A method for self-positioning of machining reference of an additively repaired bearing bush

Abstract

A kind of machining datum self-positioning method of additive repair bearing bush, comprising the following steps: S1. obtaining point cloud data of bearing bush surface;S2. preprocessing point cloud data;S3. point cloud registration aligns bearing bush surface with point cloud model;S4. after surface reconstruction, obtain feature points of bearing bush surface;S5. based on point cloud data, accurately extract feature points, and then calculate the geometric center of bearing bush;S6. determine the machining axis of bearing bush by curve fitting algorithm;S7. real-time acquisition of bearing bush positioning data determines machining datum and axis parameters, and automatically adjusts the position and parameters of machining equipment through compensation logic and three-dimensional reconstruction algorithm, so that the machining equipment matches the actual shape and position of bearing bush, so that the bearing bush is placed on the machining platform, the computer can automatically identify and determine the machining datum, axis and other parameters, and the position of machining equipment can be adjusted in real time through compensation method to match the bearing bush, improve the machining efficiency and machining precision, and have good practical value and economic value.

Description

Technical Field

[0001] This invention relates to the field of intelligent mechanical processing technology, and in particular to a self-locating method for machining reference in additive repair of bearing bushes. Background Technology

[0002] In machining, determining the machining datum is crucial for machining accuracy and quality. Traditional methods for determining the machining datum rely on manual adjustment and measurement, resulting in low machining efficiency and low accuracy. Therefore, there is an urgent need for a self-locating method for the machining datum in bearing machining to improve machining efficiency and accuracy.

[0003] Bearing bushes are important components in mechanical equipment. After additive manufacturing, they need to undergo subtractive manufacturing before they can be put into direct use. Currently, subtractive manufacturing mainly relies on manual scraping of the bushes, which requires highly skilled workers, results in poor processing accuracy, and is not conducive to widespread adoption. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a self-locating method for the machining reference of additive repair bearing bushes, which can improve the accuracy of subtractive machining of bearing bushes and solve the technical problems in the current subtractive machining process of bearing bushes.

[0005] The technical solution adopted in this invention is: a self-locating method for machining reference in additive manufacturing repair of bearing bushes, comprising the following steps:

[0006] S1. The machine tool uses a vision measurement device to perform point cloud scanning on the bearing bush clamped on the machine tool worktable to obtain point cloud data of the bearing bush surface;

[0007] S2. Preprocess the point cloud data, including filtering and denoising, outlier removal, and coordinate transformation.

[0008] S3. Point cloud registration aligns the bearing surface with the point cloud model;

[0009] S4. After surface reconstruction, obtain the feature points on the bearing surface. The feature points are determined by curvature, normal vector, and curvature change index; among them, points with significant geometric features are selected as feature points.

[0010] S5. Based on point cloud data, accurately extract feature points, and then calculate the geometric center of the bearing bush to determine the machining reference.

[0011] S6. Determine the machining axis of the bearing bush using a curve fitting algorithm;

[0012] S7. Real-time acquisition of bearing positioning data determines the machining datum and axis parameters, and automatically adjusts the position and parameters of the machining equipment through compensation logic and three-dimensional reconstruction algorithm to match the actual shape and position of the machining equipment with the bearing.

[0013] The present invention has the following beneficial effects:

[0014] 1. This invention utilizes online measurement technology to generate a three-dimensional model of the additive bearing bush, and by comparing it with the theoretical model, it rationally allocates the machining allowance and performs automatic programming to realize the automated subtractive machining of the bearing bush.

[0015] 2. The self-positioning method of the machining datum of the present invention can realize the automatic positioning of the machining datum, so that the bearing bush is clamped on the machining platform, and the position of the machining equipment can be adjusted in real time to match the bearing bush through the compensation method. The system can automatically identify and determine the machining datum, axis and other parameters, thereby improving machining efficiency and machining accuracy. Attached Figure Description

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0017] Figure 1 This is a schematic diagram of feature point extraction according to the present invention. Detailed Implementation

[0018] A self-locating method for machining reference in additive manufacturing repair of bearing bushes includes the following steps:

[0019] S1. The machine tool uses a vision measurement device to perform point cloud scanning on the bearing bush clamped on the machine tool worktable to obtain point cloud data of the bearing bush surface.

[0020] A vision measurement device is installed on the machine tool, and the vision measurement is used to scan the bearing bush to obtain point cloud data of the bearing bush surface. Obtaining the target 3D point cloud data involves reconstructing a dense point cloud from multiple sets of images, a method of reconstructing a 3D scene from multiple 2D images. Its basic principle is to capture images of the same target or scene from multiple perspectives, and then process these images using computer vision and geometric algorithms to generate a dense 3D point cloud. When measuring the bearing bush position coordinates, a 3D local coordinate system is automatically constructed during data acquisition by the laser scanning device, with the center point of the signal transmitter as the origin O. In this coordinate system, the axes are determined by the rotation axis of the instrument, where the X-axis is in the horizontal plane, the Z-axis is in the vertical plane, and the Y-axis can be determined according to the right-hand rule. The 3D coordinates of the object are obtained through certain mathematical calculations using the measurement elements acquired by the scanner. These measurement elements mainly include the straight-line distance r from the center of the signal transmitter to the object surface, the horizontal angle φ, and the vertical angle ϴ.

[0021] ;

[0022] The coordinates of three-dimensional points on the object's surface can be obtained by calculating using the formula above.

[0023] S2. Preprocess the point cloud data, including filtering and denoising, outlier removal, and coordinate transformation, to improve the accuracy and efficiency of subsequent processing.

[0024] First, by obtaining the maximum and minimum coordinates of the point cloud data on the coordinate axes, the distribution intervals in different directions, i.e., three lengths, are statistically determined. Then, the entire point cloud data is divided into voxel grids of equal size with a resolution of [resolution missing]. The point cloud density within each voxel grid is calculated. Then, point clouds with densities exceeding a set threshold within each voxel grid are downsampled, retaining only one coordinate point to represent the center point of the point cloud within that voxel grid. To reduce sampling requirements, there are two methods for selecting sampling points within a voxel: one is the centroid selection method, where the centroid coordinates of points falling within the voxel grid are calculated, and these coordinates are the voxel downsampling coordinates; the other is to find the point closest to the center coordinates of the voxel, which is the voxel downsampling point. Both methods can achieve point cloud data simplification, but their sampling effects and processing speeds differ. The centroid grid downsampling process is as follows:

[0025] a. Find the extreme values ​​of the point cloud coordinates in coordinate space. (The result is...) .

[0026] b. Calculate the mesh space size, where the voxel size is... voxels The size of the point cloud affects its level of simplification. Setting the value too low results in a large amount of point cloud data, leading to poor quality reduction. If the setting is too large, the point cloud features of the buildings will be lost.

[0027] ;

[0028] ;

[0029] ;

[0030] In the formula, Represents the size of the voxel along the X-axis; Indicates the size of the voxel along the Y-axis; Indicates the size of the voxel along the Z-axis. This is a rounding up operation.

[0031] c. Point cloud to voxel mesh mapping.

[0032] ;

[0033] ;

[0034] ;

[0035] ;

[0036] In the formula, Represents the number of voxel midpoints along the X-axis; Indicates the number of voxel midpoints along the Y-axis; This indicates the number of voxel midpoints along the Z-axis.

[0037] Grid with Arrange in order. For grid indexing, This is a rounding up operation. Calculate the position of each point within the grid, and take the grid center as the sampled point.

[0038] d. Sort according to the indices in c, and represent it using the centroid sampling method as follows:

[0039] ;

[0040] in, The number of points in the grid voxels. These are the coordinates after centroid sampling. These are the points mapped onto the grid.

[0041] Due to the unavoidable noise issues present during the acquisition of 3D point cloud data, these noise points increase the uncertainty of the point cloud data, affect its geometric structure and surface features, and thus reduce the information content. Therefore, point cloud quality optimization processing is necessary. The algorithm principle is as follows:

[0042] a. At the origin point In this context, n represents the total number of building point clouds. For any point in the tree, use a kd-tree to search for the current point. Find the k nearest neighbors of (x, y, z) by calculating the points Distance to k neighboring points Then, the average value at that point can be calculated. ;

[0043] ;

[0044] ;

[0045] b. The Gaussian distribution of the point density in the point set P is determined by the mean (μ) and standard deviation (σ). Calculate the distribution of the points in P. The mean and standard deviation are μ, which reflects the spatial distribution information of the points in the original data, and σ is the approximation of the mean to the population mean.

[0046] ;

[0047] ;

[0048] c. By setting an appropriate distance threshold t, determine whether a point is an outlier using the following formula:

[0049] ;

[0050] In the formula, L is the maximum distance threshold, when the average distance Points smaller than L are retained, while points larger than L are marked as noise points and removed by the program.

[0051] S3. Use the ICP (Iterative Closest Point) algorithm to align the point clouds of different parts, aligning the bearing surface with the point cloud model. The specific steps are as follows:

[0052] Step 1: Construct matching point pairs. Use the KD-Tree method to search for the nearest point in the target point cloud for all points in the source point cloud and construct matching point pairs.

[0053] Step 2: Remove noise values ​​by using the Random Sample Consensus (RANSAC) algorithm to remove outliers from the constructed set of matching pairs.

[0054] Step 3: Based on the constructed set of point pairs, use SVD to calculate the optimal rotation matrix R and translation vector t.

[0055] Step 4: Perform rotation and translation calculations on the source point cloud using the rotation and translation transformation matrix obtained in the previous step to obtain the new point set after the rigid body transformation of the source point cloud.

[0056] Step 5: Repeat steps 1 to 4 until the rotation and translation results obtained between two iterations converge or the maximum number of iterations has been reached.

[0057] S4. See also Figure 1 After surface reconstruction, feature point extraction algorithms are used to obtain feature points on the bearing surface. The feature points are determined by indicators such as curvature, normal vector, and curvature change. Among them, points with significant geometric features are selected as feature points.

[0058] S5. Based on point cloud data, accurately extract feature points, and then calculate the geometric center of the bearing bush to determine the machining reference.

[0059] Feature point extraction algorithms extract a set of feature points from 3D bearing data. This involves obtaining the coordinate information of these feature points, i.e., their positions in 3D space. Using this coordinate information, a mathematical algorithm calculates the geometric center of these points. Alternatively, the geometric center can be calculated by averaging the X, Y, and Z coordinates of the feature points.

[0060] Given a source point cloud Typically, a few key feature points are extracted from the point cloud first. The feature point extraction process is denoted as:

[0061] ;

[0062] in Represents a feature point extractor. Given the extracted feature point set, we have .

[0063] Feature point extraction uses a method based on the angle between normal vectors. First, the normal vector of each point in the point cloud model is calculated. For normal vector calculation, a KD-Tree is first constructed from the original point cloud. Then, for each point in the point cloud, its neighboring points are searched, and a surface is fitted to the neighboring points containing that point. Next, the covariance matrix of the neighborhood set is calculated, and the eigenvalues ​​and corresponding eigenvectors of this matrix are obtained. Principal component analysis shows that the eigenvector corresponding to the smallest eigenvalue is the normal vector of the fitted surface. Specifically, assuming the sampling points... The neighborhood set is First, calculate the centroid of the neighborhood set according to the following formula:

[0064] ;

[0065] Then, the covariance matrix of the neighborhood set is calculated according to the following formula. :

[0066] ;

[0067] Find the eigenvalues ​​of the matrix. and its corresponding eigenvectors Then the eigenvector corresponding to the smallest eigenvalue That is, sampling points The normal vector of each point is then calculated. The angle between the normal vector of each point and the normal vectors of its neighboring points is then calculated, and the overall feature points are extracted based on a certain threshold.

[0068] Surface curvature is also an important geometric feature describing the degree of curvature of a surface. Similarly, feature point extraction algorithms are used to obtain feature points on the surface, and then mathematical algorithms are used to calculate the curvature information of these feature points. For each feature point, its principal curvature and mean curvature are calculated, thus obtaining the local curvature distribution of the surface.

[0069] S6. Determine the machining axis of the bearing bush using a curve fitting algorithm.

[0070] The direction of the main axis of the bearing is obtained using a curve fitting algorithm. By fitting feature points, the vector of the main axis direction of the bearing can be obtained. The fitting methods are least squares method and principal component analysis.

[0071] Curve fitting can be performed using two-dimensional bearing data to obtain the main axis direction of the bearing. Based on the measurements of the actual models of the inner and outer surfaces of the bearing described earlier, two-dimensional data points of the inner and outer surfaces are obtained. Then, using a curve fitting algorithm and the least squares method, the discrete data points are fitted into a curve, thus obtaining the main axis direction of the bearing. Based on the reconstructed surface model, the extracted feature points, and the fitted curve, mathematical algorithms are used to extract the geometric features of the bearing. This includes calculating the geometric center, determining the main axis direction, and calculating the surface curvature.

[0072] ;

[0073] a. The sum of the distances from a point to the curve, i.e., the sum of the squares of the deviations. :

[0074] ;

[0075] b. To obtain the condition that meets the requirements Value, on the right side of the equation Find the partial derivatives:

[0076] ;

[0077] ;

[0078]

[0079] ;

[0080] c. Simplify and rearrange the equation on the left:

[0081] ;

[0082] ;

[0083]

[0084] ;

[0085] d. Convert the equations into matrices:

[0086] ;

[0087] Solving for the coefficient A yields the two-dimensional fitting curve. This curve represents the fitted bearing axis, and its parameters provide data support for subsequent algorithms.

[0088] S7. Real-time acquisition of bearing positioning data determines the machining datum and axis parameters, and automatically adjusts the position and parameters of the machining equipment through compensation logic and three-dimensional reconstruction algorithm to match the actual shape and position of the machining equipment with the bearing.

[0089] Once the machining datum and axis parameters of the bearing bush are determined, the control system can automatically adjust the position and parameters of the machining equipment to match the actual shape and position of the bearing bush. A compensation method is used to reconstruct the three-dimensional machining surface of the bearing bush, allowing the system to automatically adjust the machining parameters based on the surface features and position of the bearing bush. The main implementation process includes the following two parts: First, the machining surface is reconstructed based on NURBS reconstruction technology, and then the phase difference is calculated; then, the tool position coordinates in the toolpath file are modified according to the phase difference value to complete the compensation.

[0090] A NURBS curve is a curve composed of nodal vector parameters. sequence The decision A polynomial of degree n. Its equation is defined as:

[0091] ;

[0092] in, To control the vertices, the convex hull polygon formed by sequentially connecting these vertices is called the control polygon. It is a k-th degree B-spline basis function.

[0093] Given There are 10 control vertices, which are arranged in an array. lines and Columns are represented as: These arrayed control points constitute a control grid. (Definition) The direction of each point is The direction, where the curve in this direction has a degree of k. Definition The direction of each point is The direction, the degree of the curve in this direction is Two more node vectors are given. This defines a sheet A non-uniform rational B-spline surface with a tensor product of degree 1. Its equation can be expressed as:

[0094] ;

[0095] After determining the node vectors, and knowing the control vertices... By determining the degree k, a NURBS curve is defined. For any parameter value within the domain of the NURBS curve... Its corresponding point on the NURBS curve It can be calculated using the following formula.

[0096]

[0097] Interpolation used for inverse curve finding Data points The equation of the k-th degree NURBS curve can be written as:

[0098] ;

[0099] Define the domain of the curve Substituting the node values ​​into the above equation, the following interpolation condition should be satisfied:

[0100] ;

[0101] The specific implementation process for automatically adjusting the position and parameters of the processing equipment to match the actual shape and position of the bearing bush is as follows:

[0102] ① Based on the phase difference calculation of the reconstructed surface, the measured coordinates and tool position points are not completely coincident during the actual measurement process. Therefore, the machining surface needs to be reconstructed during the phase difference calculation. First, the NURBS curve and surface are inversely calculated based on the measured points to obtain the surface definition. Then, the NURBS curve and surface are forward calculated, the tool position point coordinates are parameterized, and the beginning and end of the tool position point are set to coincide with the beginning and end of the measured points. Based on the curve equation and the parameterized tool position point, the actual coordinate values ​​of the tool position point can be calculated, and then the difference value of each tool position point can be obtained.

[0103] NURBS surface forward solving: Given any set of parameter values ​​within the domain of the surface to be solved... To find the corresponding point on the NURBS surface The following steps can be followed.

[0104] Step 1: Calculate the intermediate polygon. Using... Parameter values ​​along Parameter direction The de Boer recursive algorithm is executed to control the polygon and obtain the result. Using points as intermediate vertices, these intermediate vertices form a path along... The intermediate polygon in the parameter direction.

[0105] Step 2: Calculate the points on the surface. Using Applying the De Boer recursive algorithm to this intermediate polygon with the parameter values, the resulting point represents the parameter on the NURBS surface. Corresponding points .

[0106] NURBS surface inverse problem: Based on the method for determining node vectors in the NURBS curve inverse problem, the node vectors in two parametric directions can be determined using given surface type points. and After the node vectors are determined, the calculation of the control vertices of the NURBS surface can be performed according to the following steps:

[0107] Step 1: Calculate the intermediate control vertices. This can be obtained along the u-parameter direction of the surface. Groups of data points, each group containing One data point. For this... By calculating each set of data points using the inverse curve method, we can obtain... The middle control vertex of the group, each group contains One intermediate control vertex.

[0108] Step 2: Calculate the control mesh. Then, along the surface v-parameter direction, calculate the intermediate control vertices obtained in Step 1 using the inverse curve calculation method to obtain the control mesh. Group control vertices, each group contains Each control vertex. This allows us to obtain... The control polygon mesh formed by the control vertices is the control vertex mesh of the NURBS surface.

[0109] ② Based on the compensation of the toolpath file, the coordinates of the toolpath points in the toolpath file are modified according to the difference of each toolpath point calculated above, and a compensated toolpath file is generated. This file is then output to the CAD software for post-processing, and finally, NC code with compensation is generated for compensated machining.

[0110] Combining the above steps, the self-locating method for machining datums of the present invention can achieve automatic positioning of machining datums, thereby improving machining efficiency and accuracy. By monitoring the machining status of the bearing bush in real time and adjusting the machining datum points according to the monitoring results, automatic positioning of the machining datums is achieved, thus improving machining efficiency and accuracy.

[0111] This invention is not limited to the above-described embodiments; all variations or modifications are within the scope of protection of this invention.

[0112] By implementing the technical solution of this invention, automatic positioning of the machining datum can be achieved, which improves machining efficiency and accuracy, and has good practical and economic value.

Claims

1. A self-locating method for machining reference in additive manufacturing of bearing bushes, characterized in that, Includes the following steps: S1. The machine tool uses a vision measurement device to perform point cloud scanning on the bearing bush clamped on the machine tool worktable to obtain point cloud data of the bearing bush surface; S2. Preprocess the point cloud data, including filtering and denoising, outlier removal, and coordinate transformation. S3. Align the bearing surface with the point cloud model and perform point cloud registration; S4. After surface reconstruction, obtain the feature points on the bearing surface. The feature points are determined by curvature, normal vector, and curvature change index; among them, points with significant geometric features are selected as feature points. S5. Based on point cloud data, accurately extract feature points, and then calculate the geometric center of the bearing bush to determine the machining reference. S6. Determine the machining axis of the bearing bush using a curve fitting algorithm; S7. Real-time acquisition of bearing positioning data to determine machining datum and axis parameters, and automatic adjustment of the position and parameters of the machining equipment through compensation logic and three-dimensional reconstruction algorithm to match the actual shape and position of the bearing; In S2, the preprocessing steps for point cloud data are as follows: S21. By obtaining the maximum and minimum coordinates of the point cloud data of the bearing tile on the coordinate axis, the distribution range in different directions is statistically determined, namely, the three lengths. S22. Divide the entire point cloud data into a voxel grid of equal size, with a resolution of [resolution missing]. And calculate the point cloud density within each voxel grid; S23. Downsample the point cloud with a density higher than a set threshold in each voxel grid, and retain only one coordinate point to represent the center point of the point cloud in that voxel grid. In S3, the steps to align the bearing surface with the point cloud model are as follows: S31. Construct matching point pairs; Use the KD-Tree method to search for the nearest point in the target point cloud for all points in the source point cloud, and construct matching point pairs; S32. Remove noise values; use the random sampling consensus algorithm to remove outliers from the constructed set of matching point pairs. S33. Based on the constructed set of point pairs, use SVD to calculate the optimal rotation matrix R and translation vector t; S34. Perform rotation and translation calculations on the source point cloud using the rotation and translation transformation matrix obtained in the previous step to obtain the new point set after rigid body transformation of the source point cloud; S35. Repeat the calculation of S31 to S34 until the rotation and translation results obtained between the two iterations converge or the maximum number of iterations has been reached.

2. The self-locating method for machining reference in additive repair of bearing bushes according to claim 1, characterized in that: In S1, a vision measurement device installed on the machine tool performs point cloud scanning on the bearing clamped on the machine tool table and measures the bearing's position coordinates. The system automatically constructs a three-dimensional local coordinate system with the center point of the vision measurement device as the origin O. In this coordinate system, the X-axis is in the horizontal plane, the Z-axis is in the vertical plane, and the Y-axis is determined according to the right-hand rule. The three-dimensional coordinates of the bearing are obtained through measurement elements acquired by the vision measurement device. The measurement elements include the straight-line distance r from the center of the vision measurement device to the object surface, the horizontal angle φ, and the vertical angle. The coordinates of three-dimensional points on the object's surface can be calculated using the following formula. 。 3. The self-locating method for machining reference of additive repair bearing bush according to claim 1, characterized in that: In S23, the steps for downsampling the point cloud within the voxel mesh are as follows: S231. Find the extreme values ​​of the point cloud coordinates in the coordinate space, and obtain... ; S232. Calculate the mesh space size, the voxel size is... voxels The size of the point cloud affects its level of simplification. Setting the value too low results in a large amount of point cloud data, leading to poor quality reduction. If the setting is too large, the point cloud features of the buildings will be lost; ； ； ； In the formula, Represents the size of the voxel along the X-axis; Indicates the size of the voxel along the Y-axis; Indicates the size of the voxel along the Z-axis. This is a rounding up operation; S233, Point Cloud and Voxel Mesh Mapping ； ； ； ； In the formula, Represents the number of voxel midpoints along the X-axis; Indicates the number of voxel midpoints along the Y-axis; Indicates the number of voxel midpoints along the Z-axis; Grid with Arrange in order. For grid indexing, calculate the position of each point within the grid, and take the center point of the grid as the sampled point; S234. Sort according to the index in S233, and represent it according to the centroid sampling method as follows: ； in, The number of points in the grid voxels. These are the coordinates after centroid sampling. These are the points mapped onto the grid.

4. The self-locating method for machining reference of additive repair bearing bush according to claim 3, characterized in that: The steps for optimizing point cloud quality are as follows: a. At the origin point In this context, n represents the total number of building point clouds. For any point in the tree, use a kd-tree to search for the current point. Find the k nearest neighbors of (x, y, z) by calculating the points Distance to k neighboring points Then, the average value at that point can be calculated. ; ； ； b. The Gaussian distribution of the point density in point set P is determined by the mean μ and the standard deviation σ. Calculate the Gaussian distribution of the points in P. The mean and standard deviation are μ, which reflects the spatial distribution information of the points in the original data, and σ is the approximation of the mean to the population mean. ； ； c. By setting an appropriate distance threshold t, determine whether a point is an outlier using the following formula: ； In the formula, L is the maximum distance threshold, when the average distance Points smaller than L are retained; points larger than L are designated as noise points and removed.

5. The self-locating method for machining reference of additive repair bearing bush according to claim 1, characterized in that: In S5, the feature point extraction algorithm extracts a set of feature points from the 3D bearing data. The coordinate information of the feature points is the position of the feature points in 3D space. Using the coordinate information of the feature points, the geometric center is calculated by averaging the X, Y, and Z coordinates of the feature points.

6. The self-locating method for machining reference of additive repair bearing bush according to claim 1, characterized in that: In S6, the main axis direction of the bearing is obtained by performing curve fitting on the two-dimensional bearing data. Two-dimensional data points on the inner and outer surfaces of the bearing are obtained by measuring the actual models of the bearing's inner and outer surfaces. Then, using a curve fitting algorithm and the least squares method, the discrete data points are fitted into a curve to obtain the main axis direction of the bearing. Based on the reconstructed surface model, the extracted feature points, and the fitted curve, the geometric features of the bearing are extracted. The steps for extracting the geometric features of the bearing are as follows: The initial curve expression for curve fitting is: ； a. The sum of the distances from a point to the curve, i.e., the sum of the squares of the deviations. : ； b. To obtain the condition that meets the requirements Value, on the right side of the equation Find the partial derivatives: ； c. Simplify the equation on the left: ； d. Convert the equations into matrices: ； By solving for the coefficient A, the two-dimensional fitting curve is obtained. This curve is the fitted axis of the bearing, and its parameters can provide data support for subsequent algorithms.

7. The self-locating method for machining reference of additive repair bearing bush according to claim 1, characterized in that: In S7, a compensation method is used to reconstruct the three-dimensional machining surface of the bearing bush, enabling the system to automatically adjust the machining parameters according to the surface features and position of the bearing bush. The process includes the following two parts: First, the machining surface is reconstructed based on NURBS reconstruction technology, and then the phase difference is calculated; then, the tool position coordinates in the toolpath file are modified according to the phase difference value to complete the compensation.

8. The self-locating method for machining reference of additive repair bearing bush according to claim 7, characterized in that: In S7, the position and parameters of the machining equipment are automatically adjusted through compensation logic and 3D reconstruction algorithm to match the actual shape and position of the machining equipment with the bearing. The specific implementation process is as follows: First, the NURBS curve and surface are inversely derived based on the measured points to obtain the surface definition. Then, the NURBS curve and surface are derived in the forward direction. The tool position coordinates are parameterized, and the beginning and end of the tool position are set to coincide with the beginning and end of the measured points. Based on the curve equation and the parameterized tool position, the actual coordinate values ​​of the tool position can be calculated, and then the difference of each tool position can be obtained. Based on the compensation of the toolpath file, the coordinates of the tool points in the toolpath file are modified according to the difference of each tool point calculated above, and a compensated toolpath file is generated. It is then output to the CAD software for post-processing, and finally, NC code with compensation is generated for compensated machining. Among them, the forward NURBS surface search: given any set of parameter values ​​within the domain of the surface to be searched. To find the corresponding point on the NURBS surface Please follow these steps. Step 1: Calculate the intermediate polygon; using... Parameter values ​​along Parameter direction The de Boer recursive algorithm is executed to control the polygon and obtain the result. Using points as intermediate vertices, these intermediate vertices form a path along... The intermediate polygon in the parameter direction; Step 2: Calculate the points on the surface using... Applying the De Boer recursive algorithm to this intermediate polygon with the parameter values, the resulting point represents the parameter on the NURBS surface. Corresponding points ; NURBS surface inverse problem: Based on the method for determining node vectors in the NURBS curve inverse problem, the node vectors in two parametric directions can be determined using given surface type points. and After the node vectors are determined, the calculation of the control vertices of the NURBS surface can be performed according to the following steps: Step 1: Calculate the intermediate control vertices; this can be obtained along the u-parameter direction of the surface. Groups of data points, each group containing These data points, for this By calculating each set of data points using the inverse curve method, we can obtain... The middle control vertex of the group, each group contains One intermediate control vertex; Step 2: Calculate the control mesh; then, along the v-parameter direction of the surface, calculate the intermediate control vertices obtained in Step 1 using the inverse curve calculation method to obtain... Group control vertices, each group contains Each control vertex can be used to obtain the result. The control polygon mesh formed by the control vertices is the control vertex mesh of the NURBS surface.

Images