precision and data processing efficiency.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of automatic assembly of radar antenna electronic products, and particularly relates to a digital measurement method for assembly gap of a large-depth-width-ratio component. BACKGROUND

[0002] With the development of modern electronic products towards multi-function and high integration, the corresponding assembly process has put forward higher requirements in terms of assembly precision and reliability; in order to reduce costs and make full use of space, the space for assembling internal components in electronic products such as radar antennas is very compact, and during the assembly process, assembly gaps between components may be caused due to part errors, assembly force and connector precision, etc. The generation of the gap will seriously hinder the heat dissipation efficiency of the TR component during operation and reduce the service life of the radar antenna.

[0003] After the assembly of the components, the components are in a state of adhesion and the space is closed, but in order to be used normally, the components still need to be detected. The traditional assembly gap measurement method usually relies on manual contact measurement, such as a plug gauge or a vernier caliper, but these methods can only measure the edge gap of the assembly area and cannot obtain the overall gap distribution between the components. In addition, contact measurement is low in efficiency and may affect the assembly precision, and cannot realize timely adjustment of the assembly process.

[0004] With the progress of digital measurement technology, the assembly gap of the components can be measured digitally by using a line laser measuring instrument. The point cloud data obtained by scanning is analyzed and processed to analyze the assembly gap. However, due to the large depth-width ratio of the gap, part of the gap with a ratio of 40:1 is also called a large-depth-width-ratio assembly gap. The line laser measuring instrument still cannot obtain measurement data inside the large-depth-width-ratio gap, which makes it difficult to digitally measure the overall appearance of the gap. Therefore, there is an urgent need for a digital measurement method for large-depth-width-ratio assembly gap of components, which can quickly, accurately and conveniently detect the data inside the large-depth-width-ratio gap, so as to truly reflect the actual situation of the gap. SUMMARY

[0005] Therefore, the present application provides a digital measurement method for large-depth-width-ratio assembly gap of components, which is applied to the field of automatic assembly of radar antenna electronic products and can solve the technical problems that the existing method can only measure the edge of the gap and cannot obtain measurement data inside the large-depth-width-ratio components.

[0006] In order to achieve the above technical purposes, the specific technical scheme adopted by the present application is as follows:

[0007] A digital measurement method for large-depth-width-ratio assembly gap of components, comprising the following steps:

[0008] S1, scanning and obtaining point cloud data P of a cold plate and components in a TR component assembly area;

[0009] S2, preliminary split the point cloud data into TR component point cloud P by clustering z and cold plate point cloud P l ;

[0010] S3, extract mounting hole features and positioning plane features from point clouds P z and P l respectively;

[0011] S4, align the corresponding TR component and cold plate theoretical model to the point cloud data according to the extracted hole features and positioning plane features;

[0012] S5, calculate the gap distribution between the transceiver assembly and the cold plate according to the relative pose of the theoretical model;

[0013] S6, convert the gap distribution into a scalar to display it as a color difference map on the model;

[0014] In step S3, for the cold plate and TR component obtained after segmentation, two positioning holes and one positioning plane are extracted as subsequent registration features, and core data is used to participate in calculation to improve registration accuracy and algorithm efficiency;

[0015] In step S4, for the TR component point cloud data and cold plate point cloud data obtained in step S3, the theoretical model is coarsely registered with the point cloud data through the extracted plane and hole features, and then finely registered through the ICP algorithm.

[0016] Further, in step S1, after the TR component is assembled to the antenna frame, a mechanical arm carries a line laser scanner to scan the assembled TR component and cold plate region. The scanner is located directly above the TR component and cold plate during scanning, and the scanning distance is within the effective measurement distance range of the line laser scanner.

[0017] Further, in step S1, the scanning path is determined according to the cold plate trend, and the scanning range covers all TR components and cold plate regions exposed to the outside.

[0018] Further, in step S2, the original point cloud data obtained by scanning is preprocessed to remove outliers in the point cloud to obtain point cloud P c , a growth clustering strategy based on radius search is used to segment the point cloud, and the point cloud P c is segmented into cold plate point cloud P l and component point cloud P z .

[0019] Further, in step S3, for the TR component, the plane p mThen, the points within a certain threshold of the plane formed in the previous step are used as input point cloud data, and the boundary point set S in the point cloud data is extracted by the region growing method based on curvature e ; Then in the boundary point set S e The RANSAC method is used to extract S e The circular features in the figure are used as the features h1 and h2 of the two holes. The structures of the segmented cold plate and TR component point cloud data are similar, so the feature extraction of the cold plate structure will not be introduced in detail. Finally, the plane data corresponding to the cold plate structure and the corresponding two positioning hole data are obtained.

[0020] Furthermore, the coarse registration matrix of the TR component theoretical model obtained by the coarse registration calculation in step S4 is W1, the coarse registration matrix of the cold plate theoretical model is W1', the fine registration matrix of the TR component theoretical model obtained by the fine registration calculation is W2, and the fine registration matrix of the cold plate theoretical model is W2'. The transformations W2·W1 and W2'·W1' are applied to the TR component theoretical model and the cold plate theoretical model respectively, so that the TR component theoretical model and the cold plate theoretical model have a relative posture consistent with that in the actual assembly state.

[0021] Furthermore, in step S5, the fitting plane of the TR component and the cold plate theoretical model is downsampled into a point cloud L z and L l , with L l As a benchmark, calculate L z Each point to point cloud L l The closest distance is taken as the gap value at that point.

[0022] Furthermore, in step S6, the gap value is converted into a corresponding color through color mapping according to the gap range, and a cloud diagram of the gap distribution between the TR component and the cold plate is generated.

[0023] By adopting the above technical solution, the present invention can also bring the following beneficial effects:

[0024] 1. The present invention proposes a digital measurement method for assembly gaps of components with large aspect ratios. Through an efficient feature extraction algorithm and registration method, a theoretical model is used to complete and recover gap data, overcoming the problem of incomplete gap data caused by spatial occlusion and narrow viewing angle. The completed fused data can be used to obtain the overall gap distribution in the gap area between components with large aspect ratios, breaking through the problem of unmeasurable internal gaps due to missing data.

[0025] 2、The digital measurement method of the assembly gap of the large-depth-width-ratio component mentioned in the application is very suitable for the assembly gap of the large-depth-width-ratio component, especially the gap data of most of the assembly gap located on the fitting surface cannot be obtained, the whole scanning and analysis process is realized automatically through the mechanical arm carrying the scanner, the automation degree and the gap detection efficiency are greatly improved, not only the accuracy is greatly improved, but also the use range of the line laser scanner measurement is expanded, which has very positive significance. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the application, and all other drawings obtained by those of ordinary skill in the art without creative labor based on these drawings also belong to the protection scope of the application.

[0027] Figure 1 The flow chart of the digital measurement method of the assembly gap of the large-depth-width-ratio component mentioned in the application;

[0028] Figure 2 The schematic diagram of the positional relationship between the line laser scanner in the use state and the radar antenna in the specific embodiment of the application;

[0029] Figure 3 The schematic diagram of the positional relationship structure between the TR component and the antenna cold plate in the specific embodiment of the application;

[0030] Figure 4 The assembly gap distribution cloud diagram of the TR component in the specific embodiment of the application;

[0031] 1, mechanical arm; 2, line laser scanner; 3, radar antenna; 4, TR component; 5, cold plate. DETAILED DESCRIPTION

[0032] The embodiments of the application will be described in detail below with reference to the drawings.

[0033] The embodiments of the application are described below through specific examples. Those skilled in the art can easily understand other advantages and effects of the application from the content disclosed in the specification. Obviously, the described embodiments are only some of the embodiments of the application, not all. The application can also be implemented or applied by other different specific embodiments, and the details in the specification can be modified or changed based on different views and applications without departing from the spirit of the application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the application, all other embodiments obtained by those of ordinary skill in the art without creative labor also belong to the protection scope of the application.

[0034] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present invention, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement an apparatus and / or practice a method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this apparatus and / or practice this method.

[0035] It should also be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. The illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0036] Additionally, in the following description, specific details are provided to provide a thorough understanding of the examples. However, one skilled in the art will appreciate that the aspects described can be practiced without these specific details.

[0037] In one embodiment of the present invention, a digital measurement method for assembly gaps of components with a large aspect ratio is provided, comprising the following steps:

[0038] Step S1, scanning and obtaining point cloud data of the TR component 4 and the cold plate 5;

[0039] Point cloud data acquisition, such as Figure 2 As shown, the point cloud data of the measured objects TR component 4 and cold plate 5 are acquired by a robot arm 1 (KUKA KR20R1810) equipped with a line laser scanner 2 (LJ-X8000). The scanning path of the robot arm 1 is determined according to the direction of the cold plate 5, the width of the line laser, the effective scanning distance of the scanner, and the direction of the radar antenna 3.

[0040] The running trajectory of the robot arm 1 can be planned according to the digital model corresponding to the radar antenna 3, and can also be determined according to the number and area of ​​the TR components 4 actually assembled on site.

[0041] Converting 2D line laser data into 3D point cloud data. Since the data directly measured by the line laser only contains the elevation data of the measured surface, in order to obtain complete point cloud data of TR component 4 and cold plate 5, it is necessary to convert the 2D line laser data into 3D point cloud data.

[0042] In this embodiment, the line laser scanner 2 acquires the surface profile of the measured object at a fixed sampling frequency h, and the mechanical arm 1 runs at a speed v, so the scanning resolution along the path of the mechanical arm 1 is v / h. Assuming that the path of the mechanical arm 1 is a zigzag type route consistent with the path of the cold plate 5 and parallel to the y axis, the x coordinate of each point in each frame of line laser has been calibrated, and assuming that the position of the mechanical arm 1 at time t0 is (x0, y0), the y direction position of the point cloud data acquired after time t is (x0, y0+v*t), and according to the above formula, the height data of each frame of line laser and the current plane coordinate value can be fused to form three-dimensional point cloud data.

[0043] Step S2, clustering and segmenting the point cloud data into TR component 4 point cloud and cold plate 5 point cloud;

[0044] After acquiring the overall point cloud data, it is necessary to segment the TR component 4 and cold plate 5 data therein. In this embodiment, since the structure of the cold plate 5 and the TR component 4 has a clear boundary, a region growing method based on geometric continuity is used to segment the point cloud, which is as follows:

[0045] Randomly select a seed point in the point cloud, search for points in the vicinity with a search radius r as the threshold, and include them in the same region. Then, take these newly added points as new seed points, repeat the above process until no more points can be added to the region.

[0046] When all the points are assigned to a region, the segmentation of the point cloud data is completed, and finally the point cloud data p z of the TR component 4 and the point cloud data p l of the cold plate 5 are obtained respectively.

[0047] The search radius r of the seed point needs to be greater than the scanning resolution v / h of the line laser and less than the minimum empirical value of the assembly gap, so as to avoid point cloud data segmentation failure caused by threshold lower than the scanning resolution or greater than the assembly gap.

[0048] Step S3, extracting positioning holes and plane features from the TR component 4 and cold plate 5 point cloud data respectively;

[0049] For the segmented cold plate 5 and TR component 4 point cloud data, their structures have similarities, and the main structure is composed of the upper surface plane of the cold plate 5 and the TR component 4 and the two positioning holes located in the plane.

[0050] Taking the TR component 4 as an example, the feature extraction is described. First, the RANSAC (Random Sample Consensus) algorithm is used to process the TR component 4 point cloud p z , and extract the plane feature in the point cloud as the main plane p m .

[0051] Take the distance P m Points within a certain threshold range as input point cloud P m Further extract the boundary features therein.

[0052] Since the input point cloud is in a plane, the present example uses an inlier method based on Euclidean distance to judge the boundary points, and the specific steps are as follows:

[0053] For any point m (x0, y0) in P m , set a neighborhood radius r, search for all points within the neighborhood range whose distance is less than 2r for each point, and form a set O;

[0054] In set O, an optional point n (x1, y1) is located in the circle with a radius of r passing through points m and n in the plane P m , two circles can be obtained, whose centers C 1 and C 2 The coordinate calculation formula is as follows:

[0055]

[0056]

[0057] Among them,

[0058]

[0059] L 2 = (x0-x1) 2 + (y0-y1) 2

[0060] For all points in set O except points m and n, calculate the distance from the two circle centers C 1 and C 2 .

[0061] If the distance of all points is greater than r, i.e. located outside the circle, then point m is determined as a boundary point.

[0062] If there is a point in set O whose distance is less than r, then other points are selected as point n, and step 3 is repeated. If there is no point in set O that satisfies step 4, then point m is a non-boundary point.

[0063] Return to step 1 to judge other points in P m , until all points are traversed.

[0064] After the boundary extraction operation, a point cloud composed of the outer shape boundary of the TR component 4 and the internal hole boundary is obtained, and the RANSAC algorithm is used to take the hole diameter as the diameter of the circle to be fitted, and the positioning hole point cloud h1, h2 in the remaining point cloud is extracted.

[0065] Feature point cloud data F extracted through the above steps p including positioning plane data p m and two positioning hole data h1, h2.

[0066] For the segmented cold plate 5 and TR assembly 4 point cloud data, the structure has similarity, and the feature extraction of the cold plate 5 structure will not be introduced in detail, and finally the corresponding plane data and the corresponding two positioning hole data of the cold plate 5 structure are obtained.

[0067] Step S4, register the TR assembly 4 and the cold plate 5 theoretical model to the point cloud data through the positioning hole and the plane feature respectively;

[0068] According to the plane and hole features obtained in the foregoing steps as constraints, register the TR assembly 4 corresponding point cloud data and the theoretical model.

[0069] Since the cold plate 5 structure is basically consistent with the TR assembly 4, the same method is used for data processing, and the corresponding positioning hole data and plane data are obtained. The plane and hole features of the cold plate 5 obtained are used as constraints to register the cold plate 5 corresponding point cloud data and the theoretical model

[0070] Still taking the registration of the TR assembly 4 point cloud and the theoretical model as an example, assuming that the TR assembly 4 theoretical model is Q, and the actual point cloud data is p z , the theoretical positioning plane and the two hole feature data are Ω, and the corresponding point cloud data are Ψ.

[0071] According to the above data, the theoretical model is registered with the actual point cloud data, in order to improve the data registration accuracy, an accurate registration method based on assembly positioning feature constraints is designed, and the optimization function is defined as follows:

[0072]

[0073] Wherein, q i ∈Q, p i ∈P z ,ω j ∈Ω, ψ j ∈Ψ and N q is the number of points on the TR assembly 4 theoretical model participating in registration, N f is the number of points on the reference feature (plane, hole) theoretical model participating in registration. At the same time, the threshold value δ=0.005 is determined through actual test.

[0074] The above optimization problem can be converted into a quadratic programming problem under the condition of bounded quadratic constraint, and finally the rotation matrix R and the translation vector t of accurate registration are obtained.

[0075] The rotation matrix R and the translation vector t are applied to the theoretical model of the TR assembly 4 to make the pose of the theoretical model the same as the pose of the actual assembly, thereby recovering the clearance data in the unmeasurable area of the TR assembly 4.

[0076] Since the structure of the cold plate 5 is basically the same as that of the TR assembly 4, the same registration process is performed to obtain the rotation matrix and the translation vector.

[0077] Step S5: Calculate the clearance distribution of the TR assembly 4 and the cold plate 5.

[0078] Calculate the assembly clearance between the theoretical model of the TR assembly 4 and the cold plate 5.

[0079] First, the theoretical model is converted into point cloud data, and then within a certain neighborhood range, generally with a neighborhood radius less than 3mm, the nearest distance from each point in the TR assembly 4 to all points in the neighborhood range of the cold plate 5 point cloud is calculated as the clearance value of the point.

[0080] Step S6: Convert the clearance value into a cloud chart and display it on the model.

[0081] Store the clearance values of all points in the distance matrix M, normalize the distance matrix M to the range of [0, 1], and map it to the color space to generate a color difference cloud chart, as shown in Figure 4 where the color of each point is determined by its clearance value to another point cloud.

[0082] In summary, the present patent uses a theoretical model for assembly clearance recovery through an efficient feature extraction algorithm and registration method, thereby realizing the calculation of complete assembly clearance in invisible areas and greatly improving the clearance calculation accuracy and data processing efficiency.

[0083] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method of digital measurement of assembly gaps of large aspect ratio components, characterized in that, The method comprises the following steps: S1, scanning to obtain point cloud data P of a TR component (4) and a cold plate (5) in an assembly area; S2, preliminary splitting of the point cloud data into TR component (4) point clouds P by clustering z and cold plate (5) point clouds P l ; S3, respectively, to the point cloud P z and P l extracting mounting hole features and positioning plane features; S4, aligning the corresponding TR component (4) and cold plate (5) theoretical model to the point cloud data according to the extracted hole features and positioning plane features; S5, calculating the gap distribution between the TR component and the cold plate (5) according to the relative pose of the theoretical model; S6, converting the gap distribution into a scalar and displaying it on the model in the form of a color difference map; In the step S3, for the cold plate (5) and the TR component (4) obtained after segmentation, two positioning holes and one positioning plane are extracted as subsequent registration features, and core data is used to participate in calculation, so as to improve the registration accuracy and algorithm efficiency; In the step S4, for the TR component (4) point cloud data and the cold plate (5) point cloud data obtained in the step S3, the theoretical model is coarsely registered with the point cloud data through the extracted plane and hole features, and then the ICP algorithm is used for fine registration.

2. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 1, wherein: In the step S1, after the TR component (4) is assembled to the antenna frame, a mechanical arm (1) carries a line laser scanner (2) to scan the assembled TR component (4) and cold plate (5) area, and the scanner is located directly above the TR component (4) and cold plate (5) during scanning, and the scanning distance is within the effective measurement distance range of the line laser scanner (2).

3. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 2, wherein: In the step S1, the scanning path is determined according to the direction of the cold plate (5), and the scanning range covers all the exposed areas of the TR component (4) and the cold plate (5).

4. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 3, wherein: The step S2 is to preprocess the original point cloud data obtained by scanning, remove outliers in the point cloud to obtain a point cloud P c , adopt a growth clustering strategy based on radius search to segment the point cloud, segment the point cloud P c into a cold plate (5) point cloud P l and a component point cloud P z .

5. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 4, wherein: In step S3, for the TR component (4), the plane p is extracted from the point cloud of the TR component (4) by the RANSAC algorithm. m Then, the points within a certain threshold of the plane formed in the previous step are used as input point cloud data, and the boundary point set S in the point cloud data is extracted by the region growing method based on curvature e ; Then in the boundary point set S e The RANSAC method is used to extract S e The circular feature in , and as the features of the two holes h1, h2.

6. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 5, wherein: In the step S4, the coarse registration matrix of the TR component (4) theoretical model obtained by coarse registration calculation is W1, the coarse registration matrix of the cold plate (5) theoretical model is W1', the registration matrix of the TR component (4) theoretical model obtained by fine registration calculation is W2, and the fine registration matrix of the cold plate (5) theoretical model is W2', the transformation W2·W1 and W2'·W1' are respectively applied to the TR component (4) theoretical model and the cold plate (5) theoretical model, so that the TR component (4) theoretical model and the cold plate (5) theoretical model have the same relative pose as in the actual assembly state.

7. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 6, wherein: The step S5 fits the plane of the TR assembly (4) and the theoretical model of the cold plate (5) to the point cloud L z and L l , as a reference, calculates the nearest distance of each point on L l to the point cloud L z as the gap value at that point. l ​ 8. A method of digital measurement of assembly gap of a large aspect ratio component as claimed in claim 7, wherein: In the step S6, the gap value is converted into the corresponding color through color mapping according to the gap range, and the gap distribution cloud diagram of the TR component (4) and the cold plate (5) is generated.

Images