Carbon tube array honeycomb surface accuracy calculation method based on machining G code guidance

Abstract

The present invention provides a method for calculating the surface accuracy of a carbon tube array honeycomb based on the guidance of a machining G code. The present invention comprises the following steps: obtaining a measurement G code that matches the surface contour of a design model; measuring the surface data of the carbon tube array honeycomb multiple times by a line laser device to generate a number of three-dimensional point cloud data; splicing and denoising the multiple acquired point cloud data to obtain a complete surface shape of the workpiece surface, and storing them as measurement point cloud data; obtaining a new position point cloud by matrix transformation of the measurement point cloud data according to the starting position information of the G code, and roughly aligning the design model with the new position point cloud; projecting the new position point cloud onto the face of the design model to obtain a model point cloud corresponding to the number of points in the measurement point cloud, and after alignment, calculating the length of the vertical line from the point cloud to the corresponding face to obtain an error length set, and calculating the surface shape error accordingly. The measurement point cloud of the present invention has a high degree of overlap with the model, and is applicable to a variety of difficult-to-measure curved surfaces.

Description

Technical Field

[0001] The invention relates to the field of large curved surface shape detection of composite material antenna reflection panels, and in particular to a carbon tube array honeycomb surface shape accuracy calculation method based on machining G code guidance. Background Art

[0002] In the field of complex surface shape detection, high-precision surface shape detection and evaluation has become the main development trend because of its leading role in subsequent correction work. However, most of the current surface shape accuracy detection methods are aimed at continuous surfaces with obvious features, and most of the results rely on some reverse engineering software. If the workpiece has no obvious sharp features, it often takes multiple repetitions to obtain more reliable results.

[0003] In the surface shape accuracy calculation method, the research methods of plane and parametric model are relatively mature. The surface shape accuracy can be obtained by parametric surface shape fitting. For some non-parametric models, the point cloud model that relies on feature points or has a high degree of overlap with the model can also obtain a more accurate surface shape by registering the point cloud with the model. However, for the non-continuous surface shape of carbon tube array honeycomb, it has the following characteristics:

[0004] 1. The design model may be a partial surface cut from a hyperboloid, lacking a fixed reference datum and no parametric equation to express the surface;

[0005] 2. The measured point cloud has less overlap with the model;

[0006] 3. The carbon tube array honeycomb measurement point cloud is in the shape of a cylindrical point cloud distributed in an array, with no obvious feature points;

[0007] The above-mentioned surface shape calculation methods are not applicable to carbon tube array honeycombs. Therefore, in order to solve this problem, it is necessary to provide a calculation method for the surface shape accuracy of non-continuous large carbon tube array honeycombs. Summary of the invention

[0008] According to the technical problems raised above, a method for calculating the surface accuracy of a carbon tube array honeycomb based on the guidance of machining G code is provided. The technical means adopted by the present invention are as follows:

[0009] A method for calculating the surface accuracy of a carbon tube array honeycomb based on machining G code guidance comprises the following steps:

[0010] Step 1, obtaining the measurement G code that matches the surface profile of the design model;

[0011] Step 2, using a line laser device to measure the carbon tube array honeycomb surface data multiple times to generate a number of three-dimensional point cloud data;

[0012] Step 3: splice and denoise the multiple point cloud data obtained to obtain the complete surface shape of the workpiece surface and store it as measured point cloud data;

[0013] Step 4: According to the starting position information of the G code, the measured point cloud data is transformed into a new position point cloud through matrix transformation, and the design model is roughly registered with the new position point cloud;

[0014] Step 5: Project the new position point cloud onto the face of the design model to obtain the model point cloud corresponding to the number of points in the measured point cloud. After registration, calculate the length of the perpendicular line from the point cloud to the corresponding face to obtain the error length set, and calculate the surface error accordingly.

[0015] Furthermore, the step 1 includes the steps of generating a code and storing a code, specifically:

[0016] Use UG software to generate G code for measurement. Specifically, for workpieces with large curvature and known design models, multi-segment broken line interpolation is used to ensure that the workpiece surface is always within the range of the online laser. During the UG programming process, the one-way parallel milling method is used to obtain the scanning G code file.

[0017] Extract the coordinate information in the G code and store it as a G code point cloud. The point set is Q'={Q'i=(x 1 'i,y 1 'i, z 1 'i)|(i=1, 2, 3,..., Nm')}.

[0018] Furthermore, the step 2 comprises the following steps:

[0019] S201. Use a line laser device that matches the machine tool and mount the line laser device on a three-axis or five-axis machine tool.

[0020] S202, adjusting the position of the calibration line laser equipment and the machine tool to calibrate the rotation error of the calibration line laser in several axes to ensure that the scanned data can be accurately spliced;

[0021] S203, execute scanning measurement G code, and the result is a plurality of bar data files, recorded as a1, a2, a3...an.

[0022] Furthermore, the step 3 comprises the following steps:

[0023] S301, the CNC machine tool drives the linear laser device to unidirectionally and cyclically scan the entire carbon tube array honeycomb surface according to the scanning trajectory generated by the G code, wherein the xy coordinates correspond to the horizontal position of the carbon tube array honeycomb surface, and the z coordinate represents the height value measured at the position; after the scanning is completed, splicing is performed according to the corresponding relationship of the number of points measured in the longitudinal direction.

[0024] S302, using point cloud filtering to remove burrs and noise, find ridges, and remove noise points within a certain threshold to obtain the complete surface shape of the workpiece surface and store it as point cloud data P, P' = {P'i = (x 2 'i,y 2 'i, z 2 'i)|(i=1, 2, 3,..., Nk')}, k is the number of points.

[0025] Furthermore, step 4 comprises the following steps:

[0026] According to the starting position information of the G code, the measured point cloud is transformed into the new position point cloud t through matrix transformation, and only the translation operation is performed.

[0027]

[0028] Where x, y, z are the three-dimensional coordinates of the measured point cloud p before transformation, T x , T y , T z is the translation distance in the xyz directions.

[0029] Furthermore, the step 5 comprises the following steps:

[0030] Project the new position point cloud onto the surface of the design model to obtain the point set U on the model corresponding to the number of points in the measured point cloud t, U'={U'i=(x 3 'i,y 3 'i, z 3 'i)|(i=1, 2, 3,..., Nk')};

[0031] Specifically, the K-neighborhood search algorithm is applied to search for several corner points closest to the point cloud;

[0032] Project to the nearest corresponding triangle patch in the direction of the centroid, and obtain the point cloud set U,. Use the BasicICP algorithm to accurately align the point cloud t and U to obtain t';

[0033] After registration, a perpendicular line is drawn from the point cloud t' to the nearest patch of the model. The sign of the error relative to the model is calculated based on the following formula. If the point is on one side of the normal line, the value is positive, otherwise it is negative:

[0034]

[0035] In the formula, is the normal direction, d i is the distance from the point cloud to the patch, O is the origin of the coordinate system, Q i p is the foot point of the perpendicular from point p to the patch,

[0036] Use this formula to loop through all points in P and obtain the distance set D, D' = {D'i = (di|(i = 1, 2, 3, ..., Nk')};

[0037] The root mean square value of the surface morphology is calculated from the set D, and the surface shape accuracy analysis is performed based on the root mean square value.

[0038] The present invention generates a measurement G code based on a model; uses a line laser device to measure the carbon tube array honeycomb surface data; performs a rough alignment based on the G code to guide the measurement point cloud coordinate system conversion; and then performs a fine alignment using a Basic ICP algorithm. Compared with the model, the measurement point cloud of the present invention has a high degree of overlap and is applicable to a variety of difficult-to-measure surfaces, providing a new idea for subsequent surface shape detection of various large components. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0040] Figure 1 A flowchart of a method for calculating the surface accuracy of a carbon tube array honeycomb based on machining G code guidance disclosed in an embodiment of the present invention;

[0041] Figure 2 It is a design model example of an embodiment of the present invention;

[0042] Figure 3 A physical diagram of a carbon tube array honeycomb according to an embodiment of the present invention;

[0043] Figure 4 This is the point cloud data diagram of the carbon tube array honeycomb surface after splicing;

[0044] Figure 5 A schematic diagram of a model for calculating surface shape error in a specific embodiment disclosed in the present invention;

[0045] Figure 6 This is a specific example result of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] like Figure 1 As shown, the embodiment of the present invention discloses a method for calculating the accuracy of the honeycomb surface of a carbon tube array based on the guidance of a machining G code, comprising the following steps:

[0048] Step 1, obtain the measurement G code that matches the surface profile of the design model;

[0049] Use as Figure 2 The design model shown in the figure uses UG software to generate G code for measurement that matches the surface contour. Specifically,

[0050] S101. For workpieces with large curvature and known design models, one-way, multi-line, and multi-segment broken line interpolation is used to ensure that the workpiece surface is always kept within the range of the online laser, and the node coordinate information of the G code is obtained as the G code point cloud data to be processed. In the process of programming with UG, the one-way parallel milling method is used to obtain the scanned G code file;

[0051] Extract the coordinate information in the G code and store it as a G code point cloud. The point set is Q'={Q'i=(x 1 'i,y 1 'i, z 1 'i)|(i=1, 2, 3, ..., Nm')}, in this embodiment, MATLAB is used to extract the coordinate information xyz in the G code, which is named Gcode, and its storage format can be a MATLAB data file or an ASCII binary readable text.

[0052] Step 2: Use a line laser device to measure the carbon tube array honeycomb surface data for multiple times. The carbon tube array honeycomb in this embodiment is as follows: Figure 3 As shown, a number of three-dimensional point cloud data are generated;

[0053] S201, use a line laser device that matches a machine tool, and mount the line laser device on a three-axis or five-axis machine tool. In this embodiment, a Gocator line laser device is selected.

[0054] S202. During the scanning process, errors in the installation of the line laser may affect the accuracy of the line laser device, so it is necessary to go through adjustment and correction steps. Specifically, the position of the line laser device and the machine tool is adjusted to calibrate the rotation errors of the line laser in several axes, such as whether the light plane of the line laser is parallel to the measurement feed direction, to ensure that the scanned data can be accurately spliced;

[0055] S203, execute scanning measurement G code, and the result is a plurality of bar data files, recorded as a1, a2, a3...an.

[0056] Step 3: splice and denoise the multiple point cloud data obtained to obtain the complete surface shape of the workpiece surface and store it as measured point cloud data;

[0057] S301, the CNC machine tool drives the linear laser device to unidirectionally scan the entire carbon tube array honeycomb surface according to the scanning trajectory generated by the G code, where the xy coordinates correspond to the horizontal position of the carbon tube array honeycomb surface, and the z coordinate represents the height value measured at the position; after the scanning is completed, the points are spliced ​​according to the corresponding relationship of the longitudinal measurement points, and the spliced ​​surface diagram is shown as follows Figure 4 shown.

[0058] S302, using point cloud filtering to remove burrs and noise, find ridges, and remove noise points within a certain threshold to obtain the complete surface shape of the workpiece surface and store it as point cloud data P, P' = {P'i = (x 2 'i,y 2 'i, z 2 'i)|(i=1, 2, 3,..., Nk')}, k is the number of points.

[0059] Step 4: According to the starting position information of the G code, the measured point cloud data is transformed into a new position point cloud through matrix transformation, and the design model is roughly registered with the new position point cloud;

[0060] As a preferred implementation, the Basic ICP algorithm is used to perform rough alignment based on the G-code guidance; the point cloud is matched with the G-code, the G-code is used as a model replacement, which is a fixed point cloud, and the measured point cloud is a moving point cloud. Basic ICP is used to obtain the new position point cloud t through matrix translation transformation.

[0061] The step 4 comprises the following steps:

[0062] S401, read files, which are G code point cloud Gcode and measurement point cloud P.

[0063] S402, coarse matching, i.e. coordinate matching, obtains the new position point cloud t by matrix transformation according to the starting position information of the G code, without rotation change, only translation operation,

[0064]

[0065] Where x, y, z are the three-dimensional coordinates of the measured point cloud p before transformation, T x , T y , T z is the translation distance in the xyz directions

[0066] Step 5: Project the new position point cloud onto the face of the design model to obtain the model point cloud corresponding to the number of points in the measured point cloud. After registration, calculate the length of the perpendicular line from the point cloud to the corresponding face to obtain the error length set, and calculate the surface error accordingly.

[0067] S501, read the converted point cloud file t, STL file.

[0068] S502, extracting information such as corner point no, normals, and face elements of the STL file.

[0069] S503, perform precise registration, project the new position point cloud onto the surface of the design model STL, and obtain the point set U on the model corresponding to the number of points in the measured point cloud t, U'={U'i=(x 3 'i,y 3 'i, z 3 'i)|(i=1, 2, 3,..., Nk')};

[0070] Specifically, the K-neighborhood search algorithm is applied to search for several corner points closest to the point cloud;

[0071] Project to the nearest corresponding triangle patch in the direction of the centroid, and obtain the point cloud set U,. Use the BasicICP algorithm to accurately align the point cloud t and U to obtain t';

[0072] After registration, a perpendicular line is drawn from the point cloud t' to the nearest patch of the model. To determine whether the error is positive or negative relative to the model, the normal direction of the patch is extracted. If the point is on one side of the normal, the value is positive, otherwise it is negative. Figure 5 As shown, it is calculated based on the following formula:

[0073]

[0074] In the formula, is the normal direction, d i is the distance from the point cloud to the patch, O is the origin of the coordinate system, Qi p is the foot point of the perpendicular from point p to the patch,

[0075] Use this formula to loop through all points in P and obtain the distance set D, D' = {D'i = (di|(i = 1, 2, 3, ..., Nk')};

[0076] The root mean square value of the surface topography is calculated from the set D, and the error length set based on the root mean square value is used to calculate the surface error.

[0077] Execute the embodiment, the calculation results are as follows Figure 6 shown.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for calculating the honeycomb surface shape accuracy of a carbon nanotube array guided by machining G-code, characterized in that, it includes the following steps: Step 1: Obtain the measured G-code that conforms to the surface contour of the design model; Step 2: Use a line laser device to measure the honeycomb surface shape data of the carbon nanotube array multiple times to generate several three-dimensional point cloud data; Step 3: Stitch and denoise the obtained multiple point cloud data to obtain the complete surface shape of the workpiece surface and store it as measured point cloud data; Step 4: According to the starting position information of the G-code, transform the measured point cloud data through matrix transformation to obtain a new position point cloud, and roughly register the design model with the new position point cloud; Step 5: Project the new position point cloud onto the patches of the design model to obtain a model point cloud corresponding to the number of points in the measured point cloud. After registration, calculate the perpendicular length from the point cloud to the corresponding patch to obtain a set of error lengths, and calculate the surface shape error accordingly.

2. The method for calculating the honeycomb surface shape accuracy of a carbon nanotube array guided by machining G-code according to claim 1, characterized in that, in step 1, it includes the steps of generating code and storing code. Specifically: Use UG software to generate the measured G-code. Specifically, for workpieces with large curvature and known design models, use multi-segment broken line interpolation to ensure that the workpiece surface is always within the measuring range of the line laser. During the UG programming process, draw on the unidirectional parallel milling method to obtain the scanned G-code file; Extract the coordinate information in the G code and store it as a G code point cloud. The point set is Q'={Q'i=(x 1 'i,y 1 'i, z 1 'i)|(i=1, 2, 3,..., Nm')}.

3. The method for calculating the honeycomb surface shape accuracy of a carbon nanotube array guided by machining G-code according to claim 2, characterized in that, step 2 includes the following steps: S201: Use a line laser device matching the machine tool and mount the line laser device on a three-axis or five-axis machine tool; S202: Adjust and calibrate the position of the line laser device and the machine tool to calibrate the rotational errors of the line laser in several axial directions to ensure that the scanned data can be accurately stitched; S203: Execute the scanned measurement G-code, and the result is multiple bar data files, denoted as a1, a2, a3... an.

4. The method for calculating the honeycomb surface shape accuracy of a carbon nanotube array guided by machining G-code according to claim 3, characterized in that, step 3 includes the following steps: S301: The numerical control machine tool drives the line laser device to unidirectionally and circularly scan the entire honeycomb surface shape of the carbon nanotube array according to the scanning trajectory generated by the G-code, where the xy coordinates correspond to the horizontal position of the honeycomb surface shape of the carbon nanotube array, and the z coordinate represents the measured height value at that position; After scanning, perform stitching according to the corresponding relationship of the number of points measured longitudinally; S302, using point cloud filtering to remove burrs and noise, find ridges, and remove noise points within a certain threshold to obtain the complete surface shape of the workpiece surface and store it as point cloud data P, P' = {P'i = (x 2 'i,y 2 'i, z 2 'i)|(i=1, 2, 3,..., Nk')}, k is the number of points.

5. The method for calculating the honeycomb surface shape accuracy of a carbon nanotube array guided by machining G-code according to claim 4, characterized in that, step 4 includes the following steps: Transform the measured point cloud according to the starting position information of the G-code to obtain a new position point cloud t, and only perform translation operations, Where x, y, z are the three-dimensional coordinates of the measured point cloud p before transformation, T x , T y , T z is the translation distance in the xyz directions.

6. The method for calculating the honeycomb surface shape accuracy of a carbon nanotube array guided by machining G-code according to claim 5, characterized in that, step 5 includes the following steps: Project the new position point cloud onto the surface of the design model to obtain the point set U on the model corresponding to the number of points in the measured point cloud t, U'={U'i=(x 3 'i,y 3 'i, z 3 'i)|(i=1, 2, 3,..., Nk')}; Specifically, apply the K-nearest neighbor search algorithm to search for several corner points closest to the point cloud; Project to the nearest corresponding triangle patch in the direction of the centroid, and obtain the point cloud set U. Use the BasicICP algorithm to accurately align the point cloud t and U to obtain t'; After registration, a perpendicular line is drawn from the point cloud t' to the nearest patch of the model. The sign of the error relative to the model is calculated based on the following formula. If the point is on one side of the normal line, the value is positive, otherwise it is negative: In the formula, is the normal direction, d i is the distance from the point cloud to the patch, O is the origin of the coordinate system, is the foot point of the perpendicular from point p to the patch, This formula is used to loop through all points in P to obtain the distance set D, D' = {D'i = (d i |(i=1, 2, 3,..., Nk')}; The root mean square value of the surface morphology is calculated from the set D, and the surface shape accuracy analysis is performed based on the root mean square value.

Images