CN118003031B - Self-adaptive machining method for material-adding repairing bearing bush - Google Patents
Self-adaptive machining method for material-adding repairing bearing bush Download PDFInfo
- Publication number
- CN118003031B CN118003031B CN202410418027.6A CN202410418027A CN118003031B CN 118003031 B CN118003031 B CN 118003031B CN 202410418027 A CN202410418027 A CN 202410418027A CN 118003031 B CN118003031 B CN 118003031B
- Authority
- CN
- China
- Prior art keywords
- point
- point cloud
- points
- bearing bush
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003754 machining Methods 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims description 110
- 238000012545 processing Methods 0.000 claims abstract description 89
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 45
- 238000005259 measurement Methods 0.000 claims abstract description 15
- 239000000654 additive Substances 0.000 claims abstract description 12
- 230000000996 additive effect Effects 0.000 claims abstract description 12
- 238000003672 processing method Methods 0.000 claims abstract description 12
- 239000013598 vector Substances 0.000 claims description 64
- 239000011159 matrix material Substances 0.000 claims description 49
- 230000008569 process Effects 0.000 claims description 48
- 230000009466 transformation Effects 0.000 claims description 25
- 239000000463 material Substances 0.000 claims description 19
- 238000005520 cutting process Methods 0.000 claims description 18
- 238000013519 translation Methods 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 16
- 230000003044 adaptive effect Effects 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 11
- 238000013461 design Methods 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 8
- 230000014509 gene expression Effects 0.000 claims description 8
- 238000007514 turning Methods 0.000 claims description 8
- 230000003287 optical effect Effects 0.000 claims description 6
- 238000007619 statistical method Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 4
- 230000000903 blocking effect Effects 0.000 claims description 3
- 238000012512 characterization method Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000010438 heat treatment Methods 0.000 claims description 3
- 239000010687 lubricating oil Substances 0.000 claims description 3
- 238000003801 milling Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000004441 surface measurement Methods 0.000 claims description 3
- 238000004381 surface treatment Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 2
- 230000008439 repair process Effects 0.000 abstract description 12
- 230000006870 function Effects 0.000 description 24
- 238000005516 engineering process Methods 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 6
- 230000009467 reduction Effects 0.000 description 5
- 238000005299 abrasion Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000007790 scraping Methods 0.000 description 3
- 238000005211 surface analysis Methods 0.000 description 3
- 229910000897 Babbitt (metal) Inorganic materials 0.000 description 2
- 241001522296 Erithacus rubecula Species 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 230000003746 surface roughness Effects 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23P—METAL-WORKING NOT OTHERWISE PROVIDED FOR; COMBINED OPERATIONS; UNIVERSAL MACHINE TOOLS
- B23P6/00—Restoring or reconditioning objects
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Numerical Control (AREA)
Abstract
The invention relates to a self-adaptive processing method of an additive repairing bearing bush, which comprises the following steps: s1, determining machining allowance of a bearing bush tolerance surface according to machining process requirements of the bearing bush; s2, measuring a free-form surface of an actual machining surface of the bearing bush to obtain surface parameters, and generating an accurate point cloud model of the measured free-form surface of the bearing bush by a point cloud processing method; s3, performing surface matching on the point cloud model and the CAD model, and determining a machining positioning reference to obtain a CAD theoretical model; s4, comparing the actual model of the bearing bush obtained through online measurement with a theoretical model to determine deviation, and automatically adjusting a processing path and processing parameters through a self-adaptive processing algorithm to realize accurate processing of the bearing bush. The invention can realize the accurate repair and processing of the bearing bush, improve the repair efficiency, reduce the influence of human factors on the repair quality, and better ensure the geometric precision and the surface quality of the bearing bush.
Description
Technical Field
The invention relates to the technical field of metal processing, in particular to a self-adaptive processing method for an additive repairing bearing bush.
Background
The bearing bush is a key part of the hydroelectric generating set, and is easy to wear and damage in the long-term use process. The damage of the bearing bush not only reduces the power generation efficiency, but also increases the vibration in the running process of the water turbine, directly influences the service life and running safety of the water turbine, and therefore, the abrasion and damage of the bearing bush are required to be repaired. The bearing bush after material addition cannot be directly used and needs material reduction processing. At present, the material reduction processing of the bearing bush mainly depends on manual scraping, has high requirements on the skills of constructors, and is not beneficial to popularization.
Disclosure of Invention
The invention aims to solve the technical problems that: the self-adaptive processing method for the material-increasing and repairing bearing bush is characterized in that in the material-decreasing process, an online measurement technology is utilized to generate a three-dimensional model of the bearing bush after material increase, the processing allowance is reasonably distributed and automatically programmed through comparison with a theoretical model, and through the integration of the technology, the self-adaptive processing of the bearing bush can realize accurate repair and processing of the bearing bush, improve the repair efficiency, reduce the influence of human factors on repair quality and better ensure the geometric precision and surface quality of the bearing bush.
The technical scheme adopted by the invention is as follows: an adaptive processing method for an additive repairing bearing bush comprises the following steps:
S1, determining machining allowance of a bearing bush tolerance surface according to machining process requirements of the bearing bush;
S2, measuring a free-form surface of an actual machining surface of the bearing bush to obtain surface parameters, and generating an accurate point cloud model of the measured free-form surface of the bearing bush by a point cloud processing method;
s3, performing surface matching on the point cloud model and the CAD model, and determining a machining positioning reference to obtain a CAD theoretical model;
S4, comparing the actual model of the bearing bush obtained by online measurement with a theoretical model to determine deviation, and automatically adjusting a processing path and processing parameters through a self-adaptive processing algorithm to realize accurate processing of the bearing bush.
In S1, according to the machining process requirements of the bearing bush, including material selection, machining precision, surface treatment, size requirements, heat treatment and design of a lubricating oil hole, and in combination with the thermal deformation characteristics of the material and the thermal influence during cutting, the bearing bush repairing process adopts milling, and simultaneously, the bearing bush tolerance surface machining allowance is determined according to the monitored cutter wear and cutting temperature.
In S2, the free-form surface measurement of the actual machined surface of the bearing shell includes the steps of,
S21, recording a bearing bush image through an optical scanner, acquiring comprehensive surface characteristics, and processing the image to generate point cloud data;
S22, filtering point cloud; the average distance from each point to k nearest points is calculated, the distance distribution of all points in the point cloud is constructed, and outlier points are removed by means of the average value and the variance;
wherein if a certain point coordinate in the point cloud is There are k points in the neighborhood, and the point reaches any point/>, in the neighborhoodThe distance of (2) is as follows:
；
traversing and calculating all points in the neighborhood Screening out points meeting normal distribution conditions, and calculating to obtain the average value and standard deviation of normal distribution, wherein the average value and standard deviation are as follows:
；
；
Set standard deviation multiple as Then when the distance between any point in the neighborhood of a point is within the rangeWhen in, the point is reserved; conversely, if the distance is outside the range, the point is considered to be an outlier and is removed;
s23, registering the point cloud; the point cloud surface is regarded as a set of curvature and normal vector, and then statistical analysis is carried out on the set of curvature and normal vector;
S24, registering the point cloud RANSAC; the initial transformation matrix is calculated through an algorithm, so that the square sum of the distances between corresponding points in two point clouds is minimized, the specific expression is as follows,
;
In the method, in the process of the invention,Representing a transformation matrix; p represents a source point cloud; p represents a matrix of source points Yun Zhongdian; q represents a back point cloud point matrix; /(I)The extremum function of T represents the minimum value of the point set in the function;
S25, reconstructing point cloud data and fitting the surface by using a poisson reconstruction method to obtain an accurate point cloud model with complete detail characteristics and geometric surface characteristics.
In S3, constructing a CAD model according to the original design size of the bearing bush, extracting features from the point cloud model and the CAD model, wherein the features can comprise curvature, normal vector and feature points of a curved surface, the curved surface features in the point cloud data and the curved surface features in the CAD model are required to be corresponding in curved surface registration, an optimal conversion matrix is found, the curved surface in the point cloud data and the curved surface in the CAD model are aligned, and curved surface matching is completed, so that a processing positioning reference is determined.
In S4, the self-adaptive processing algorithm automatically adjusts processing parameters including cutting speed, feeding speed and cutting depth according to local characteristics and processing requirements of the curved surface, so that accurate processing of different curved surface characteristics is achieved, in the processing process, the self-adaptive processing algorithm monitors the processing state in real time, and the processing parameters are dynamically adjusted according to actual processing conditions, so that processing accuracy and processing stability are guaranteed.
In S23, the statistical analysis of the set of curvatures and normal vectors is performed as follows,
S231, weighting the base points; in order to establish a local coordinate system of the base point, the base point and its neighboring points are weighted to overcome the non-uniformity of the point cloud sampling, the specific expression is as follows,
；
Wherein,Representative pointWeights of (2); /(I)Is toRadius of sphere as center of circle, expressA weighted range; p j represents a dot matrix of dots j;
s232, analyzing a base point characteristic value; at the datum point Eigenvalue analysis is performed at the adjacent point riss _frame, the formula of the covariance matrix is shown in the following formula,
；
S233, a local reference system; calculation ofEigenvalues {,And arranged in order of magnitude, feature vectorAre also arranged in order of size toIs the origin,And its productEstablishing a local reference system for x, y and z axes;
S234, extracting feature points; setting a threshold value AndThe points that satisfy the formula are called ISS feature points, where the formula is as follows:
；
In the method, in the process of the invention, ,Three eigenvalues representing i-points,AndRepresenting the set threshold.
In S24, the algorithm needs to input the downsampled source point cloud P 'and target point cloud Q' and their corresponding description subsets, and the main steps are as follows:
s241, searching for a corresponding point; random selection Personal dotIterating in P ', finding their corresponding points in Q' by characterization；
S242, calculating a differential vector of the corresponding edge; firstly, calculating Euclidean distance between n sampling points, then calculating the ratio of the difference between two corresponding side lengths to the larger side length to form a differential vector; IfLess than side-length similarity thresholdContinuing the next step, otherwise returning to S241; wherein, whenIn the case of sample,The calculation formula of (2) is as follows:
；
In the method, in the process of the invention, Representing a point/>, in a source point cloud PAndDistance betweenAnd the same is done; /(I)Representing points/>, in the back point cloud QAndDistance betweenAnd the same is done;
S243, corresponding point transformation; estimating a temporary transformation matrix Ti from the corresponding pair of points and converting P 'to Pi';
s244, judging an inner point; calculating the nearest neighbor Euclidean distance between Pi 'and Q', and setting the distance smaller than the set distance threshold value Is taken as an inner point; if the number of interior points is too small, returning to S241;
S245, calculating a transformation matrix; calculating a transformation matrix according to the corresponding relation of the inner points; the iteration number k of the RANSAC is determined by a preset success probability p and a preset interior point proportion w, and a specific calculation formula is as follows:
；
Wherein n represents the number of sampling points, k represents the iteration times, j represents the preset success probability, and w represents the preset interior point proportion;
s246, model estimation and updating; testing other points using the interior point fitting parameterized model and updating the interior points to continue iterative computation until the best estimated model is obtained or the maximum number of iterations Riter is reached, when the error value is minimal And setting the result as the best estimation result, and stopping iteration.
In S25, using the poisson reconstruction method, the steps of reconstructing point cloud data and fitting the surface are as follows,
S251, normal line estimation and direction unification; defining a local quadric as; The calculation formula of the first partial derivative of the curved surface at a certain point is as follows:
；
In the method, in the process of the invention, ; The calculation formula of the normal direction of the curved surface at the point is shown as follows; whereinIs a coefficient of a curved surface equation,
；
Performing direction unification processing on normals in the model by using a minimum spanning tree; the minimum spanning tree algorithm defines a cost function to act on the three-dimensional point cloud, and the specific expression is as follows:
；
In the method, in the process of the invention, Representing the adjacent point,Representing the slave pointsPointing PointIs a unit vector of (a); /(I)AndRespectively represent pointsSum dotIs a normal to (2);
S252, introducing a shielding factor; in poisson's equation, the directed function is solved by solving the directed point set so that an optimal approximation is formed between the directed point set and the function gradient, i.e. by minimizing the scale function, the following equation is solved:
；
wherein; Representing a scale function; /(I) Representing the functional gradient of h; /(I)Representing a set of directed points; /(I)Is thatIn shorthand form, representing the scale between the set of directed points and the gradient of the function;
designating a set of points with weights Adding gradient constraints and discrete point value constraints in the scale function minimization problem, re-calculating a sample point function, the constraint equation is as follows,
；
In the method, in the process of the invention,Representing sample point weights; s represents a corresponding point set; /(I)Representing the area of the region defined by the set of points S; /(I)A value constraint representing a discrete point;
For ease of calculation, each sample point is weighted Setting the value to 1, simplifying the above formula,
；
In the method, in the process of the invention,Representing sample point weights; s represents a corresponding point set; /(I)A value constraint representing a discrete point;
S253, boundary constraint; and (3) adopting Robin boundary constraint, combining Dirichlet and Neumann boundary constraint, and simultaneously combining function values and normal vectors of the point cloud, and processing the point cloud with the mixed characteristic to obtain an accurate point cloud model with complete detail characteristic and geometric surface characteristic.
In S3, the step of performing surface matching between the point cloud model and the CAD model is as follows,
S31, the point cloud data is initially matched with the CAD model; placing the blank point cloud to be matched and the CAD model in the same system coordinate system by adopting a three-point method, and respectively taking three corresponding points on the blank point cloud and the CAD modelAndConstructing vector units:
；
；
In the method, in the process of the invention, 、Is an intermediate vector for solving、；
；
Unit vectorAndTwo local coordinate systems are formed, after Euclidean transformation, the two coordinate systems are completely overlapped, and then the matrixAnd translation vectorThe method comprises the following steps of:
；
；
after the coordinate system is adjusted, the original point cloud is basically overlapped with the CAD model, and the initial matching target is achieved;
s32, further unconstrained matching; obtaining a simplified point set in a uniform sampling mode, and solving by using ICP iteration;
S33, performing automatic blocking of the point cloud; firstly, a process semantic guide model is established as a block template, then seed points of corresponding structural features on a blank are quickly searched according to guide points on each curved surface in the guide model, and all compatible points are added into corresponding data blocks through a region growing algorithm, wherein the specific steps of the region growing algorithm are as follows:
S331, calculating normal vector and curvature local differential information of each measuring point in the point cloud data;
s332, traversing each surface in the process semantic guide model, and searching guide points on the current surface And calculates the normal vector/>, at that pointThen find the distance point/>, in the point cloud modelNearest pointAs a preliminary seed point for the corresponding face on the split point cloud;
S333, calculating the point cloud model at the seed point Normal vector at; IfThen byAs a seed point of the divided block, go to S335; otherwise go to S334;
s334, prompting the seed point to search for errors, and then Is an offset element, toFor the offset direction, performing equal-step iterative offset on the guide point on the curved surface, and then finding the nearest point/> to the offset point in the point cloud modelGo to S333;
S335, growing around the seed point as the center, judging whether the adjacent points have similar geometric or technological structural characteristics with the seed point according to the local differential information, if so, continuing growing, and turning to S335; if the condition is not satisfied, stopping the growth, and turning to S336;
s336, sequentially outputting the point cloud data blocks and the CAD model curved surface pieces corresponding to the point cloud data blocks, and if the point cloud data blocks are not segmented, turning to S332 to continue segmentation; otherwise the algorithm ends.
In S32, if the rotation matrix obtained by ICP isTranslation vector isThe ICP iterative solution steps are as follows:
s321, initializing a rotation matrix Is an identity matrix, translation vectorZero vector, iteration numberSet to 0 and set the maximum iteration number；
S322, for a given reduced point setFinding the nearest point/>, on a CAD modelWherein；
S323, minimizeClosed solving of the rotation matrix/>, in this iterationAnd translation vector; Wherein:
；
S324, use AndFor measurement data pointsCoordinate transformation is performed andAnd；
；
In the method, in the process of the invention,A rotation matrix representing k iterations; /(I)A translation vector representing k iterations; /(I)A distance value representing the rotation matrix and the translation vector;
s325, if ，OrThen go to S326; no makeContinuing S322;
Wherein the method comprises the steps of AndIs an allowable error;
S326, if The algorithm is terminated, the ICP is converged to the local optimum, and the optimal alignment position is output; otherwise, the iteration is finished, and the current alignment position is output.
The invention has the following beneficial effects:
1. According to the invention, in the process of material reduction and repair, an on-line measurement technology is utilized to generate an actual model of the bearing bush, machining allowance is reasonably distributed through comparison with a theoretical model, automatic programming machining repair is performed, by integrating the technology, the bearing bush self-adaptive machining can realize accurate repair and machining of the bearing bush, the repair efficiency is improved, the influence of human factors on repair quality is reduced, and the geometric precision and the surface quality of the bearing bush can be better ensured. The technology has important significance for improving the reliability of the hydroelectric generating set, prolonging the service life and reducing the maintenance cost.
Drawings
The invention is further described below with reference to the drawings and examples.
Fig. 1 is a schematic view of a machining surface, a tolerance surface and a theoretical surface on a curved surface of a blank of a bearing bush.
Fig. 2 is a partial reference frame of the base point in the present invention.
Fig. 3 is a schematic diagram of the present invention for calculating differential vectors.
Detailed Description
An adaptive processing method for an additive repairing bearing bush comprises the following steps:
S1, determining machining allowance of a bearing bush tolerance surface according to machining process requirements of the bearing bush;
S2, measuring a free-form surface of an actual machining surface of the bearing bush to obtain surface parameters, and generating an accurate point cloud model of the measured free-form surface of the bearing bush by a point cloud processing method;
s3, performing surface matching on the point cloud model and the CAD model, and determining a machining positioning reference to obtain a CAD theoretical model;
S4, comparing the actual model of the bearing bush obtained by online measurement with a theoretical model to determine deviation, and automatically adjusting a processing path and processing parameters through a self-adaptive processing algorithm to realize accurate processing of the bearing bush.
The utility model provides a to the self-adaptation processing method that needs to subtract accurate processing of material, to the material increase manufacturing, the bearing bush work piece after the restoration can't directly use, need subtract material processing, but prior art still needs artifical scraping, has free curved surface to the bearing bush after the material increase, because free curved surface on the blank does not have the processing benchmark that can fix a position, alignment location is extremely difficult, and processing location benchmark precision is difficult to guarantee. The machining allowance of a key molded surface on a blank with a complex integral structure is smaller, the thin-wall structure on the blank is often more, larger deformation can be generated in the machining process, the actual appearance of each blank with the same counting die is different, and if the blank before machining is positioned improperly on a machine tool, the problems of partial machining allowance, insufficient contour or position tolerance and the like, the problems of positioning and contact points of a cutter machining curved surface, and the matching transition of the machining curved surface and an additive are easily caused in the subsequent cutting machining. The self-adaptive processing method is used for carrying out material reduction processing on the material-increasing bearing bush.
Referring to fig. 1, for the upper surface of the bearing bush with the free-form surface, the upper surface of the blank can be divided into a surface to be processed, which needs to be allocated with allowance, a non-processed surface which needs to be ensured with allowance of a processing technology, an allowance surface and a theoretical surface of a three-dimensional model of the theoretical bearing bush.
According to the processing technology requirements of the bearing bush, the method mainly comprises the aspects of material selection, processing precision, surface treatment, size requirements, heat treatment, lubricating oil hole design and the like. During machining, machining is required to be performed strictly according to design requirements, and good machining environment and process control are required to be kept in the machining process so as to ensure the quality and performance of the bearing bush. The repair process material is Babbitt metal, and the Babbitt metal has a thermal expansion coefficient of aboutAnd the shrinkage is generally between 1.5% and 3%, and has a lower thermal expansion coefficient and higher thermal conductivity. Therefore, consideration is required to be given to the thermal deformation characteristics of the material and the thermal influence at the time of cutting. Milling is adopted in the bearing bush repairing process, and a certain machining allowance is determined according to the monitored cutter abrasion and cutting temperature; the diameter tolerance in precision requirements is typically tens to hundreds of microns, and the surface roughness is typically in the range of Ra 0.4 to Ra 1.6 microns; and determining the machining allowance of the bearing bush tolerance surface according to the machining process requirements.
Combining with a mechanical manual practical bearing manual, a process flow and empirical data, wherein the diameter of the thin-wall bearing bush is 0-75mm, and the machining allowance of the wall thickness is 0.008+/-0.001 mm; the diameter of the thin-wall bearing bush is 75-110mm, and the wall thickness machining allowance is 0.010mm plus or minus 0.002mm; the diameter of the thin-wall bearing bush is 110-200mm, and the machining allowance of the wall thickness is 0.015+/-0.004 mm; the diameter of the thin-wall bearing bush is 200-250mm, and the machining allowance of the wall thickness is 0.020+/-0.004 mm.
And measuring the free-form surface of the bearing bush machining surface, wherein the acquisition of object surface point cloud data is a precondition and a foundation for realizing the optimal distribution of machining allowance of the blank with the complex integral structure. Aiming at the problems that the surface of the bearing bush after material addition is a free curved surface, the structure is complex and is thin-wall processing, the self shielding possibly exists, the parts needing to be measured and analyzed are many, the thickness is uneven, the accumulated error is easily caused by multi-view measurement data, and the like, the three-dimensional digital comprehensive measurement is carried out on the surface of the bearing bush by adopting an optical scanning means. Before optical scanning, preparation of the measurement area is required; second, the positions of the light source and camera need to be set to ensure that the entire measurement area is adequately illuminated and that the camera is able to capture a clear surface reflection image. The optical scanner uses a light source such as laser or white light to irradiate the surface to be measured to form a light spot. The position and shape of the spot will vary with the curvature of the surface. The cameras capture light reflected from the surface, record topography and color information of the surface, and four cameras are used to capture reflected images of the surface at different angles and positions to obtain more comprehensive surface information. The captured image is processed through data processing software Polyworks, the steps of image registration, denoising, point cloud generation and the like are included, point cloud data are reconstructed and surface fitting is carried out, and an accurate three-dimensional model of the free-form surface to be measured of the bearing bush is generated. The three-dimensional model can include information such as the overall shape of the surface being measured, surface details, and occlusion locations.
The free-form surface measurement of the actual machining surface of the bearing bush comprises the following steps:
Recording a bearing bush image through an optical scanner, acquiring comprehensive surface characteristics, and processing the image to generate point cloud data;
And step two, preprocessing the point cloud for image processing comprises point cloud filtering and point cloud registration. These methods are typically used prior to reconstruction. The statistical filter is mainly used for removing obvious outliers. According to the characteristics, when the density of the point cloud at a certain position is less than a certain threshold value, the point cloud is invalid. In the filtering process, the distance distribution of all points in the point cloud can be constructed by calculating the average distance from each point to the nearest k points, and outlier points are removed by using the designated mean and variance. Assuming that a certain point coordinate in the point cloud is There are k points in the neighborhood, and the point reaches any point/>, in the neighborhoodThe distance of (2) is as follows:
；
traversing and calculating all points in the neighborhood Screening out points meeting normal distribution conditions, and calculating to obtain the average value and standard deviation of normal distribution, wherein the average value and standard deviation are as follows:
；
；
Set standard deviation multiple as Then when the distance between any point in the neighborhood of a point is within the rangeWhen in, the point is reserved. Conversely, if the distance is outside of this range, the point is considered an outlier and is removed.
Step three, point cloud initial registration, an intrinsic shape feature (INTRINSIC SHAPE Signatures, ISS), is a method of describing the local shape of the point cloud. The algorithm treats the point cloud surface as a set of curvature and normal vectors, and then performs statistical analysis on the set.
The steps for statistical analysis of the set of curvature and normal vectors are as follows,
(1) And weighting the base points. To establish a local coordinate system of the base points, the base points and their neighboring points are weighted to overcome the non-uniformity of the point cloud sampling. Specific expressions are shown below, in whichRepresentative pointWeights ofIs toRadius of sphere as center of circle, expressA weighted range;
；
Wherein, Representative pointWeights of (2); /(I)Is toRadius of sphere as center of circle, expressA weighted range; p j represents a dot matrix of dots j;
(2) And analyzing the characteristic value of the base point. At the datum point Eigenvalue analysis is performed at the point riss _frame adjacent to the eigenvalue analysis, and the covariance matrix is calculated as shown in the following formula.
；
(3) Local reference frame. Calculation ofEigenvalues {,And arranged in order of magnitude, feature vectorAlso arranged in order of magnitude. ToIs the origin,And its productFor the x, y, z axes, a local reference frame is established, see fig. 2.
(4) And extracting characteristic points. Setting a threshold valueAndThe points satisfying the formula are called ISS feature points, and the formula is as follows:
；
In the method, in the process of the invention, ,Three eigenvalues representing i-points,AndRepresenting the set threshold.
And step four, the point cloud RANSAC registration, namely the RANSAC algorithm can automatically fit a model in the data, and the position of a corresponding point can be calculated for solving the positioning problem, so that the method can be used for three-dimensional point cloud registration. The sum of squares of the distances between corresponding points in two point clouds is minimized by calculating an initial transformation matrix, and the specific expression is as follows:
；
In the method, in the process of the invention, Representing a transformation matrix; p represents a source point cloud; p represents a matrix of source points Yun Zhongdian; q represents a back point cloud point matrix; /(I)Is an extremum function for T, representing the minimum of the set of points in the function.
The algorithm needs to input a downsampled source point cloud P 'and a target point cloud Q' and their corresponding description subsets, and the main steps are as follows:
(1) Searching for a corresponding point: random selection Personal dotIterating in P ', finding their corresponding points in Q' by characterization。
(2) Calculating a differential vector of the corresponding edge: firstly, calculating Euclidean distance between n sampling points, then calculating the ratio of the difference between two corresponding side lengths to the larger side length to form a differential vector. IfLess than side-length similarity thresholdAnd (5) continuing the next step, otherwise, returning to the step (1). Referring to FIG. 3, whenIn the case of sample,The calculation formula of (2) is as follows:
；
In the method, in the process of the invention, Representing a point/>, in a source point cloud PAndDistance betweenAnd the same is done; /(I)Representing points/>, in the back point cloud QAndDistance betweenAnd the same is true.
(3) Corresponding point transformation: a temporary transformation matrix Ti is estimated from the corresponding pair of points and P 'is converted to Pi'.
(4) Judging an inner point: calculating the nearest neighbor Euclidean distance between Pi 'and Q', and setting the distance smaller than the set distance threshold valueIs taken as an inner point. If the number of interior points is too small, return to step (1).
(5) Calculating a transformation matrix: and calculating a transformation matrix according to the corresponding relation of the inner points. The iteration number k of the RANSAC is determined by a preset success probability p and a preset interior point proportion w, and a specific calculation formula is as follows:
；
in the formula, n represents the number of sampling points, k represents the iteration times, j represents the preset success probability, and w represents the preset interior point proportion.
(6) Model estimation and updating: other points are tested using the interior point fitting parameterized model and the interior points are updated to continue the iterative computation until the best estimate model is obtained or the maximum number of iterations Riter is reached. When the error value is minimumAnd setting the result as the best estimation result, and stopping iteration.
And fifthly, reconstructing point cloud data and fitting a surface, wherein a poisson reconstruction method is used. Poisson reconstruction is a surface reconstruction method based on hidden functions, which can obtain a model with complete detail characteristics and geometric surface characteristics, and has good closure.
(1) Normal estimation and direction normalization: defining a local quadric as. The calculation formula of the first partial derivative of the curved surface at a certain point is as follows:
；
Wherein, . The calculation formula of the normal direction of the curved surface at this point is shown as follows. WhereinIs a surface equation coefficient.
；
The normal direction calculated based on the moving least square method has ambiguity, i.e., only the straight line where the normal is located can be determined and the pointing direction cannot be determined. Therefore, a minimum spanning tree (SPANNING TREE, MST) is required to perform a direction unification process on the normals in the model to ensure its accuracy. The MST algorithm defines a cost function to act on the three-dimensional point cloud, and the specific expression is as follows:
；
In the method, in the process of the invention, Representing the adjacent point,Representing the slave pointsPointing PointIs a unit vector of (a); /(I)AndRespectively represent pointsSum dotIs defined by the normal to (d).
(2) Introducing a shielding factor: for some error reasons, the indicator function in poisson's equation may shift, resulting in non-watertight surface connection at the edge portion, thereby producing a large number of pseudo-closed surfaces. In poisson's equation, the directed function is solved by solving the directed point set so that an optimal approximation is formed between the directed point set and the function gradient, i.e. by minimizing the scale function, the following equation is solved:
；
In the method, in the process of the invention, Representing a scale function; /(I)Representing the functional gradient of h; /(I)Representing a set of directed points; /(I)Is thatIn shorthand form, represents the scale between the set of directed points and the gradient of the function.
Designating a set of points with weightsAnd adding a gradient constraint and a value constraint of a discrete point in the scale function minimization problem to calculate a sample point function, wherein a constraint equation is shown as follows.
；
In the method, in the process of the invention,Representing sample point weights; s represents a corresponding point set; /(I)Representing the area of the region defined by the set of points S; /(I)A value constraint representing a discrete point.
For ease of calculation, each sample point is weightedLet 1, the above formula reduces to:
；
In the method, in the process of the invention, Representing sample point weights; s represents a corresponding point set; /(I)A value constraint representing a discrete point.
(3) Boundary constraint: the boundary constraints are used to optimize the shape and smoothness of the reconstructed surface to reduce offset errors. To satisfy the solution of the mask poisson equation and the boundary condition in the specified region, we introduce a boundary constraint condition when solving the partial differential equation. Here, we use Robin boundary constraints, comprehensively considering Dirichlet and Neumann boundary constraints, and considering function values and normal vectors of the point cloud to process the point cloud with hybrid characteristics.
In order to ensure that the point cloud data can be correctly corresponding to the curved surface of the CAD model, curved surface matching is needed, and finally, a processing positioning reference is determined.
The measurement point cloud data and the CAD model are respectively in different coordinate systems, and thus pre-alignment is required for the segmentation. Typically, the best curve alignment requires two stages of initial matching and exact matching. The initial matching can reduce rotation and translation errors between coordinate systems and improve the efficiency of the subsequent accurate matching. The accurate matching is to finely adjust the gesture of the model based on the initial matching so as to meet the best matching requirement in actual requirements. Before surface matching is performed, features need to be extracted from the point cloud data and the CAD model in order to perform matching. These features may include curvature of the surface, normal vectors, feature points, and the like. Feature extraction can be achieved by surface fitting, curvature analysis, and the like. The core of the surface matching is the registration of the surface, namely, the surface features in the point cloud data are corresponding to the surface features in the CAD model. This typically requires the use of registration algorithms, such as least squares registration, feature matching algorithms, and the like. The goal of the registration is to find the optimal transformation matrix, aligning the surface in the point cloud data with the surface in the CAD model. Once the surface matching is completed, a tooling positioning reference can be determined. This means that the correspondence between the point cloud data and the CAD model is determined, as well as the coordinate system conversion and positioning information required during processing. This information will be used for subsequent process path planning and process parameter optimization.
Step one, the point cloud data is initially matched with a CAD model. The blank point cloud to be matched and the design digital model are placed in the same system coordinate system by adopting a three-point method, and three corresponding points are respectively taken from the blank point cloud and the design digital modelAndConstructing vector units:
；
；/>
In the method, in the process of the invention, 、Is an intermediate vector for solving、；
；
Unit vectorAndTwo local coordinate systems are formed, and after Euclidean transformation, the two coordinate systems are completely overlapped. Thus rotate matrixAnd translation vectorThe method comprises the following steps of:
；
；
after the coordinate system is adjusted, the original point cloud is basically coincident with the design model, and the aim of initial matching is achieved.
Step two, we perform further unconstrained matching. The accuracy of point cloud chunking depends to a large extent on the relative position between the measured point cloud data and the CAD model. To reduce computation time for unconstrained matching, we acquire a simplified set of points in a uniformly sampled fashion and solve iteratively using ICP. Assuming that the ICP results in a rotation matrix ofTranslation vector is. The following are the specific steps of the algorithm:
(1) Initializing a rotation matrix Is an identity matrix, translation vectorZero vector, iteration numberSet to 0 and set the maximum iteration number。
(2) For a given set of reduced pointsFinding the nearest point/>, on a CAD modelWherein。
(3) Minimization ofClosed solving of the rotation matrix/>, in this iterationAnd translation vector. Wherein:
；
(4) By using AndFor measurement data pointsCoordinate transformation is performed andAnd。
；
(5) If it is，OrThen go to step 6; no makeAnd (2) continuing the step (2). WhereinAndTo allow for errors.
(6) If it isThe algorithm is terminated, the ICP is converged to the local optimum, and the optimal alignment position is output; otherwise, the iteration is finished, and the current alignment position is output.
And thirdly, performing automatic blocking of the point cloud. Firstly, a process semantic guide model is established as a block template, then 'seed points' of corresponding structural features on a blank are quickly searched according to guide points on each curved surface in the guide model, and all compatible points are added into corresponding data blocks through a region growing algorithm. The specific steps of the algorithm are as follows:
Step1: and calculating the normal vector, curvature and other local differential geometric properties of each measuring point in the point cloud data.
Step2: traversing each face in the process semantic guide model, and searching guide points on the current faceAnd calculates the normal vector/>, at that point. Then find the distance point/>, in the point cloud modelNearest pointAs a preliminary seed point for partitioning the corresponding face on the point cloud.
Step3: calculating the point cloud model at the seed pointNormal vector at. IfThen byTurning to Step5 as a seed point of the segmentation block; otherwise go to Step4.
Step4: prompting seed point search error, then toIs an offset element, toFor the offset direction, the guide points on the curved surface are subjected to iterative offset with equal step length. Then find the nearest point/>, to the offset point, in the point cloud modelGo to Step3.
Step5: and growing around the seed point as the center, and judging whether the adjacent points have similar geometric or technological structural characteristics with the seed point according to the local differential information (normal vector and curvature change). If the conditions are met, continuing to grow, and turning to Step5; if the condition is not met, the growth is stopped and the process goes to Step6.
Step6: and sequentially outputting the point cloud data blocks and the corresponding CAD model curved surface pieces. If the undivided block exists, the method goes to Step2 to continue segmentation; otherwise the algorithm ends.
Through the algorithm, not only is the automatic partitioning of the point cloud data realized, but also the one-to-one correspondence between each partitioned data and the corresponding curved surface piece on the CAD model is established. In this way, a minimum margin constraint can be applied to the measured data of each surface to be processed, and a different tolerance constraint can be applied to each tolerance surface.
The self-adaptive machining algorithm can automatically adjust the machining path and the machining parameters according to the actual measurement data so as to realize accurate material reduction machining of the bearing bush. After the matched CAD curved surface model is obtained, curved surface analysis is needed, including curvature analysis, normal vector calculation, curved surface feature extraction and the like. These analyses can help determine local features of the surface, providing basis for adaptive processing algorithms. Based on the result of the curved surface analysis, the self-adaptive processing algorithm can plan a processing path, and the motion trail and the processing parameters of the cutter are automatically adjusted to adapt to curved surfaces with different curvatures and different shapes. This step requires consideration of machining precision requirements, surface characteristics, tool size, etc. The self-adaptive machining algorithm can automatically adjust machining parameters including cutting speed, feeding speed, cutting depth and the like according to local characteristics of the curved surface and machining requirements. The method can help to realize accurate machining of different curved surface characteristics, and improves machining efficiency and machining quality. In the processing process, the self-adaptive processing algorithm can monitor the processing state in real time, and dynamically adjust the processing parameters according to the actual processing conditions so as to ensure the processing precision and stability. In actual software development, these steps may be implemented in the programming language Python, the following being a simplified pseudocode example:
# pseudo code example
# Data acquisition and preprocessing
point_cloud = get_point_cloud_data()、
processed_data = preprocess_point_cloud(point_cloud)
# Surface reconstruction and fitting
surface_model = reconstruct_surface(processed_data)
# Surface analysis
surface_properties = analyze_surface(surface_model)
# Machining path planning
tool_path = generate_tool_path(surface_model, surface_properties)
Numerical control machining
start_machining(tool_path)
One or more technical schemes provided in the patent have at least the following technical effects or advantages:
and (3) high-efficiency processing: the self-adaptive machining algorithm can dynamically adjust the machining path and the cutting parameters of the cutter according to the local curvature and the shape characteristics of the surface of the bearing bush, so that unnecessary cutting is reduced to the greatest extent, and the machining efficiency is improved. This means that when the bearing bush with complex curved surface is processed, the processing time can be reduced, the productivity can be improved, and the low-efficiency processing mode of the traditional manual scraping shoe is changed.
And (3) accurate processing: by performing curvature analysis and feature extraction on the bearing bush surface, a more accurate machining can be achieved by the adaptive machining algorithm. The method can adjust the processing path and the cutting parameters according to the uneven curvature and the complex shape of the curved surface, thereby improving the processing precision, reducing the processing error and ensuring that the final product meets the design requirement.
Optimizing the service life of the cutter: the self-adaptive machining algorithm can adjust cutting parameters according to the local characteristics of the surface of the bearing bush, and reduce the abrasion of a cutter in the cutting process. This not only prolongs the service life of the tool, but also reduces the frequency of tool replacement, thereby reducing the production cost.
The cost is reduced: the self-adaptive machining algorithm can reduce the overall cost of bearing bush machining by improving machining efficiency and precision and optimizing the service life of a cutter. This includes reduced machining time, reduced scrap rate, reduced tooling costs and maintenance costs, thereby improving production efficiency.
The adaptability is strong: the self-adaptive processing algorithm can automatically adjust the processing path and parameters according to the different shapes and curvature characteristics of the bearing bush surface, so that the self-adaptive processing algorithm is suitable for bearing bush processing requirements of different shapes and curvatures. This increases the flexibility and adaptability of the process, making the process more intelligent and automated.
And (3) digital production: the self-adaptive processing algorithm is combined with numerical control processing equipment, so that the digital production of the bearing bush processing is realized. This means that the processing course is more accurate and controllable, reduces the influence of human factors on the processing precision, and ensures that the production process is more intelligent and efficient.
Claims (8)
1. The self-adaptive processing method for the material-adding repairing bearing bush is characterized by comprising the following steps of: the method comprises the following steps:
S1, determining machining allowance of a bearing bush tolerance surface according to machining process requirements of the bearing bush;
S2, measuring a free-form surface of an actual machining surface of the bearing bush to obtain surface parameters, and generating an accurate point cloud model of the measured free-form surface of the bearing bush by a point cloud processing method;
s3, performing surface matching on the point cloud model and the CAD model, and determining a machining positioning reference to obtain a CAD theoretical model;
S4, comparing the actual model of the bearing bush obtained by online measurement with a theoretical model to determine deviation, and automatically adjusting a processing path and processing parameters through a self-adaptive processing algorithm to realize accurate processing of the bearing bush;
In S2, the free-form surface measurement of the actual machined surface of the bearing shell includes the steps of,
S21, recording a bearing bush image through an optical scanner, acquiring comprehensive surface characteristics, and processing the image to generate point cloud data;
S22, filtering point cloud; the average distance from each point to k nearest points is calculated, the distance distribution of all points in the point cloud is constructed, and outlier points are removed by means of the average value and the variance;
wherein if a certain point coordinate in the point cloud is K points are in the neighborhood, and the point reaches any point in the neighborhoodDistanceThe following formula is shown:
；
traversing and calculating all points in the neighborhood Screening out the points meeting the normal distribution condition, and calculating to obtain the average value/>, of the normal distributionAnd standard deviationThe method is specifically as follows:
；
；
Set standard deviation multiple as Then when the distance between any point in the neighborhood of a point is within the rangeWhen in, the point is reserved; conversely, if the distance is outside the range, the point is considered to be an outlier and is removed;
s23, registering the point cloud; the point cloud surface is regarded as a set of curvature and normal vector, and then statistical analysis is carried out on the set of curvature and normal vector;
S24, registering the point cloud RANSAC; calculating an initial transformation matrix by an algorithm The square sum of the distances between the corresponding points in the two point clouds is minimized, and the specific expression is as follows,
;
In the method, in the process of the invention,Representing a transformation matrix; p represents a source point cloud; p represents a matrix of source points Yun Zhongdian; q represents a back point cloud point matrix; The extremum function of T represents the minimum value of the point set in the function;
S25, reconstructing point cloud data and fitting the surface by using a poisson reconstruction method to obtain an accurate point cloud model with complete detail characteristics and geometric surface characteristics.
2. The adaptive machining method of the additive repairing bearing bush according to claim 1, wherein the method comprises the following steps of: in S1, according to the machining process requirements of the bearing bush, including material selection, machining precision, surface treatment, size requirements, heat treatment and design of a lubricating oil hole, and in combination with the thermal deformation characteristics of the material and the thermal influence during cutting, the bearing bush repairing process adopts milling, and simultaneously, the bearing bush tolerance surface machining allowance is determined according to the monitored cutter wear and cutting temperature.
3. The adaptive machining method of the additive repairing bearing bush according to claim 1, wherein the method comprises the following steps of: in S3, constructing a CAD model according to the original design size of the bearing bush, extracting features from the point cloud model and the CAD model, wherein the features can comprise curvature, normal vector and feature points of a curved surface, the curved surface features in the point cloud data and the curved surface features in the CAD model are required to be corresponding in curved surface registration, an optimal conversion matrix is found, the curved surface in the point cloud data and the curved surface in the CAD model are aligned, and curved surface matching is completed, so that a processing positioning reference is determined.
4. The adaptive machining method of the additive repairing bearing bush according to claim 1, wherein the method comprises the following steps of: in S4, the deviation is determined by comparing the actual model measured on line with the CAD theoretical model, the self-adaptive processing algorithm automatically adjusts processing parameters to eliminate or reduce the deviation according to the local characteristics of the curved surface and the processing requirements, including cutting speed, feeding speed and cutting depth, so that accurate processing of different curved surface characteristics is realized, in the processing process, the self-adaptive processing algorithm monitors the processing state in real time, and dynamically adjusts the processing parameters according to the actual processing conditions, so that the processing precision and stability are ensured.
5. The adaptive machining method of the additive repairing bearing bush according to claim 1, wherein the method comprises the following steps of: in S23, the statistical analysis of the set of curvatures and normal vectors is performed as follows,
S231, weighting the base points; in order to establish a local coordinate system of the base point, the base point and its neighboring points are weighted to overcome the non-uniformity of the point cloud sampling, the specific expression is as follows,
；
Wherein,Representative pointWeights of (2); /(I)Is toRadius of sphere as center of circle, expressA weighted range; p j represents a dot matrix of dots j;
s232, analyzing a base point characteristic value; at the datum point Eigenvalue analysis is performed at the adjacent point riss _frame, the formula of the covariance matrix is shown in the following formula,
；
S233, a local reference system; calculation ofEigenvalues {,And arranged in order of magnitude, feature vectorsAre also arranged in order of size toIs the origin,And its productEstablishing a local reference system for x, y and z axes;
S234, extracting feature points; setting a threshold value AndThe points that satisfy the formula are called ISS feature points, where the formula is as follows:
；
In the method, in the process of the invention, ,Three eigenvalues representing i-points,AndRepresenting the set threshold.
6. The adaptive machining method of the additive repairing bearing bush according to claim 1, wherein the method comprises the following steps of: in S24, the algorithm needs to input the downsampled source point cloud P 'and target point cloud Q' and their corresponding description subsets, and the main steps are as follows:
s241, searching for a corresponding point; random selection Personal dotIterating in P ', finding their corresponding points in Q' by characterization；
S242, calculating a differential vector of the corresponding edge; firstly, calculating Euclidean distance between n sampling points, then calculating the ratio of the difference between two corresponding side lengths to the larger side length to form a differential vector; IfLess than side-length similarity thresholdContinuing the next step, otherwise returning to S241; wherein, whenIn the case of sample,The calculation formula of (2) is as follows:
；
In the method, in the process of the invention, Representing a point/>, in a source point cloud PAndDistance betweenAnd the same is done; /(I)Representing points/>, in the back point cloud QAndDistance betweenAnd the same is done;
S243, corresponding point transformation; estimating a temporary transformation matrix Ti from the corresponding pair of points and converting P 'to Pi';
s244, judging an inner point; calculating the nearest neighbor Euclidean distance between Pi 'and Q', and setting the distance smaller than the set distance threshold value Is taken as an inner point; if the number of interior points is too small, returning to S241;
S245, calculating a transformation matrix; calculating a transformation matrix according to the corresponding relation of the inner points; the iteration number k of the RANSAC is determined by a preset success probability p and a preset interior point proportion w, and a specific calculation formula is as follows:
；
In the method, in the process of the invention, Representing the number of sampling points,Representing the number of iterations,Representing a preset success probability,Representing a preset interior point proportion;
s246, model estimation and updating; testing other points using the interior point fitting parameterized model and updating the interior points to continue iterative computation until the best estimated model is obtained or the maximum number of iterations Riter is reached, when the error value is minimal And setting the result as the best estimation result, and stopping iteration.
7. The adaptive machining method of the additive repairing bearing bush according to claim 1, wherein the method comprises the following steps of: in S3, the step of performing surface matching between the point cloud model and the CAD model is as follows,
S31, the point cloud data is initially matched with the CAD model; placing the blank point cloud to be matched and the CAD model in the same system coordinate system by adopting a three-point method, and respectively taking three corresponding points on the blank point cloud and the CAD modelAndConstructing vector units:
；
；
In the method, in the process of the invention, 、Is an intermediate vector for solving、；
；
Unit vectorAndTwo local coordinate systems are formed, after Euclidean transformation, the two coordinate systems are completely overlapped, and then the matrixAnd translation vectorThe method comprises the following steps of:
；
；
after the coordinate system is adjusted, the original point cloud is basically overlapped with the CAD model, and the initial matching target is achieved;
s32, further unconstrained matching; obtaining a simplified point set in a uniform sampling mode, and solving by using ICP iteration;
S33, performing automatic blocking of the point cloud; firstly, a process semantic guide model is established as a block template, then seed points of corresponding structural features on a blank are quickly searched according to guide points on each curved surface in the guide model, and all compatible points are added into corresponding data blocks through a region growing algorithm, wherein the specific steps of the region growing algorithm are as follows:
S331, calculating normal vector and curvature local differential information of each measuring point in the point cloud data;
s332, traversing each surface in the process semantic guide model, and searching guide points on the current surface And calculates the normal vector/>, at that pointThen find the distance point/>, in the point cloud modelNearest pointAs a preliminary seed point for the corresponding face on the split point cloud;
S333, calculating the point cloud model at the seed point Normal vector at; IfThen byAs a seed point of the divided block, go to S335; otherwise go to S334;
s334, prompting the seed point to search for errors, and then Is an offset element, toFor the offset direction, performing equal-step iterative offset on the guide point on the curved surface, and then finding the nearest point/> to the offset point in the point cloud modelGo to S333;
S335, growing around the seed point as the center, judging whether the adjacent points have similar geometric or technological structural characteristics with the seed point according to the local differential information, if so, continuing growing, and turning to S335; if the condition is not satisfied, stopping the growth, and turning to S336;
s336, sequentially outputting the point cloud data blocks and the CAD model curved surface pieces corresponding to the point cloud data blocks, and if the point cloud data blocks are not segmented, turning to S332 to continue segmentation; otherwise the algorithm ends.
8. The adaptive machining method of the additive repairing bearing bush according to claim 7, wherein the method comprises the following steps of: in S32, if the rotation matrix obtained by ICP isTranslation vector isThe ICP iterative solution steps are as follows:
s321, initializing a rotation matrix Is an identity matrix, translation vectorZero vector, iteration numberSet to 0 and set the maximum iteration number；
S322, for a given reduced point setFinding the nearest point/>, on a CAD modelWherein；
S323, minimizeClosed solving of the rotation matrix/>, in this iterationAnd translation vector; Wherein:
；
S324, use AndFor measurement data pointsCoordinate transformation is performed andAnd；
；
In the method, in the process of the invention,A rotation matrix representing k iterations; /(I)A translation vector representing k iterations; /(I)A distance value representing the rotation matrix and the translation vector;
S325, if ，OrThen go to S326; no makeContinuing S322;
Wherein the method comprises the steps of AndIs an allowable error;
S326, if The algorithm is terminated, the ICP is converged to the local optimum, and the optimal alignment position is output; otherwise, the iteration is finished, and the current alignment position is output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410418027.6A CN118003031B (en) | 2024-04-09 | 2024-04-09 | Self-adaptive machining method for material-adding repairing bearing bush |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410418027.6A CN118003031B (en) | 2024-04-09 | 2024-04-09 | Self-adaptive machining method for material-adding repairing bearing bush |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118003031A CN118003031A (en) | 2024-05-10 |
| CN118003031B true CN118003031B (en) | 2024-06-14 |
Family
ID=90954441
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410418027.6A Active CN118003031B (en) | 2024-04-09 | 2024-04-09 | Self-adaptive machining method for material-adding repairing bearing bush |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118003031B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106203726A (en) * | 2016-07-21 | 2016-12-07 | 中国民用航空飞行学院 | Precision component surface defect repair system and restorative procedure |
| CN109202378A (en) * | 2018-08-30 | 2019-01-15 | 大连交通大学 | A kind of increase and decrease composite intelligent restorative procedure of metal parts |
| CN109345523A (en) * | 2018-09-21 | 2019-02-15 | 中国科学院苏州生物医学工程技术研究所 | Surface defects detection and three-dimensional modeling method |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104674210A (en) * | 2013-12-02 | 2015-06-03 | 齐欢 | Workpiece laser automatic repair method |
| CN105608733B (en) * | 2015-12-23 | 2018-04-03 | 大连海洋大学 | A kind of high-precision 3D printing restorative procedure of complex curved surface parts |
| CN105598445B (en) * | 2015-12-23 | 2017-08-08 | 大连海洋大学 | A kind of high-precision 3D printing system for repairing metal parts |
| CN107856309B (en) * | 2017-11-06 | 2020-04-28 | 沈阳航空航天大学 | Rapid path planning method for additive manufacturing and remanufacturing |
| CN109118589A (en) * | 2018-09-07 | 2019-01-01 | 北京航空航天大学青岛研究院 | Geometrical model full range details restorative procedure |
| DE102019121947B3 (en) * | 2019-08-14 | 2020-12-10 | Sauer Gmbh | DEVICE AND METHOD FOR REPAIR OF COMPONENTS BY ADDITIVE MANUFACTURING |
| CN110782521A (en) * | 2019-09-06 | 2020-02-11 | 重庆东渝中能实业有限公司 | Mobile terminal three-dimensional reconstruction and model restoration method and system |
| EP3933527A1 (en) * | 2020-07-03 | 2022-01-05 | Siemens Aktiengesellschaft | Method and assembly for repairing a workpiece |
| CN112815843B (en) * | 2021-01-07 | 2023-12-29 | 西安理工大学 | On-line monitoring method for printing deviation of workpiece surface in 3D printing process |
| CN112981395A (en) * | 2021-02-18 | 2021-06-18 | 中国长江电力股份有限公司 | Thrust pad repairing method based on laser cladding process |
| CN116373304A (en) * | 2022-12-27 | 2023-07-04 | 西北工业大学 | In-situ repair path planning method for composite damaged member |
-
2024
- 2024-04-09 CN CN202410418027.6A patent/CN118003031B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106203726A (en) * | 2016-07-21 | 2016-12-07 | 中国民用航空飞行学院 | Precision component surface defect repair system and restorative procedure |
| CN109202378A (en) * | 2018-08-30 | 2019-01-15 | 大连交通大学 | A kind of increase and decrease composite intelligent restorative procedure of metal parts |
| CN109345523A (en) * | 2018-09-21 | 2019-02-15 | 中国科学院苏州生物医学工程技术研究所 | Surface defects detection and three-dimensional modeling method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118003031A (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111915730B (en) | Method and system for automatically generating indoor three-dimensional model by taking semantic slave point cloud into consideration | |
| CN111077844B (en) | Part accurate machining method based on measured data feature guidance | |
| CN111489432B (en) | Bent pipe reconstruction and allowance calculation method based on point cloud data | |
| Li et al. | An automatic and accurate method for tool wear inspection using grayscale image probability algorithm based on bayesian inference | |
| CN116402866A (en) | Point cloud-based part digital twin geometric modeling and error assessment method and system | |
| Ghorbani et al. | Construction of damage-free digital twin of damaged aero-engine blades for repair volume generation in remanufacturing | |
| CN112697058A (en) | Machine vision-based large-size plate assembly gap on-line measurement system and method | |
| Feng et al. | A novel feature-guided trajectory generation method based on point cloud for robotic grinding of freeform welds | |
| CN116204990B (en) | Three-dimensional measured data driven precise coordination repair method for large-scale framework of aircraft | |
| CN114868128A (en) | Computer-implemented method for personalizing spectacle frame elements by determining a parameterized replacement model of spectacle frame elements, and device and system using such a method | |
| CN114972640B (en) | Cable three-dimensional reconstruction and parameter calculation method based on point cloud | |
| Yue et al. | A new plane segmentation method of point cloud based on mean shift and RANSAC | |
| CN116664889A (en) | Aircraft flap skin repair allowance determination method | |
| CN115797414A (en) | Complex curved surface measurement point cloud data registration method considering measuring head radius | |
| CN118003031B (en) | Self-adaptive machining method for material-adding repairing bearing bush | |
| Mu et al. | A point cloud registration-based calibration algorithm for robot offline programming automatic loading in aero-grinding applications | |
| CN112581521B (en) | Method for extracting central line of magnetic suspension track | |
| CN118081767A (en) | Automatic programming system and method for post-processing machining of casting robot | |
| CN116841246A (en) | Robot polishing path automatic planning method based on three-dimensional point cloud data | |
| CN115641418A (en) | Radar cover sub-region self-adaptive modification method based on point cloud information | |
| Zou et al. | Feature-based optimization analysis technology of precision casting machining allowance | |
| CN113792359B (en) | Method for reconstructing quasi-scattered point cloud data curved surface of aero-engine blade | |
| CN118628474B (en) | Industrial automatic three-dimensional detection system and method | |
| Raible et al. | Automatic Path Planning for Robotic Grinding and Polishing Tasks based on Point Cloud Slicing | |
| CN118171503B (en) | Method for coordinating canopy based on point cloud measured data virtual assembly |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |