Gear tooth profile measurement result processing method based on robust Gaussian regression filter

By using a robust Gaussian regression filter to process tooth profile measurement data in gear surface measurement, the problem of intimate integration of filtering methods and measurement methods in the prior art is solved, and the accuracy and robustness of gear tooth profile measurement results are improved. The edge effect and outlier influence are achieved, and a more efficient filtering effect is achieved.

CN119003972APending Publication Date: 2024-11-22BEIJING UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411184807.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In the existing gear surface measurement technology, the filtering method and the measurement method are not closely combined, resulting in low accuracy of the gear tooth profile measurement results, and there is a boundary effect and low robustness, which affects the assessment accuracy.

Method used

Using a method based on a robust Gaussian regression filter, the edge effect and outlier value are solved by calculating the relative error value of the gear tooth profile and performing Gaussian filtering. Combining the robust weight function and Gaussian weight function, a robust filtering model is established, and the filtering results iteratively calculates the filtering results, solving the edge effect and outlier value. Impact issues.

Benefits of technology

It improves the accuracy and efficiency of gear tooth profile measurement results, enhances the robust filtering ability of abnormal data, overcomes the edge effect and low robustness problems of traditional filtering methods, and improves the consistency of measurement results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119003972A_ABST
    Figure CN119003972A_ABST
Patent Text Reader

Abstract

The invention discloses a gear tooth profile measurement result processing method based on robust Gaussian regression filtering. The gear tooth profile measurement result processing method comprises the steps that a Klinberg P26 model gear measurement center is adopted to measure a gear; calculating a relative error value of a gear tooth profile measurement result relative to an actual surface according to a generalized polar coordinate measurement principle for bottom layer data recorded by a gear measurement center; converting the relative error value to a normal direction along the involute of the gear and a direction vertical to the involute of the gear to obtain a curve of a tooth profile measurement result of the gear; establishing a robust Gaussian regression filter according to nominal parameters of the tooth surface of the gear to obtain a tooth profile deviation curve for tooth profile deviation evaluation; and for the tooth profile deviation curve obtained after filtering and used for tooth profile deviation evaluation, tooth profile deviation calculation is carried out on the measurement result of the gear tooth profile. According to the method, the accuracy of a gear measurement result curve can be improved, and the limitation that a traditional filtering method generates an end effect and is poor in robustness is overcome.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of precision instruments and measurement technology, and in particular relates to a gear measurement result processing method based on a robust Gaussian regression filtering method. Background Art

[0002] Surface measurement technology, surface assessment theory and methods have become an important technical branch in the field of precision measurement. Filtering technology is an important part of surface measurement. Different filtering methods will have a great impact on the accuracy and precision of surface measurement results. Therefore, filtering accuracy directly affects the quality and results of surface profile assessment.

[0003] GPS is the abbreviation of "Geometrical Product Specification and Verification" by ISO / TC 213 of the International Organization for Standardization, which refers to the technical specifications and certification of geometric products. It is the most influential standard and metrological information system in the world. This series proposes a series of specifications for geometric surface morphology, including: the definition, classification, extraction method, calculation and evaluation method of surface morphology. The measurement and evaluation methods of the surface should be carried out in accordance with this series of international standards proposed by the International Organization for Standardization ISO / TC.

[0004] In 2010, the second-order robust Gaussian regression filter was included in the international standard ISO16610-31. The principle of the second-order robust Gaussian regression filter is to apply the Tukey biweighted estimation function in the M-estimation to the Gaussian regression model to establish a robust Gaussian regression filter. (ISO / TS16610-31:2010Geometrical Product Specifications(GPS)-Filtration-Part21:RobustProfile Filters:Gaussian Regression Filters.)

[0005] Gears are widely used as key components for transmitting motion and power. In order to smoothly transmit motion and power, the gear tooth profile is designed to be an involute shape. Except for a small number of gears with special transmission requirements or designed to reduce gear damage such as gear root cutting and pitting, the tooth surface of most gears is an involute surface.

[0006] However, in the existing gear surface measurement technology, the filtering method used is not closely integrated with the gear measurement method. The filtering method used often cannot effectively and accurately filter the gear tooth profile measurement data with abnormal value signals, and has boundary effects. The robustness of the filter is low, which in turn has a disadvantageous impact on the accuracy of the gear measurement result evaluation. Summary of the invention

[0007] The present invention provides a method for processing gear tooth profile measurement results based on robust Gaussian regression filtering, which solves the problems that the existing gear surface measurement method is not closely integrated with the surface filtering and evaluation method, the results of different surface filtering methods are different, and the gear tooth profile detection accuracy is low. The method can improve the accuracy and efficiency of the evaluation of the gear tooth surface profile measurement results and improve the intuitiveness of gear evaluation.

[0008] To achieve the above objectives, the present invention provides the following technical solutions:

[0009] A method for processing gear tooth profile measurement results based on a robust Gaussian regression filter, comprising:

[0010] Use Klingelnberg P26 gear measuring center to measure the involute gear to be measured, so as to obtain the measurement data record of the gear measuring center;

[0011] Furthermore, the circular grating reading, X-axis grating reading, Y-axis grating reading, Z-axis grating reading, X-axis direction probe value, Y-axis direction probe value, and Z-axis direction probe value are extracted from the underlying data records of the gear measurement center.

[0012] According to the measurement principle of the gear measurement center - the generalized polar coordinate method, the relative error value of the gear tooth profile on the measurement trajectory is calculated using the extracted underlying data of the gear measurement center, including;

[0013] Calculate the systematic errors generated during gear profile measurement, including:

[0014] Errors caused by circular grating rotation readings, errors caused by X-axis grating readings, errors caused by Y-axis grating readings, and errors caused by Z-axis grating readings.

[0015] Furthermore, the calculated system error in the measurement process is synthesized in the measurement direction and the error occurrence direction to obtain a final measurement error value;

[0016] Furthermore, the measured value obtained by measuring the gear tooth profile is subtracted from the calculated measurement error value, and the result obtained is the relative error value of the gear tooth profile.

[0017] The calculated relative error value of the gear tooth profile is converted to the normal direction along the gear involute, and the profile curve of the gear tooth profile measurement result is constructed, including:

[0018] Calculate the base circle radius of the gear and generate the measurement line in the normal direction of the gear involute;

[0019] Furthermore, the length l of the trajectory line during the measurement process is calculated according to the gear parameters and the measurement position;

[0020] Read the data dimension and calculate the minimum interval dx of the measured data;

[0021] Establish a rectangular coordinate system in the normal direction along the involute of the gear and perpendicular to the direction;

[0022] Furthermore, in the established coordinate system, the relative error value of the gear tooth profile is characterized.

[0023] Establishing a robust Gaussian regression filter, and performing Gaussian filtering on the relative error value curve of the gear measurement result according to the established robust Gaussian regression filter to obtain a tooth profile curve, including:

[0024] Introducing robust weight function When |v l |≤c,l=1,...,n:

[0025] Where: v is the deviation of the tooth profile data points before and after filtering, that is, the difference between the tooth profile data after filtering and the tooth profile data before filtering, using a digital representation method;

[0026] n is the data dimension, that is, the data dimension of the digital expression of the tooth profile data before and after filtering;

[0027] c=4.4478*median|v|, and median means taking the median value, that is, taking the median value of the deviation of the results before and after filtering as the calculation threshold;

[0028] Further, according to the robust weight function, according to the Gaussian weight function Establish the weight function matrix S k ,

[0029] Where Δx represents the actual interval between the numerical points of the digitized tooth profile data, that is, the sampling interval of the gear measurement center when measuring the tooth profile; x l,k =(lk)*Δx represents the conjugate distance between each point of all numerical points of the digitized tooth profile data and other data points; s n,k is a robust Gaussian weight function, λ c is the cut-off wavelength, and the Gaussian weight function g(x) is at the cut-off wavelength λc The curve amplitude passing rate at is 50%, and the parameter γ≈0.7309 is calculated by inverse Fourier transform;

[0030] Furthermore, according to the weight function matrix S k , establish the robust Gaussian regression filter equation

[0031] Where Z is the initial curve matrix, that is, a one-dimensional matrix composed of the digitized data of the tooth profile measurement results. is the conjugate distance x between the numerical points of the tooth profile l,k Constructed coefficient regression matrix;

[0032] Furthermore, by solving the filtering equation, the deviation curve of the tooth profile measurement result after filtering using the robust Gaussian regression filtering method is obtained, including:

[0033] Determine the initial value w(1) of the robust weight function iteration using the robust filtering model established in step 3) of claim 6, and calculate the deviation value v(1);

[0034] Using the robust weight function obtained in step 1) of claim 6 and the weight function matrix obtained in step 2) of claim 6, iteratively calculate the Gaussian filter centerline w(i);

[0035] Use the calculated Gaussian filter center line w(i) as the original curve z(i+1) for the next calculation, iteratively calculate the filter result w(i+1) until the deviation value v(i+1) meets the requirements, and take w(i+1) as the final robust Gaussian filter result.

[0036] According to the tooth profile measurement result deviation curve obtained by Gaussian filtering, different types of tooth profile deviation value calculations are performed on the result curve for evaluation, including:

[0037] According to the national standard GB / T 10095.1-2022, the gear profile deviation is calculated based on the deviation curve of the gear profile measurement result;

[0038] F a The total tooth profile deviation is the distance between two parallel lines of the designed tooth profile that include the measured tooth profile within the tooth profile measurement range;

[0039] f fa The tooth profile shape deviation is the distance between two parallel lines of the average tooth profile that contain the measured tooth profile within the tooth profile gauge value range;

[0040] f Ha The tooth profile inclination deviation is the distance between two parallel lines of the designed tooth profile that intersect the intersection of the extension line of the average tooth profile line and the tooth top circle diameter da within the tooth profile measurement range.

[0041] The present invention has the following characteristics and beneficial effects:

[0042] 1. The robust Gaussian regression filtering method used in the present invention can completely overcome the edge effect problem of missing data within the cutoff wavelength of the first and last sections of the profile to be processed due to the inherent characteristics of the convolution operation in the traditional filtering method.

[0042] 2. The robust Gaussian regression filtering method used in the present invention has extremely high robustness and ability to resist extreme deviations for the surface signals of peaks and valleys on the measured tooth surface, such as scratches, burrs, pits and other extreme outliers. This method uses the theory of robust statistics to solve the problem that the traditional filtering method is easily affected by outliers, resulting in the filtering results deviating from the real surface and low confidence, which in turn affects the assessment precision and accuracy.

[0043] 3. The present invention combines the robust Gaussian regression filtering method with the gear tooth profile measurement and evaluation process, which can particularly solve the problem that the filtering method and the gear measurement method are not closely integrated in the current gear measurement and the consistency of the filtering results is poor, and can improve the accuracy of the gear tooth profile measurement results.

[0044] The invention is particularly suitable for the field of precision measurement, and can particularly solve the problem of eliminating the edge effect of the profile and filtering the robustness of abnormal data in the measurement and evaluation of the tooth profile of an involute gear. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the specific embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below.

[0046] Figure 1 This is a schematic diagram of the gear tooth profile measurement result processing method proposed in the present invention.

[0047] Figure 2 A diagram of the underlying data log of a gear measurement center provided for an example of the present invention.

[0048] Figure 3 Schematic diagram of the gear measurement center measurement principle provided as an example of the present invention.

[0049] Figure 4 A gear measurement process trajectory diagram provided for an example of the present invention.

[0050] Figure 5 The original profile curve diagram of the gear tooth profile measurement results provided for the example of the present invention.

[0051] Figure 6 This is a graph of the gear tooth profile measurement results after filtering provided by the example of the present invention.

[0052] Figure 7A curve chart comparing the gear tooth profile measurement results before and after filtering provided by the example of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0054] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0055] The present invention mainly aims at the problem that the combination of the filtering method and the gear measurement method in the current gear measurement is not tight enough, the consistency of the filtering results is poor, and the problem of balancing the elimination of the edge effect of the profile and the robust filtering of abnormal data in the gear tooth profile measurement and evaluation. A method for processing the gear tooth profile measurement results based on robust Gaussian regression filtering is provided, the underlying data of the measurement results are obtained, the tooth profile deviation value is calculated according to the generalized polar coordinate measurement principle adopted by the gear measurement center, a coordinate system is established in the gear measurement direction, the measurement result data coordinate system is converted to form a gear tooth profile deviation value curve, a robust Gaussian regression filter is established and Gaussian filtering is performed on the tooth profile relative deviation value curve, and finally the filter result is subjected to tooth profile deviation evaluation processing. The method of the present invention can improve the accuracy of the gear tooth profile measurement results, and solve the problem of balancing the elimination of the edge effect of the profile and the robust filtering of abnormal data in the involute gear tooth profile measurement and evaluation.

[0056] like Figure 1 As shown, the present invention proposes a method for detecting cigarette ash performance based on a novel robust Gaussian filter, and the specific steps include:

[0057] S1: Use Klingelnberg P26 gear measuring center to measure the involute gear to be measured, and obtain the measurement data record of the gear measuring center;

[0058] S2: Calculate the relative error value of the gear tooth profile measurement result relative to the actual surface;

[0059] S3: coordinate transformation to the normal direction along the gear involute and perpendicular to the direction to obtain the curve of the gear tooth profile measurement result;

[0060] S4: Establish a robust Gaussian regression filter to process the tooth profile measurement results and obtain a tooth profile deviation curve for tooth profile deviation evaluation;

[0061] S5: Calculate the tooth profile deviation, including: total tooth profile deviation, tooth profile shape deviation, and tooth profile inclination deviation.

[0062] Specifically, the gear measurement center measures the gear tooth profile by linking the machine tool X-axis with the turntable C-axis. Taking the right tooth surface as an example, the turntable rotates counterclockwise by θ radians while the slide moves L distance along the negative direction of the X-axis, and L=rb*θ is satisfied. At the same time, the probe can swing along the positive or negative direction of the X-axis and record the swing amount information. The results are as follows: Figure 3 shown.

[0063] Furthermore, from the underlying data records of the gear measurement center, the circular grating reading, X-axis grating reading, Y-axis grating reading, Z-axis grating reading, X-axis direction probe value, Y-axis direction probe value, and Z-axis direction probe value are extracted. The results are as follows: Figure 2 shown.

[0064] Furthermore, the systematic error caused by the measurement principle during the gear tooth profile measurement is calculated.

[0065] System errors include: errors caused by circular grating rotation readings, errors caused by X-axis grating readings, errors caused by Y-axis grating readings, and errors caused by Z-axis grating readings.

[0066] Furthermore, the calculated errors caused by the circular grating rotation reading and the errors caused by the moving axis grating reading during the measurement process are converted into the measurement direction.

[0067] The positive and negative sign and size of the error value are determined according to the positive direction of the measurement center coordinate system, and the error is synthesized in the measurement direction to obtain the final measurement error value.

[0068] Calculate the gear base circle radius rb according to the basic parameters of the gear being measured.

[0069] The tangent direction of the gear base circle is the normal direction of the generated gear involute, and the measurement line is also located in the normal direction of the gear involute. The result is as follows: Figure 4 shown.

[0070] In the normal direction of the generated gear involute, along the measuring direction, the trajectory line length l during the measuring process is calculated according to the gear measurement starting circle position.

[0071] In the normal direction of the generated gear involute, a rectangular coordinate system is established with the unfolded length of the involute as the horizontal coordinate and the relative deviation value as the vertical coordinate.

[0072] The calculated relative error value of the gear tooth profile is converted to the normal direction along the gear involute, and the contour curve of the gear tooth profile measurement result is constructed. The result is as follows: Figure 5 shown.

[0073] A robust Gaussian regression filter is established and Gaussian filtering is performed on the tooth profile relative deviation curve, including: Construct a Gaussian weight function: In the formula, x is the distance between the data to be processed and the center of the weight function, λc is the filter cutoff wavelength, and α is a constant. α≈0.4697 is calculated by inverse Fourier transform;

[0074] According to the national standard GB / T 10095.1-2022, the filter cutoff wavelength λc used for gear profile measurement results is calculated as follows: c = l a / 30, la is the unfolded length of the gear.

[0075] Construct a robust Gaussian weight function: In the formula, the parameters γ≈0.7309 and λ are calculated by inverse Fourier transform. c is the cut-off wavelength, x l,k =(lk)*Δx, Δx is the actual interval between the measured tooth profile data points;

[0076] Combine the robust weight function and the Gaussian weight function to establish the weight function matrix S.

[0077] Introducing robust weight function When |v l |≤c,l=1,...,n:

[0078] Where: v is the deviation of the results before and after filtering, that is, the difference between the tooth profile data after filtering and the tooth profile data before filtering, using a digital representation method;

[0079] n is the data dimension, that is, the data dimension of the digital expression of the tooth profile data before and after filtering;

[0080] c=4.4478*median|v|, and median means taking the median value, that is, taking the median value of the deviation of the results before and after filtering as the calculation threshold;

[0081] Establish the conjugate distance x between the tooth profile numerical points l,k Constructed coefficient regression matrix regression matrix

[0082] Establish the initial tooth profile curve vector Z, which is: a one-dimensional matrix composed of the data of the digitized tooth profile measurement results;

[0083] According to the weight function matrix S, a robust Gaussian regression filter equation is established.

[0084] Determine the initial value w(1) of the robust weight function iteration and calculate the deviation value v(1);

[0085] The weight function matrix iteratively calculates the Gaussian filter midline w(i) to calculate the deviation value v(i);

[0086] Use the calculated Gaussian filter center line w(i) as the original curve z(i+1) for the next calculation, iteratively calculate the filter result w(i+1), until the deviation value v(i+1) meets the requirements, and take w(i+1) as the final robust Gaussian filter result to obtain the filter result of the robust spline filter as shown in Figure 6 shown.

[0087] The curve comparison of gear tooth profile measurement results before and after filtering is as follows: Figure 7 shown.

[0088] For the tooth profile deviation curve obtained after filtering for tooth profile deviation evaluation, the tooth profile deviation is calculated based on the measurement results of the gear tooth profile, including: total tooth profile deviation, tooth profile shape deviation, and tooth profile inclination deviation.

[0089] The embodiments described above are only descriptions of the preferred modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field, all changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the description and illustrations, should fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for processing gear tooth profile measurement results based on a robust Gaussian regression filter, characterized in that: include: The gear to be measured is measured using the Klingelnberg P26 gear measuring center to obtain the underlying data records of the gear measuring center; Extract the circular grating reading, X-axis grating reading, Y-axis grating reading, Z-axis grating reading, X-axis direction probe value, Y-axis direction probe value, and Z-axis direction probe value from the underlying data records of the gear measurement center; According to the measurement principle of the gear measurement center - the generalized polar coordinate method measurement principle, the relative error value of the gear tooth profile is calculated using the extracted underlying data of the gear measurement center; The calculated relative error value of the gear tooth profile is converted to the normal direction along the gear involute, and the profile curve of the gear measurement result is constructed; Establishing a robust Gaussian regression filter, and performing Gaussian filtering on the contour curve of the gear measurement result according to the established robust Gaussian regression filter to obtain a tooth profile curve; The tooth profile deviation is calculated for the obtained tooth profile curve, including: total tooth profile deviation, tooth profile shape deviation, tooth profile inclination deviation, to determine the deviation of the gear tooth profile.

2. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The gear profile deviation measurement process measurement center bottom data recording and extraction includes: The records of circular grating angle readings, X-axis grating readings, Y-axis grating readings, Z-axis grating readings, X-axis direction probe value records, Y-axis direction probe value records, and Z-axis direction probe value records during the measurement process are extracted from the built-in software records of the measurement center.

3. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The measuring principle of the gear tooth profile measurement result includes: The generalized polar coordinate method is used to measure the deviation in the direction of the involute gear tooth profile. The probe is deviated from the base cylindrical axis plane by a certain distance. During the measurement process, the relative displacement between the actual position of the probe and the polar diameter of the tooth profile is measured.

4. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The calculation method of the gear tooth profile measurement result includes: Read the circular grating angle reading record and calculate the angle change during the measurement process; Read the X-axis grating reading record and calculate the displacement of the probe during the measurement process; Read the probe indication record in the X-axis direction and calculate the measurement results of the tooth profile during the measurement process; The tooth profile measurement result is calculated based on the change in the rotation angle, the displacement of the probe and the measurement result of the tooth profile.

5. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The method for converting the coordinate system of the gear tooth profile measurement result comprises: Establish a coordinate system in the normal direction of the gear along the involute of the gear: The rotation amount in the process of gear tooth profile measurement is converted into a linear amount. The horizontal axis is the distance in the normal direction of the gear involute during the gear measurement process, and the vertical axis is the calculated value of the tooth profile measurement result, and a curve of the gear tooth profile measurement result is obtained.

6. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The method of establishing a robust Gaussian regression filter includes performing Gaussian filtering on the involute gear tooth profile measurement result curve according to the robust Gaussian regression filter to obtain a tooth profile deviation curve, including: 1) Based on the dual-weight influence function asymmetric function When |x|≤c, Introducing the robust weight function in robust constraint theory When |v l |≤c,l=1,...,n, Wherein: v is the deviation of the tooth profile data points before and after filtering, that is, the difference between the tooth profile data after filtering and the tooth profile data before filtering, using a digital representation method; n is the data dimension, that is, the data dimension of the digital expression of the tooth profile data before and after filtering; l traverses from 1 to n, indicating that the above calculation is performed on all numerical points of the digitized tooth profile data; c = 4.4478*median|v|, and median represents the median, that is, the median of the deviation of the results before and after filtering is taken as the calculation threshold; 2) According to the robust weight function, according to the Gaussian weight function Establish the weight function matrix S k , In the formula, s n,k is a robust Gaussian weight function, Parameters γ≈0.7309, λ c is the cut-off wavelength, x l,k =(lk)*Δx, where: The mathematical meaning of the Gaussian weight function g(x) is a bell-shaped curve with Gaussian mathematical law. x represents the center position of the bell-shaped curve, and λ c is the cut-off wavelength, and the Gaussian weight function g(x) is at the cut-off wavelength λ c The curve amplitude passing rate at is 50%, and the parameter γ≈0.7309 is calculated by inverse Fourier transform; the Gaussian weight function g(x) is intended to perform convolution operation on all data points of the original tooth profile within the range of the Gaussian function; Combining the Gaussian weight function g(x) with the theory of robust weight function, we construct a robust Gaussian weight function s n,k , where: Δx represents the actual interval between the numerical points of the digitized tooth profile data, that is, the sampling interval of the gear measurement center when measuring the tooth profile; x l,k =(lk)*Δx, which represents the conjugate distance between each point of all numerical points of the digitized tooth profile data and other data points; through the conjugate distance x between the numerical points of the tooth profile l,k , construct a robust Gaussian weight function s n,k , robust Gaussian weight function s n,k *δ n It represents the data points of the tooth profile. Robust constraints are established between each point and other data points. The established robust constraints are then arranged diagonally to form a weight function matrix S k ; 3) According to the weight function matrix S k , establish the robust Gaussian regression filter equation Where: Z is the initial curve matrix, that is, a one-dimensional matrix composed of the digitized data of the tooth profile measurement results. is the conjugate distance x between the numerical points of the tooth profile l,k The coefficient regression matrix constructed has the first column of the matrix as 1 and the second column as the conjugate distance x. l,k , the third column is x l,k 2 The purpose of constructing the coefficient matrix is ​​to make the filtering process have a second-order robust property; by calculating the robust Gaussian regression filtering equation, the tooth profile data point matrix w after robust Gaussian regression filtering can be calculated.

7. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The method of establishing a robust Gaussian regression filter includes performing Gaussian filtering on the involute gear tooth profile measurement result curve according to the robust Gaussian regression filter to obtain a tooth profile deviation curve, and further includes: Performing Gaussian filtering on the involute gear tooth profile measurement result curve, using the robust filtering model established in step 3) of claim 6 to determine the initial value w(1) of the robust weight function iteration, and calculating the deviation value v(1); Using the robust weight function obtained in step 1) of claim 6 and the weight function matrix obtained in step 2) of claim 6, iteratively calculate the Gaussian filter centerline w(i); Use the calculated Gaussian filter center line w(i) as the original curve z(i+1) for the next calculation, iteratively calculate the filter result w(i+1) until the deviation value v(i+1) meets the requirements, and take w(i+1) as the final robust Gaussian filter result.

8. The method for processing gear tooth profile measurement results based on a robust Gaussian regression filter according to claim 1, characterized in that: The method for evaluating the deviation curve of the gear tooth profile measurement result comprises: According to the national standard GB / T 10095.1-2022, the tooth profile deviation is calculated based on the deviation curve of the gear tooth profile measurement result, including: total tooth profile deviation, tooth profile shape deviation, and tooth profile inclination deviation.