A high-fidelity and fast measurement method for the assembly surface of a motor core stamping die

Abstract

The present invention discloses a high-fidelity and rapid measurement method for the assembly surface of a motor core stamping die; the process is as follows: 1. Collect multiple point cloud coordinate data with different point density. 2. Select the maximum measurement interval and layer interval that meet the surface error measurement accuracy requirements. 3. Based on the measurement interval and layer interval obtained in step 2, further screening is performed using correlation analysis to obtain the maximum measurement interval and optimal layer interval that meet the geometric error modeling point accuracy requirements. The present invention performs a correlation analysis on the radius values ​​of the same points after modeling of small measurement intervals and layer interval data with other large measurement intervals and layer interval data, and selects a strongly correlated measurement interval based on the correlation degree judgment range, indicating that the data obtained by the measurement interval is strongly correlated with the high-precision data measured by the small measurement interval, and finally selects the maximum measurement interval among them, further achieving the dual guarantee of high efficiency and high precision.

Description

Technical Field

[0001] The invention relates to the field of precision measurement of motor core stamping dies for new energy vehicles, and in particular to a high-fidelity and rapid measurement method for an assembly surface of a motor core stamping die. Background Art

[0002] Due to the existence of processing errors, the actual assembly surfaces of the template holes and inserts of the stamping die for the iron core of new energy vehicles have profile errors compared to the ideal assembly surfaces. The existence of profile errors will directly affect the assembly performance of the template holes and inserts, and thus have an important impact on the assembly deformation of the die concave template. At present, in order to achieve high-precision measurement, researchers often use a small measurement interval when measuring the assembly surface. Although this can ensure the measurement accuracy, the measurement efficiency is reduced due to the dense sampling points. Under some working conditions, only a small number of measurement points are taken to increase the measurement speed, thereby sacrificing the measurement accuracy. Therefore, there is an urgent need for a high-fidelity and fast measurement method for the assembly surface of the motor core stamping die to achieve high-precision and high-efficiency measurement of the profile errors of the mold template holes and insert assembly surfaces, thereby providing an accurate model basis for the assembly surface for the subsequent stress deformation analysis of the template hole and insert assembly. Summary of the invention

[0003] In view of the above-mentioned defects of the prior art, the purpose of the present invention is to provide a high-fidelity and rapid measurement method for the assembly surface of a motor core stamping die, which can perform high-precision and high-efficiency measurement of the assembly surface profile errors of the template holes and inserts of the motor core stamping die for new energy vehicles.

[0004] The specific technical solution adopted by the present invention to solve the above technical problems is: a high-fidelity rapid measurement method for the assembly surface of a motor core stamping die, comprising the following steps:

[0005] S100. The template hole and the insert assembly surface are measured with different point density by using the circular line layered point method to obtain a plurality of point cloud coordinate data corresponding to different point density. The parameters corresponding to the point density include the measurement interval and the layer interval.

[0006] S200. Select the best measurement interval L for surface error measurement s , optimal layer spacing P s , the specific process is:

[0007] S210. Multiple point cloud coordinate data A with the same layer interval and different measurement intervals 1 ~A n ; Point cloud coordinate data A 1 ~A n The corresponding measurement intervals increase successively; n is the number of measurement intervals analyzed in this step. 1 ~A n, respectively collect the surface errors as the measurement targets. Take the maximum measurement interval that the surface error meets the accuracy requirements as the optimal measurement interval L for surface error measurement s .

[0008] S220. The measurement interval is the optimal measurement interval L S , multiple point cloud coordinate data B with different layer intervals 1 ~B m ; Point cloud coordinate data B 1 ~B m The corresponding layer intervals increase successively; m is the number of layer intervals analyzed in this step. 1 ~B m , respectively collect the surface errors as the measurement target, and take the maximum layer interval that meets the accuracy requirements of the surface error as the optimal layer interval P for surface error measurement s .

[0009] S300. Select the best measurement interval L for sampling points for geometric error modeling z , optimal layer spacing P z .

[0010] As a preferred method, if the profile error of the template hole and the insert assembly surface needs to be measured, the optimal measurement interval L for profile error measurement is as follows: s , optimal layer spacing P s Points are collected, and the surface errors of the template hole and the insert assembly surface are calculated based on the obtained point cloud data; if three-dimensional modeling of the template hole and the insert assembly surface is required, the optimal measurement interval L of the points collected for geometric error modeling is calculated. z , optimal layer spacing P z Points are collected and three-dimensional models of template holes and inserts are established based on the obtained point cloud data.

[0011] Preferably, in steps S210 and S220, the process of determining whether the surface error calculated by the point cloud coordinate data meets the accuracy requirements is as follows: the surface error corresponding to the point cloud coordinate data with the smallest measurement interval or layer interval is used as the error benchmark; the error rate between the surface error calculated by the remaining point cloud coordinate data and the error benchmark is calculated; if the error rate is less than the permitted threshold, the measurement interval or layer interval corresponding to the error rate meets the accuracy requirements

[0012] Preferably, in step S210, the error rate E i The expression is as follows:

[0013] ;

[0014] Where i=2,...,n. is the profile error used as a reference; is the profile error for calculating the error rate.

[0015] Preferably, the permission threshold in step S210 is 5% to 7%.

[0016] Preferably, in step S200, the surface error as the measurement target is straightness. In steps S210 and S220, for each point cloud coordinate data, the coordinates of the center of each layer of sampling points are fitted, and the straightness between the centers of different layers is calculated. The straightness corresponding to the point cloud coordinate data with the smallest measurement interval or layer interval is used as the benchmark; the error rate between the straightness of the remaining measurement intervals or layer intervals and the straightness used as the benchmark is calculated respectively. Among the measurement intervals and layer intervals whose error rate is less than or equal to the permitted threshold, the largest measurement interval and layer interval are taken as the optimal measurement interval L for measuring the surface error. s and the optimal layer spacing P s .

[0017] Preferably, the process of calculating the straightness is: determining the target straight line based on the coordinates of each circle center corresponding to the same point cloud coordinate data. The target straight line is the straight line with the smallest range of distances to all circle centers. The range of all circle center distances to the target straight line is used as the straightness of the point cloud coordinate data.

[0018] As a preference, in step S100, only multiple point cloud coordinate data with the same layer spacing and different measurement intervals are measured; in step S230, point cloud coordinate data with different layer spacings B 1 ~B m The optimal measurement interval L determined in step S210 S The corresponding point cloud coordinate data is obtained by taking points according to different layer spacings. Point cloud coordinate data B 1 ~B m The interlayer spacing increases successively.

[0019] Preferably, the process of step S300 is as follows:

[0020] S310. Multiple point cloud coordinate data A with the same layer interval and increasing measurement intervals 1 ~A s Perform surface modeling. Point cloud coordinate data A s The measurement interval is the best measurement interval for surface error measurement L s The same number of points are collected on each surface; the distances from the sampling points to the axis of the point cloud coordinate data center are calculated; and the distance groups corresponding to each surface are obtained. R 1 , R 2 , ... , R s. For distance groups R 2 , ... , R s With distance group R 1 Perform correlation analysis to obtain the correlation coefficient; among the measurement intervals with a correlation coefficient greater than the correlation threshold, take the largest measurement interval as the optimal measurement interval for geometric error modeling sampling points L z .

[0021] S320. The measurement interval is the best measurement interval L z , multiple point cloud coordinate data B with increasing layer intervals 1 ~B s Perform surface modeling; point cloud coordinate data B s The layer spacing is the optimal layer spacing for surface error measurement. P s The same number of points are collected on each surface; the distances from the sampling points to the axis of the point cloud coordinate data center are calculated; and the distance groups corresponding to each surface are obtained. R ' 1 , R ' 2 , ... , R ' s . For distance groups R ' 2 , ... , R ' s With distance group R ' 1 Perform correlation analysis to obtain the correlation coefficient; among the layer intervals with correlation coefficients greater than the correlation threshold, take the largest layer interval as the optimal layer interval for geometric error modeling sampling points P z .

[0022] Preferably, in steps S310 and S320, the correlation coefficient used in the correlation analysis is the Pearson correlation coefficient.

[0023] Preferably, the point cloud coordinate data collected in step S100 is subjected to gross error elimination, point filling processing, system error correction and random error filtering.

[0024] As a preferred embodiment, the process of system error correction is:

[0025] (1) Select a gauge block with a nominal length of L.

[0026] (2) Use one end face of the gauge block as the reference surface and place it on the CMM so that the nominal length direction of the gauge block is parallel to the x-axis direction of the CMM coordinate system. Repeat the measurement of a point on the other end face of the gauge block p times. Calculate the x-direction system error based on the x-coordinates x(i) of the p times. .

[0027] (3) Use one end face of the gauge block as the reference surface and place it on the three-dimensional coordinate measuring machine so that the nominal length direction of the gauge block is parallel to the y-axis direction of the three-dimensional coordinate measuring machine coordinate system. Repeat the measurement of a point on the other end face of the gauge block q times. Calculate the y-direction system error based on the y-coordinates y(i) of the q times. .

[0028] (4) According to the system error f x 、f y The point cloud coordinate data is corrected by using the opposite number of

[0029] Preferably, the direction vector of the center axis of the point cloud coordinate data and the coordinates of a point on the axis are obtained by fitting using the least squares method.

[0030] The present invention has the following beneficial effects:

[0031] 1. The present invention samples points on the template hole and the insert assembly surface with different point sampling densities, and uses the point cloud data obtained with the maximum sampling density as a benchmark to analyze the influence of sparse sampling density on the acquisition accuracy of the profile errors of the template hole and the insert assembly surface; thereby obtaining measurement intervals and layer intervals that are as sparse as possible with less influence on the profile error measurement accuracy, thereby improving the measurement efficiency while ensuring the detection accuracy of the profile error of the assembly surface.

[0032] 2. The present invention takes the optimal point density for obtaining the surface error as the lower limit, and further analyzes the influence of the point density on the geometric error modeling; and uses the surface constructed by the point cloud data with the maximum sampling density as the reference surface, and calculates the Pearson correlation coefficient between the modeling surface constructed by the point cloud data obtained with different point densities and the reference surface; thereby obtaining the sparsest measurement interval and layer interval while ensuring the accuracy of assembly surface modeling, and achieving the dual guarantee of high efficiency and high precision in the geometric error modeling of the motor core stamping die assembly surface.

[0033] 3. The present invention realizes the correlation analysis between point cloud data with different sample numbers by constructing a surface from point cloud data, taking points equally and calculating the radius, and then obtains the influence of different sparsity levels of point density on geometric error modeling. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the process of the present invention.

[0035] Figure 2 It is a schematic diagram of the stamping die template of the new energy motor core in the present invention.

[0036] Figure 3 It is a schematic diagram of the insert of the stamping die for the iron core of the new energy motor in the present invention.

[0037] Figure 4 It is a schematic diagram of point selection at different measurement intervals in step S100 of the present invention.

[0038] Figure 5 It is a schematic diagram of using a gauge block to collect system errors in the present invention.

[0039] Figure numerals: 1, large template hole; 2, small template hole; 3, large insert; 4, small insert. DETAILED DESCRIPTION

[0040] The present invention will be further described below in conjunction with the accompanying drawings.

[0041] like Figure 1 As shown in the figure, a high-fidelity and rapid measurement method for the assembly surface of a motor core stamping die is as follows:

[0042] S100. Use the circumferential line layered point method to measure the template holes and insert assembly surfaces of each size range of the new energy vehicle motor core mold with different point density, and separate the obtained point cloud coordinate data.

[0043] The specific process is as follows:

[0044] S110. Divide the design dimensions of all the template holes and shafts on the concave template into multiple size intervals. Take the diameter of the smallest template hole and insert as the lower size limit; take the diameter of the largest template hole and insert as the upper size limit. In this embodiment, the lower size limit and the upper size limit are 8 mm and 136 mm respectively, and 8 size intervals are evenly divided between the lower size limit and the upper size limit. Different size intervals need to select the measurement interval and layer interval respectively.

[0045] In this embodiment, the concave template to be tested is as follows Figure 2 As shown; inserts as Figure 3 As shown, the concave template is provided with a large template hole 1 with a diameter of 136 mm and a small template hole 2 with a diameter of 8 mm. The inserts include a large insert 3 with a diameter of 136 mm and a small insert 4 with a diameter of 8 mm.

[0046] S120. Figure 4As shown, multiple different measurement intervals are set. The template holes and inserts to be measured are selected in the first size interval, and multiple different measurement intervals are set; in this embodiment, the number of measurement intervals set is four; the measurement intervals are 1mm, 2mm, 3mm, and 4mm from small to large. Since the number of points is dense when the measurement interval is 1mm, it can better express the assembly surface error information. Therefore, 1mm is used as the minimum measurement interval, and the surface errors of the data measured at other measurement intervals are compared. The maximum interval with a surface error corresponding to the 1mm measurement interval that is within 7% is selected as the optimal measurement interval, which improves the measurement efficiency while ensuring the accuracy of the surface error calculation.

[0047] S130. Measure the template holes and inserts that match each other according to different measurement intervals. In each measurement, the template holes and insert assembly surfaces of the new energy vehicle motor core mold are measured using the circular line layered point method to obtain point cloud coordinate data. Among them, the measuring instrument used is a high-precision three-dimensional coordinate measuring machine, the measurement point scheme is the circular line layered point method, the point cloud measurement interval is L, the number of measurement layers N is 11, and the layer interval P is 2mm; the point cloud measurement interval represents the distance between two adjacent sampling points in the same layer; finally, the point cloud coordinate data A of the assembly surface corresponding to four different measurement intervals are measured. 1 , A 2 , A 3 , A 4 .

[0048] S140. Separate the point cloud coordinate data obtained in step S130. The separation process includes gross error elimination, point filling, system error correction and random error filtering. The specific steps of S140 include:

[0049] S141. Use Laida's law to eliminate gross errors in point cloud coordinate data.

[0050] (1) Import the point cloud coordinate data into Matlab software through the readtable function.

[0051] (2) Use the least squares method to perform a preliminary fit on the point cloud coordinate data to obtain the direction vector (dx 00 ,dy 00 ,dz 00 ) and the coordinates of a point on the axis (x 00 ,y 00 , z 00 ).

[0052] The process of fitting the axis by the least squares method is:

[0053] ①. Assume the center point Q of an axis 0 (x 0 ,y0 , z 0 )、Direction vector M 0 (dx 0 ,dy 0 ,dz 0 ), calculate the distance from each data point to the axis .

[0054] d(i)= x i - x 0 2 + y i - y 0 2 + z i - z 0 2 -[ x i - x 0 dx 0 + y i - y 0 dy 0 + z i - z 0 dz 0 ] dx 0 2 + dy 0 2 + dz 0 2

[0055] in,( , ) is any data point in the point cloud coordinate data.

[0056] ②. Calculate distance The square of the difference between the ideal radius r of the template hole or insert being measured and the sum .

[0057] l(i)= [d(i)-r] 2

[0058]

[0059] Where n is the number of data points in the point cloud coordinate data.

[0060] ③. Use the optimization function fmincon in Matlab software to optimize the sum l and obtain the optimal fitting axis center point Q 00 (x 00 ,y 00 , z 00 )、Direction vector M 00 (dx 00 ,dy 00 ,dz 00 ).

[0061] (3) Calculate the distance from all point cloud coordinate data to the fitting axis , calculate the average of all distances and standard deviation .

[0062]

[0063]

[0064]

[0065] (4) If the distance corresponding to any point in the point cloud coordinate data , and the average distance The difference is greater than , then the point is considered to be a gross error and is removed.

[0066] S142 performs point filling processing on the point cloud coordinate data after eliminating the gross errors. The specific implementation steps include:

[0067] (1) First, the point cloud coordinate data is layered, and then points are added layer by layer.

[0068] (2) Start filling in points from the first layer. First, fit the first layer of point cloud coordinate data using the least squares method to obtain the coordinates of the center of the fitting circle. Then calculate the distance D and angle O from each point cloud coordinate data to the center of the circle, and take the average value of the angle. .

[0069] (3) Starting from the first point, traverse the angle difference between two adjacent points ,like < < , then for this angle difference Fill in the points between the two corresponding points (denoted as point U and point V).

[0070] (4) Calculate the number of points to be filled between two adjacent points .

[0071]

[0072] in, This is a round-up operation.

[0073] (5) Fill in the points. Insert evenly between points U and V. The radius of the filling point is determined by the average radius of the U and V points.

[0074] (6) Repeat the above steps to complete each layer of point filling in turn.

[0075] S143 performs system error correction on the point cloud coordinate data of the completed points.

[0076] (1) If Figure 5 As shown, select a grade 0 gauge block with a nominal length of 50 mm.

[0077] (2) Use one high-precision surface of the gauge block as the reference surface and place it on a three-dimensional coordinate measuring machine so that the nominal length direction of the gauge block is parallel to the x-axis direction of the coordinate measuring machine coordinate system. Randomly measure a point on the other high-precision surface of the gauge block 30 times. The measured x-coordinate is marked as x(i); i = 1, 2, ..., 30.

[0078] (3) The difference between the average value of the infinite number of measurement results of the same measured quantity under repeatability conditions and the true value of the measured quantity is taken as the systematic error f; calculate the x-direction systematic error f x .

[0079] -50

[0080] (4) Use one high-precision surface of the gauge block as the reference surface and place it on a three-dimensional coordinate measuring machine so that the nominal length direction of the gauge block is parallel to the y-axis direction of the coordinate measuring machine coordinate system. Randomly measure a point on the other high-precision surface of the gauge block 30 times. The measured y-coordinate is marked as y(i).

[0081] (5) Calculate the system error f in the y direction y .

[0082] -50

[0083] (6) According to the calculated system error f x 、f y The point cloud coordinate data is corrected by using the opposite number of

[0084] S144. Perform random error filtering on the point cloud coordinate data that has completed system error correction.

[0085] (1) Based on the 30 repeated measurements of the same point in step S143, calculate the random error W x (i) W y (i) Then the spatial distribution of random errors is analyzed.

[0086]

[0087]

[0088] (2) Select an appropriate filtering method based on the spatial distribution of random errors to filter out random errors in the point cloud coordinates.

[0089] S200 performs surface error analysis on the point cloud coordinate data obtained by measuring different measurement intervals and different layer intervals, and obtains the maximum measurement interval L that can ensure the surface error accuracy while improving the measurement efficiency. s , layer spacing P s The specific process is as follows:

[0090] S210 performs surface error analysis on the point cloud coordinate data corresponding to different measurement intervals to obtain the maximum measurement interval L that can ensure the surface error accuracy while improving the measurement efficiency. s The specific implementation process is as follows:

[0091] (1) Analyze and calculate the profile error according to the point cloud coordinate data corresponding to different measurement intervals obtained in step S100. In this embodiment, straightness is selected as the profile error to be analyzed and calculated.

[0092] The straightness calculation method is as follows:

[0093] The least squares method is used to fit each layer of point cloud data to obtain the center of the circle, and then the target line is determined based on the center of the circle corresponding to each layer of point cloud. The target line is the line with the smallest range of distances to all the center of the circle (i.e., the difference between the maximum distance and the minimum distance). Specifically, assuming a straight line, let the x, y, and z coordinates of a point Q on the straight line (x 0 ,y 0 , z 0 ) are the average values ​​of the x, y, and z coordinates of all the circle centers. , the straight line direction vector M is (0, 0, 1), and the optimization object is optimized to obtain the target straight line. The formula for the optimization object is:

[0094] All distances from the center of the circle to the straight line:

[0095]

[0096] Optimization object: f=max(D)-min(D), that is, find the maximum value max(D) and the minimum value min(D) of the distance from all points to the straight line, and then use the fmincon optimization function to optimize f, and finally get the optimized straight line, and f is the straightness.

[0097] (2) Straightness S calculated from the point cloud coordinate data corresponding to a 1 mm measurement interval 1 As a benchmark, calculate the straightness S measured at 2mm, 3mm, and 4mm measurement intervals respectively. 2 ~S 4 ; Calculate the straightness S 2 ~S 4 With straightness S 1 The error rate E 2 、E 3 、E 4 as follows:

[0098]

[0099] Among them, i=2,3,4.

[0100] At the error rate E i Among the measurement intervals less than or equal to 7%, the largest measurement interval is taken as the optimal measurement interval L for measuring the surface error. s .

[0101] S220. Perform surface error analysis on the point cloud coordinate data corresponding to different layer intervals to obtain the maximum layer interval P that can ensure the surface error accuracy while improving the measurement efficiency. s .

[0102] S221. Take the maximum measurement interval L obtained in S210 s The corresponding point cloud coordinate data are extracted from 11 layers of data with a layer interval of 2 mm, which are recorded as B 1 , extract 6 layers of data with a layer interval of 4 mm and record it as B 2 , extract 4 layers of data with a layer interval of 6 mm and record it as B 3 , take the 3-layer data with a layer interval of 8 mm and record it as B 4 .

[0103] S222. Based on point cloud coordinate data B 1 , B 2 , B 3 , B 4 , analyze and calculate the profile error. The calculation method is the same as step S210. Then, the point cloud coordinate data B corresponding to the 2mm layer interval is 1 Calculated straightness S' 1 As a benchmark, calculate the straightness S' measured at 4mm, 6mm, and 8mm measurement intervals respectively. 2 ~S' 4 ;

[0104] Calculate the straightness S' 2 ~S' 4 and straightness S' 1 The error rate E' 2 , E' 3 , E' 4 as follows:

[0105]

[0106] Among them, i=2,3,4.

[0107] At the error rate E' i Among the layer intervals less than or equal to 7%, the largest layer interval is taken as the optimal layer interval P for measuring the profile error. s .

[0108] S300. Perform correlation analysis on the assembly surface point cloud coordinate data obtained by measuring different measurement intervals and different layer intervals, and obtain the optimal measurement interval L that can improve measurement efficiency while ensuring the accuracy of geometric error modeling. z , layer spacing P z .

[0109] S310. Perform correlation analysis on the assembly surface point cloud coordinate data obtained by measuring at different measurement intervals to obtain the optimal measurement interval L that can improve measurement efficiency while ensuring the accuracy of geometric error modeling. z .

[0110] S311. Take the measurement intervals obtained in step S100 from small to large, 1 mm, ..., L s The corresponding point cloud coordinate data.

[0111] S312. Perform surface modeling on the point cloud coordinate data corresponding to different measurement intervals according to the NURBS surface modeling method.

[0112] S313. Uniformly interpolate points from the surfaces corresponding to different measurement intervals, and the number and position of points in each surface are the same; the same points specifically refer to the same azimuth and height of the corresponding sampling points in different surfaces in the cylindrical coordinate system; then fit the axis, and the axis fitting method is as shown in steps (1) and (2) of step S141; finally, calculate the distance from each surface point to the fitting axis, which is recorded as distance group R. Among them, the distance calculation method is as shown in step (3) of step S141, and the distance groups obtained after executing each point cloud coordinate data are respectively distance groups R 1 ,R 2 ,...,R s s is the number of point cloud coordinate data selected in step S311. 1 ,R 2 ,...,R s The number of distances in are equal.

[0113] S314. Calculate and evaluate the Pearson correlation coefficient to obtain the optimal measurement interval L z .

[0114] When calculating the Pearson correlation coefficient, the distance group R corresponding to the data is measured at the minimum measurement interval (1mm). 1 The data in is variable a, and then the distance group R corresponding to the data is measured at other measurement intervals. 2 ,……,R s The data in is variable b. The calculation formula of Pearson correlation coefficient is

[0115] The calculated correlation coefficients are recorded as 2 ,……, s .

[0116] In order to obtain point cloud coordinate data with strong correlation to the accuracy of point cloud coordinate data measured at the minimum measurement interval, the correlation coefficient is selected according to the Pearson correlation coefficient evaluation standard. 2 ,……, s For values ​​greater than 0.6, the corresponding point cloud coordinate data is strongly correlated with the accuracy of the point cloud coordinate data measured at the minimum measurement interval. On this basis, the corresponding maximum measurement interval L is selected. z .

[0117] S320. Perform correlation analysis on the assembly surface point cloud coordinate data obtained by measuring different layer intervals to obtain the optimal layer interval P that can improve measurement efficiency while ensuring the accuracy of geometric error modeling. z .

[0118] S321. Take the best measurement interval L obtained in S310 z The corresponding point cloud coordinate data are taken as layer interval P 1 The data is recorded as , take the layer interval as P 2 The data is recorded as , ..., take the layer interval as P s The data is recorded as .

[0119] S322. Perform surface modeling on the point cloud coordinate data corresponding to different layer intervals according to the NURBS surface modeling method.

[0120] S323. Uniformly interpolate points from the surfaces corresponding to different layer intervals, and the number and position of points in each surface are the same; then fit the axis, and the axis fitting method is as shown in step (1) (2) of step S141; finally, calculate the distance from each surface point to the fitting axis, which is recorded as the distance group The distance calculation method is as shown in step (3) of step S141. The distance groups obtained after executing each point cloud coordinate data are respectively: , ,……, .

[0121] S324. Calculate and evaluate the Pearson correlation coefficient to obtain the optimal layer interval P z .

[0122] When calculating the Pearson correlation coefficient, the distance group corresponding to the data was measured with the minimum layer interval (0.5 mm). The data in is variable a, and then the distance groups corresponding to the data are measured at other layer intervals. ,……, The data in is variable b. The calculation formula of Pearson correlation coefficient is

[0123] The calculated correlation coefficients are recorded as 2,……, s .

[0124] In order to obtain point cloud coordinate data with strong correlation with the accuracy of point cloud coordinate data measured by the minimum layer interval, the correlation coefficient is selected according to the Pearson correlation coefficient evaluation standard. 2 ,……, s For values ​​greater than 0.6, the corresponding point cloud coordinate data is strongly correlated with the accuracy of the point cloud coordinate data measured by the minimum layer interval. On this basis, the corresponding maximum layer interval P is selected. z .

[0125] The following is the execution data of this embodiment:

[0126] This data is the execution data of the 12mm template hole in the first size interval.

[0127] Measuring interval / mm 1 2 3 4 Surface error (straightness) 0.000484 0.000462 0.000507 0.000525 Percent change (relative to 0.000484) / 4.5454547% 4.7520661% 8.4710744%

[0128] Therefore, the maximum measurement interval L corresponding to the linearity change of less than 7% relative to the minimum measurement interval (1mm) is selected. s , that is 3mm.

[0129] Layer interval / mm 2 4 6 8 Surface error (straightness) 0.000507 0.000502 0.0003 0.000453 Percent change (relative to 0.001431) / 1.042600% 40.862111% 10.701788%

[0130] Therefore, based on the 4 mm measurement interval, the maximum layer interval P corresponding to the linearity change of less than 7% relative to the minimum layer interval (2 mm) is selected. s , that is 4mm.

[0131] Measuring interval / mm Correlation coefficient 1 and 2 0.694 1 and 3 0.003

[0132] Therefore, the measurement interval L with a correlation coefficient greater than 0.6 with the surface accuracy of the 1mm measurement interval is selected. z , that is, take a 2mm measurement interval.

[0133] Layer interval / mm Correlation coefficient 2 and 4 0.946

[0134] Therefore, based on the 2 mm measurement interval, the layer interval P with a correlation coefficient greater than 0.6 with the surface accuracy of the 2 mm layer interval is selected. z , that is, take 4mm layer interval.

[0135] In summary, selecting a 2mm measurement interval and a 4mm layer interval to measure the template holes in this size range can improve the measurement efficiency while ensuring the surface error accuracy and measurement accuracy.

Claims

1. A high-fidelity and rapid measurement method for the assembly surface of a motor core stamping die, characterized in that: The following steps are involved: S100. The template hole and the insert assembly surface are measured with different point density by the circular line layered point method to obtain multiple point cloud coordinate data corresponding to different point density; the parameters corresponding to the point density include the measurement interval and the layer interval; S200. Select the best measurement interval L for surface error measurement s , optimal layer spacing P s , the specific process is: S210. Multiple point cloud coordinate data A1 to A2 with the same layer interval and different measurement intervals n ; For point cloud coordinate data A1~A n , respectively collect the surface errors as the measurement targets; take the maximum measurement interval that the surface error meets the accuracy requirements as the optimal measurement interval L for surface error measurement s ; S220. The measurement interval is the optimal measurement interval L S , multiple point cloud coordinate data B1~B with different layer intervals m ; For point cloud coordinate data B1~B m , respectively collect the surface errors as the measurement target, and take the maximum layer interval that meets the accuracy requirements of the surface error as the optimal layer interval P for surface error measurement s ; In step S200, the surface error as the measurement target is straightness; in steps S210 and S220, for each point cloud coordinate data, the coordinates of the center of each layer of sampling points are fitted, and the straightness between the centers of different layers is calculated; the straightness corresponding to the point cloud coordinate data with the smallest measurement interval or layer interval is used as the benchmark; the error rate between the straightness of the remaining measurement intervals or layer intervals and the straightness used as the benchmark is calculated respectively; among the measurement intervals and layer intervals whose error rates are less than or equal to the permitted threshold, the largest measurement interval and layer interval are taken as the optimal measurement interval L for measuring the surface error. s and the optimal layer spacing P s ; In steps S210 and S220, the process of judging whether the profile error calculated by the point cloud coordinate data meets the accuracy requirement is as follows: the profile error corresponding to the point cloud coordinate data with the smallest measurement interval or layer interval is used as the error reference; Calculate the error rate between the surface error calculated by the remaining point cloud coordinate data and the error benchmark; If the error rate is less than the permissible threshold, the measurement interval or layer interval corresponding to the error rate meets the accuracy requirement; S300. Select the best measurement interval L for sampling points for geometric error modeling z , optimal layer spacing P z ; The process of step S300 is as follows: S310. Multiple point cloud coordinate data A1 to A1 with the same layer interval and increasing measurement intervals s Perform surface modeling; point cloud coordinate data A s The measurement interval is the best measurement interval for surface error measurement L s ; Sampling the same number of points on each surface; calculating the distances from the sampling points to the axis of the point cloud coordinate data center; obtaining the distance groups corresponding to each surface R 1, R 2, ... , R s ; For distance groups R 2, ... , R s With distance group R 1. Perform correlation analysis to obtain the correlation coefficient; among the measurement intervals with a correlation coefficient greater than the correlation threshold, take the largest measurement interval as the optimal measurement interval for geometric error modeling sampling points. L z ; S320. The measurement interval is the best measurement interval L z , multiple point cloud coordinate data B1~B with increasing layer intervals s Perform surface modeling; point cloud coordinate data B s The layer spacing is the optimal layer spacing for surface error measurement. P s ; Sampling the same number of points on each surface; calculating the distances from the sampling points to the axis of the point cloud coordinate data center; obtaining the distance groups corresponding to each surface R '1, R '2, ... , R ' s ; For distance groups R '2, ... , R ' s With distance group R '1 Carry out correlation analysis to obtain the correlation coefficient; among the layer intervals with correlation coefficients greater than the correlation threshold, take the largest layer interval as the optimal layer interval for geometric error modeling sampling points P z .

2. A high-fidelity rapid measurement method for the assembly surface of a motor core stamping die according to claim 1, characterized in that: If you need to measure the profile error of the template hole and the insert assembly surface, the optimal measurement interval L for profile error measurement is s , optimal layer spacing P s Points are collected, and the surface errors of the template hole and the insert assembly surface are calculated based on the obtained point cloud data; if three-dimensional modeling of the template hole and the insert assembly surface is required, the optimal measurement interval L of the points collected for geometric error modeling is calculated. z , optimal layer spacing P z Points are collected and three-dimensional models of template holes and inserts are established based on the obtained point cloud data.

3. The high-fidelity rapid measurement method for the assembly surface of a motor core stamping die according to claim 1 is characterized in that: In step S210, the error rate E i The expression is as follows ； Where i=2,...,n; is the profile error used as a reference; is the profile error for calculating the error rate.

4. The high-fidelity rapid measurement method for the assembly surface of a motor core stamping die according to claim 1 is characterized in that: The permission threshold in step S210 is 5% to 7%.

5. The high-fidelity rapid measurement method for the assembly surface of a motor core stamping die according to claim 1, characterized in that: The process of calculating straightness is as follows: determine the target straight line based on the coordinates of each circle center corresponding to the same point cloud coordinate data; the target straight line is the straight line with the smallest range of distances to all circle centers; the range of all circle center distances to the target straight line is used as the straightness of the point cloud coordinate data.

6. The high-fidelity rapid measurement method for the assembly surface of a motor core stamping die according to claim 1, characterized in that: In step S100, only the same layer interval is measured, and multiple point cloud coordinate data with different measurement intervals are measured; in step S230, the point cloud coordinate data B1 to B1 with different layer intervals are measured. m The optimal measurement interval L determined in step S210 S The corresponding point cloud coordinate data is obtained by taking points according to different layer spacings.

7. The high-fidelity rapid measurement method for the assembly surface of a motor core stamping die according to claim 1, characterized in that: The point cloud coordinate data collected in step S100 is subjected to gross error elimination, point filling processing, system error correction and random error filtering; The process of system error correction is: (1) Select a gauge block with a nominal length of L; (2) Use one end face of the gauge block as the reference surface and place it on the three-dimensional coordinate measuring machine so that the nominal length direction of the gauge block is parallel to the x-axis direction of the coordinate measuring machine coordinate system. Repeat the measurement of a point on the other end face of the gauge block p times; calculate the x-direction system error based on the x-coordinates x(i) of p times. ; (3) Use one end face of the gauge block as the reference surface and place it on the three-dimensional coordinate measuring machine so that the nominal length direction of the gauge block is parallel to the y-axis direction of the three-dimensional coordinate measuring machine coordinate system. Repeat the measurement of a point on the other end face of the gauge block q times; calculate the y-direction system error based on the y-coordinates y(i) of the q times. ; (4) According to the system error f x 、f y The point cloud coordinate data is corrected by using the opposite number of

Images