CN118288107A - Free-form surface machine tool contact type in-situ detection error compensation method - Google Patents
Free-form surface machine tool contact type in-situ detection error compensation method Download PDFInfo
- Publication number
- CN118288107A CN118288107A CN202410569195.5A CN202410569195A CN118288107A CN 118288107 A CN118288107 A CN 118288107A CN 202410569195 A CN202410569195 A CN 202410569195A CN 118288107 A CN118288107 A CN 118288107A
- Authority
- CN
- China
- Prior art keywords
- measurement
- error
- measuring
- sampling
- measuring head
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 98
- 238000001514 detection method Methods 0.000 title claims abstract description 36
- 238000011065 in-situ storage Methods 0.000 title claims abstract description 33
- 238000005259 measurement Methods 0.000 claims abstract description 146
- 238000012549 training Methods 0.000 claims abstract description 24
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 18
- 238000004088 simulation Methods 0.000 claims abstract description 11
- 230000008569 process Effects 0.000 claims abstract description 10
- 238000004441 surface measurement Methods 0.000 claims abstract description 6
- 238000005070 sampling Methods 0.000 claims description 101
- 230000036544 posture Effects 0.000 claims description 16
- 238000012625 in-situ measurement Methods 0.000 claims description 14
- 230000003044 adaptive effect Effects 0.000 claims description 11
- 238000012795 verification Methods 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 5
- 238000011176 pooling Methods 0.000 claims description 5
- 238000002474 experimental method Methods 0.000 abstract description 3
- 241001422033 Thestylus Species 0.000 description 6
- 239000000523 sample Substances 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 230000006872 improvement Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000010979 ruby Substances 0.000 description 2
- 229910001750 ruby Inorganic materials 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000003754 machining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
Landscapes
- Length Measuring Devices With Unspecified Measuring Means (AREA)
Abstract
The invention discloses a free-form surface machine tool contact type in-situ detection error compensation method, which comprises the following steps: s1: obtaining a measurement initial compensation value by a precompensation method; s2: acquiring an actual measurement error value and an actual measuring head posture; s3: training a convolutional neural network structure to obtain a curved surface measurement error prediction model; s4: performing measurement error compensation to obtain a final measurement coordinate value; s5: verifying the measurement accuracy through simulation measurement and actual measurement; the measuring error compensation method does not need to go deep into an error source, is simple in error identification method and high in operability, and avoids complex error compensation modeling and error decoupling processes; the measurement error compensation only needs three-coordinate measurement equipment, does not need a large amount of complex experiments and a plurality of expensive special measurement equipment to participate, and can save time and capital cost.
Description
Technical Field
The invention relates to a free-form surface machine tool contact type in-situ detection error compensation method, and belongs to the technical field of digital manufacturing and geometric measurement.
Background
The free curved surface has complex geometric shape, the processing precision of the free curved surface is difficult to ensure in the traditional processing, and in-situ detection is required to be introduced in the processing process to obtain geometric measurement data in the processing process so as to facilitate the follow-up correction processing. The in-situ detection system adopting the contact type measuring head is an important means for acquiring geometric information in the numerical control machining process of the workpiece, and particularly, the measurement is realized by installing the contact type measuring head on a main shaft of a machine tool. The in-situ detection method is characterized in that the end part of the measuring needle of the measuring head contacts a workpiece through controlling the multi-axis action of the machine tool to obtain a trigger signal, and the space coordinates when the trigger is recorded and used as a measurement result. The final measurement accuracy is limited to a certain extent by factors such as manufacturing and assembly errors of the shafts on the machine tool, a pre-stroke error generated by the measurement principle of the contact probe, and the like.
The measurement error compensation method is a necessary means for improving the in-situ detection accuracy. In the existing measuring error compensation methods, some methods adopt a laser tracker, a laser interferometer, an R-test, a double-club instrument and the like to identify and compensate errors of a linear shaft and a rotating shaft of a numerical control machine tool so as to improve measuring accuracy; other methods are used for triggering measurement at a position of a standard component such as a standard ball, a standard block and the like, which is planned in advance, by controlling the measuring head, and obtaining a measurement error condition by comparing the deviation condition of the measured size and the actual size of the standard component. The existing method mainly starts from the source of generating measurement errors, and realizes the integral measurement error improvement through respective compensation. However, the defects of the existing method cannot be ignored, the error compensation usually needs to be performed by a large number of experiments, and a plurality of expensive special instruments and equipment are needed to cooperate to complete the error compensation, and meanwhile, a complex measurement error compensation model is needed to be established and corresponding solution is needed to be performed.
Disclosure of Invention
The invention aims to: aiming at the defects existing in the prior art, the invention provides a free-form surface machine tool contact type in-situ detection error compensation method.
The technical scheme is as follows: a free-form surface machine tool contact type in-situ detection error compensation method comprises the following steps:
S1: obtaining a measured initial compensation value by a precompensation method: setting sampling points on a standard sphere, calibrating and measuring the sampling points by adopting different measuring head postures, comparing a measuring result with the actual size of the standard sphere to obtain a measuring error, and correspondingly obtaining an initial measuring error compensation value from a standard database by the measuring error; inputting the initial measurement error compensation value into a measuring head system to form preliminary compensation;
S2: acquiring an actual measurement error value and an actual measuring head posture;
S3: training a convolutional neural network structure according to the actual measuring head posture and the actual measurement error value in the S2, acquiring a curved surface measurement error prediction model, and predicting the measurement error of the curved surface to be measured by adopting the curved surface measurement error prediction model;
s4: performing measurement error compensation according to the measurement error predicted value obtained in the step S3 to obtain a final measurement coordinate value;
s5: and verifying the measurement accuracy through simulation measurement and actual measurement.
Preferably, the step S1 includes:
S101: setting n sampling points on a standard sphere;
s102: fitting the sampling points in the step S101 to obtain a spherical center coordinate C (x 0,y0,z0);
s103: calibrating and measuring the standard ball by adopting different positions and postures of the measuring head to obtain actual measurement coordinates of each sampling point;
s104: correcting the sphere center position C New;
s105: obtaining a measurement error value Err i;
s106: the initial measurement error compensation value of each direction corresponding to the error value Err i is obtained in the standard database, and the initial measurement error compensation value is input into the measuring head system to form initial compensation.
Preferably, the step S101 specifically includes:
the contact point of the standard ball surface and the extension line of the axis of the measuring head is set as an initial sampling point, the included angle between adjacent sampling points is theta, and the included angles between each sampling point and the axis of the measuring head are respectively Setting the values of m and n according to the calibration precision requirement,M represents the number of standard sphere surface rings.
Preferably, the step S104 is specifically:
the distance from the coordinate value of the initial sampling point to the sphere center is the sum d s of the standard sphere radius R b and the radius R p of the measuring head:
ds=Rb+rp
Assuming that the measured coordinate point obtained at the initial sampling point is P 1(x1,y1,z1), the corrected center position C New is:
Wherein, Representing the vector from point C to point P 1.
Preferably, the step S105 is specifically:
calculating the distance d i between C New and each of the sampling point measurement coordinates { P i(xi,yi,zi) |i=1, 2, …, m×n }, the error value Err i is:
Preferably, the S2 specifically is:
Setting sampling points on a curved surface to be measured, and measuring measurement data of the sampling points by using a three-coordinate measuring machine as reference data, wherein the measurement data is measured by using a standard measuring head; measuring actual measurement data of a sampling point by adopting an in-situ measurement method, wherein the measuring head subjected to compensation in the step S1 is adopted for measurement; the difference between the actual measurement data and the reference data is the measurement error value of each sampling point in the X, Y, Z three-coordinate direction;
And (3) planning a measuring path of a curved surface sampling point according to the interference condition and the sampling efficiency of the measuring head and the workpiece, and recording measuring position information, namely five-axis movement information of the machine tool X, Y, Z, B, C and I, J, K direction information of the axis of the measuring head, so as to obtain the posture of the measuring head when measuring each sampling point.
Preferably, the step S3 specifically includes:
The convolutional neural network structure comprises an input layer, a convolutional layer, a ReLU activation function layer, a pooling layer, a full connection layer and an output layer; the method comprises the steps of taking the gesture of a measuring head as an input parameter, taking a measurement error value as an output parameter, training a convolutional neural network structure together, obtaining a curve measurement error prediction model after training, and measuring and predicting a curve to be measured by adopting the curve measurement error prediction model to obtain measurement errors delta X, delta Y and delta Z of the curve to be measured.
Preferably, the step S4 specifically includes:
and obtaining the sum of error prediction results DeltaX, deltaY and DeltaZ in the sampling point actual coordinate values XYZ and S3 obtained by in-situ measurement, namely the final measurement coordinate value.
Preferably, the simulation measurement of S5 specifically includes:
The method comprises the steps of respectively obtaining dense point cloud and sparse point cloud by adopting an equal-divided curved surface parameter domain method and an adaptive method, taking a result of sampling by adopting an equal-parameter method as a training set, taking a sample obtained by adopting the adaptive sampling method as a verification set, obtaining sampling points by adopting the equal-parameter sampling method, obtaining the sampling points by adopting the adaptive sampling method, and using the sampling points obtained by adopting the equal-parameter sampling method for convolutional neural network structure training so as to generate error distribution experience, wherein the training process obtains measurement experience of in-place detection in a working space of a machine tool, and the sampling points obtained by adopting the characteristic-based sampling method are used as the verification set for testing the prediction effect of the error compensation method on measurement errors.
Preferably, the actual measurement of S5 is specifically:
Measuring the actual curved surface by adopting a three-coordinate measuring machine and in-situ measurement respectively to obtain measurement results, and comparing the curved surface measured in-situ measurement with the curved surface measured by the three-coordinate measuring machine to obtain in-situ detection errors, thereby obtaining maximum errors and average errors before the in-situ measurement system compensates; and calculating the maximum error and average error of the in-situ measurement according to the measurement error after the prediction and compensation of the trained convolutional neural network structure.
The beneficial effects are that: the measuring error compensation method does not need to go deep into an error source, is simple in error identification method and high in operability, and avoids complex error compensation modeling and error decoupling processes; the measurement error compensation only needs three-coordinate measurement equipment, does not need a large amount of complex experiments and a plurality of expensive special measurement equipment to participate, and can save time and capital cost.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a flow chart of the overall process of the method of the present invention;
FIG. 3 (a) is a view perpendicular to the plane formed by the stylus axis and the standard club axis;
FIG. 3 (b) is a schematic diagram showing the angle θ between two standard points on the surface ring of the standard sphere;
FIGS. 4 (a-c) are schematic diagrams of gauge head calibration;
FIG. 5 is a measurement head calibration result;
FIG. 6 is a measurement position versus error correspondence;
FIG. 7 is a convolutional neural network structure;
FIG. 8 is a simulation process training sample and verification sample;
FIGS. 9 (a-b) are simulation results of measurement error compensation;
FIG. 10 shows training samples and verification samples for the actual measurement process;
FIG. 11 is a schematic view of free-form surface measurement;
FIG. 12 is a graph of measurement error before compensation;
fig. 13 is a graph of measurement error after compensation.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. indicate orientations or positional relationships based on the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present invention.
In the present invention, unless expressly stated or limited otherwise, a first feature "above" or "below" a second feature may include both the first and second features being in direct contact, as well as the first and second features not being in direct contact but being in contact with each other through additional features therebetween. Moreover, a first feature being "above," "over" and "on" a second feature includes the first feature being directly above and obliquely above the second feature, or simply indicating that the first feature is higher in level than the second feature. The first feature being "under", "below" and "beneath" the second feature includes the first feature being directly under and obliquely below the second feature, or simply means that the first feature is less level than the second feature.
The free-form surface in this embodiment refers to a surface that can be expressed by a B-spline, a non-uniform rational B-spline, or a T-spline curve, and is not limited to the expression form of the surface, and may be a continuous surface or a surface represented by a discrete point cloud.
As shown in fig. 1 and 2, a free-form surface machine tool contact type in-situ detection error compensation method comprises the following steps:
S1: obtaining a measured initial compensation value by a precompensation method: setting sampling points on a standard sphere, calibrating and measuring the sampling points by adopting different measuring head postures, comparing a measuring result with the actual size of the standard sphere to obtain a measuring error, and correspondingly obtaining an initial measuring error compensation value from a standard database by the measuring error; inputting the initial measurement error compensation value into a measuring head system to form preliminary compensation;
as shown in fig. 3, S101: setting n sampling points on a standard sphere;
the contact point of the standard ball surface and the extension line of the axis of the measuring head is set as an initial sampling point, the included angle between adjacent sampling points is theta, and the included angles between each sampling point and the axis of the measuring head are respectively Setting the values of m and n according to the calibration precision requirement,M represents the number of standard sphere surface rings.
And when in sampling, the ruby ball at the end part of the measuring head approaches to and touches the sampling point along the normal direction of the sampling point, and the coordinate data measured at the sampling point is recorded. FIG. 3 (a) is a view of a plane perpendicular to the axis of the stylus and the axis of the standard sphere, with a sampling point located closest to the ruby ball of the stylus, B being the angle between the axis of the stylus and the axis of the standard sphere, and B being the angle between the axis of the stylus and the axis of the standard sphereN sampling points are equally distributed on the ring of the surface of the standard sphere, and the included angles between every two sampling points are theta, as shown in fig. 3 (b). The values of m and n are set according to the calibration precision requirement,The larger the values of m and n, the higher the calibration accuracy but the longer the required calibration time. In the example of fig. 3: m=3, n=8, and m and n can be selected to be different values according to the calibration accuracy requirement in practical implementation. The calibration of the stylus using different stylus attitudes is shown in figures 4 (a-c). The purpose of the test is to test the influence of the gesture of the measuring head on the measurement error of the in-situ detection system during measurement.
S102: fitting the sampling points in the step S101 to obtain a spherical center coordinate C (x 0,y0,z0);
s103: calibrating and measuring the standard ball by adopting different positions and postures of the measuring head to obtain actual measurement coordinates of each sampling point; the obtained measured coordinates have higher accuracy because the coordinates of the sphere center are known.
S104: correcting the sphere center position C New:
The deformation of the measuring needle and the pre-stroke error when the measuring head contacts the standard ball in the axis direction are very small, and the measuring error is minimum relative to other triggering directions. Therefore, when the measuring head performs measurement in the axial direction, neglecting the measurement error, the distance from the coordinate value of the initial sampling point to the center of the sphere as shown in fig. 3 (a) is the sum d s of the standard sphere radius R b and the radius R p of the measuring head:
ds=Rb+rp
A new center position C New(xNew0,yNew0,zNew0 can be reversely deduced according to the measured coordinate value obtained at the sampling point), assuming that the measured coordinate point obtained at the initial sampling point is P 1(x1,y1,z1), the corrected center position C New is:
Wherein, Representing the vector from point C to point P 1.
S105: obtaining a measurement error value Err i:
calculating the distance d i between C New and each of the sampling point measurement coordinates { P i(xi,yi,zi) |i=1, 2, …, m×n }, the error value Err i is:
S106: the initial measurement error compensation value of each direction corresponding to the error value Err i is obtained in the standard database, and the initial measurement error compensation value is input into the measuring head system to form initial compensation. The calibration result corresponding to fig. 4 can be obtained through the pre-compensation method treatment, as shown in fig. 5.
S2: acquiring an actual measurement error value and an actual measuring head posture:
Setting sampling points on a curved surface to be measured, and measuring measurement data of the sampling points by using a three-coordinate measuring machine as reference data, wherein the measurement data is measured by using a standard measuring head; measuring actual measurement data of a sampling point by adopting an in-situ measurement method, wherein the measuring head subjected to compensation in the step S1 is adopted for measurement; the difference between the actual measurement data and the reference data is the measurement error value of each sampling point in the X, Y, Z three-coordinate direction;
And (3) planning a measuring path of a curved surface sampling point according to the interference condition and the sampling efficiency of the measuring head and the workpiece, and recording measuring position information, namely five-axis movement information of the machine tool X, Y, Z, B, C and I, J, K direction information of the axis of the measuring head, so as to obtain the posture of the measuring head when measuring each sampling point.
As shown in fig. 6, it is assumed that different measurement errors are also corresponding when different probe postures are adopted at different positions on the curved surface to be measured, and a complex corresponding relationship exists between the two. Under the condition of this assumption, firstly, a certain number of sampling points are planned relatively densely on a free curved surface on a workpiece clamped at a specific position on a machine tool by adopting an isoparametric method, then, measurement path planning is carried out, the planned sampling points are respectively measured by adopting an in-situ detection and three-coordinate measuring machine, input and output data shown in fig. 6 are recorded after measurement, the input data, namely five-axis movement information of the machine tool X, Y, Z, B, C and I, J, K direction information of a measuring head axis are directly read in from the planned measurement path, and the output data are errors delta X, delta Y and delta Z in three coordinate axis directions when the same sampling points are measured between the in-situ detection and the three-coordinate measuring machine.
S3: training a convolutional neural network structure according to the actual measuring head posture and the actual measuring error value in the S2, acquiring a curved surface measuring error prediction model, and predicting measuring errors by adopting the curved surface measuring error prediction model:
The convolutional neural network structure comprises an input layer, a convolutional layer, a ReLU activation function layer, a pooling layer, a full connection layer and an output layer; and (3) taking the gesture of the measuring head in the step (S2) as an input parameter, taking the measurement error value in the step (S2) as an output parameter, training the convolutional neural network structure together, obtaining a curve measurement error prediction model after training, and adopting the curve measurement error prediction model to conduct measurement prediction on the curve to be measured to obtain measurement errors delta X, delta Y and delta Z of the curve to be measured.
The built model is shown in fig. 7, the input layer is 1 image 8*1, which is the input shown in fig. 6, the output layer is 1 image 3*1, which is the output shown in fig. 6. The input layer is followed by a convolution layer, the convolution layer and the ReLU activation function layer are used in groups, three pairs of convolution layers and the ReLU activation function layer are used in total, the number of convolution kernels in the three convolution layers is 128, 256 and 512 respectively, and then the total connection layer is used for outputting regression prediction values. The pooling layer is not included in the structure because the pooling layer has the main effect of reducing the size and complexity of the features, thereby reducing the computational effort, while the feature map itself in the present model is not complex and therefore does not require the use of this layer. In fig. 7, which is only an embodiment, the present invention is not limited to the structure shown in this embodiment, and it is also within the framework of the present invention to select an error prediction model including different layers and being constructed by optimizing the convolutional neural network structure for a specific measurement error compensation problem.
S4: performing measurement error compensation according to the measurement error predicted value obtained in the step S3 to obtain a final measurement coordinate value:
and obtaining the sum of error prediction results DeltaX, deltaY and DeltaZ in the sampling point actual coordinate values X, Y, Z and S3 obtained by in-situ measurement, namely the final measurement coordinate value.
S5: and verifying the measurement accuracy through simulation measurement and actual measurement:
The simulation measurement is specifically as follows:
As shown in fig. 8, a dense point cloud and a sparse point cloud are obtained by adopting a method of equally dividing a curved surface parameter domain and an adaptive method, a sampling result of the equal parameter method is used as a training set, and a sample obtained by the adaptive sampling method is used as a verification set. The number of sampling points obtained by adopting the equal parameter sampling method is 30×30=900, and the number of sampling points obtained by adopting the adaptive sampling method is 57. 900 sampling points obtained by the equal parameter sampling method are used for CNN training, so that error distribution experience is generated. The training process obtains measurement experience for in-situ detection in the machine tool working space. The 57 sampling points obtained by the feature-based sampling method are used as verification sets for testing the prediction effect of the error compensation method on the measurement error. The experimental results are shown in fig. 9 (a-b), and the error compensation method provided by the invention can effectively predict the measurement errors in three coordinate directions X, Y, Z. As can be easily observed from fig. 9 (b), the value of the measurement accuracy improvement after compensation is positive, that is, the measurement accuracy is improved at all sampling points after compensation, the measurement accuracy is obtained through calculation, the average value of the measurement error after compensation is reduced by 0.0316mm relative to the measurement error before compensation, the average value of the original measurement error is 0.0337mm, the average error after compensation is reduced by 93.8% relative to the original measurement error, and the measurement accuracy improvement effect is obvious.
The actual measurement is specifically as follows:
Fig. 10 is a training set and a verification set generated by a similar method to the simulation experiment, the three-coordinate measuring machine and the in-situ detection shown in fig. 11 are adopted to respectively measure the actual curved surface and obtain the measurement result, the curved surface measured by the in-situ detection and the curved surface measured by the three-coordinate measurement are compared to obtain the in-situ detection error, the obtained in-situ detection error distribution is shown in fig. 12, the maximum error before the in-situ detection system compensation is calculated to be 0.0615mm, and the average error is 0.0226mm. The measurement error predicted and compensated by the trained convolutional neural network is shown in fig. 13, and the calculated maximum error of the in-place detection is 0.0277mm, and the average error is 0.0075mm. Compared with the error distribution shown in fig. 12, the measurement accuracy of the in-situ detection after compensation is improved by 55.0% in terms of maximum error, and the average error is improved by 64.8%.
It should be noted that in simulation and actual verification, the sampling method generates the training set by adopting an isoparametric method, and the adaptive method generates the verification set, which does not represent the requirement of the invention on sampling, in fact, the data of the generated training set is distributed densely enough on the curved surface, the degree of the density depends on the precision requirement of the application occasion, and the adaptive method can also adopt different methods.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A free-form surface machine tool contact type in-situ detection error compensation method is characterized in that: the method comprises the following steps:
S1: obtaining a measured initial compensation value by a precompensation method: setting sampling points on a standard sphere, calibrating and measuring the sampling points by adopting different measuring head postures, comparing a measuring result with the actual size of the standard sphere to obtain a measuring error, and correspondingly obtaining an initial measuring error compensation value from a standard database by the measuring error; inputting the initial measurement error compensation value into a measuring head system to form preliminary compensation;
S2: acquiring an actual measurement error value and an actual measuring head posture;
S3: training a convolutional neural network structure according to the actual measuring head posture and the actual measurement error value in the S2, acquiring a curved surface measurement error prediction model, and predicting the measurement error of the curved surface to be measured by adopting the curved surface measurement error prediction model;
s4: performing measurement error compensation according to the measurement error predicted value obtained in the step S3 to obtain a final measurement coordinate value;
s5: and verifying the measurement accuracy through simulation measurement and actual measurement.
2. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 1, wherein the method comprises the following steps: the S1 comprises the following steps:
S101: setting n sampling points on a standard sphere;
s102: fitting the sampling points in the step S101 to obtain a spherical center coordinate C (x 0,y0,z0);
s103: calibrating and measuring the standard ball by adopting different positions and postures of the measuring head to obtain actual measurement coordinates of each sampling point;
s104: correcting the sphere center position C New;
s105: obtaining a measurement error value Err i;
s106: the initial measurement error compensation value of each direction corresponding to the error value Err i is obtained in the standard database, and the initial measurement error compensation value is input into the measuring head system to form initial compensation.
3. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 2, wherein the method comprises the following steps: the step S101 specifically includes:
the contact point of the standard ball surface and the extension line of the axis of the measuring head is set as an initial sampling point, the included angle between adjacent sampling points is theta, and the included angles between each sampling point and the axis of the measuring head are respectively Setting the values of m and n according to the calibration precision requirement,M represents the number of standard sphere surface rings.
4. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 2, wherein the method comprises the following steps: the step S104 specifically includes:
the distance from the coordinate value of the initial sampling point to the sphere center is the sum d s of the standard sphere radius R b and the radius R p of the measuring head:
ds=Rb+rp
Assuming that the measured coordinate point obtained at the initial sampling point is P 1(x1,y1,z1), the corrected center position C New is:
Wherein, Representing the vector from point C to point P 1.
5. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 2, wherein the method comprises the following steps: the step S105 specifically includes:
calculating the distance d i between C New and each of the sampling point measurement coordinates { P i(xi,yi,zi) |i=1, 2, …, m×n }, the error value Err i is:
6. the free-form surface machine tool contact type in-situ detection error compensation method according to claim 1, wherein the method comprises the following steps: the step S2 is specifically as follows:
Setting sampling points on a curved surface to be measured, and measuring measurement data of the sampling points by using a three-coordinate measuring machine as reference data, wherein the measurement data is measured by using a standard measuring head; measuring actual measurement data of a sampling point by adopting an in-situ measurement method, wherein the measuring head subjected to compensation in the step S1 is adopted for measurement; the difference between the actual measurement data and the reference data is the measurement error value of each sampling point in the X, Y, Z three-coordinate direction;
And (3) planning a measuring path of a curved surface sampling point according to the interference condition and the sampling efficiency of the measuring head and the workpiece, and recording measuring position information, namely five-axis movement information of the machine tool X, Y, Z, B, C and I, J, K direction information of the axis of the measuring head, so as to obtain the posture of the measuring head when measuring each sampling point.
7. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 1, wherein the method comprises the following steps: the step S3 is specifically as follows:
The convolutional neural network structure comprises an input layer, a convolutional layer, a ReLU activation function layer, a pooling layer, a full connection layer and an output layer; the method comprises the steps of taking the gesture of a measuring head as an input parameter, taking a measurement error value as an output parameter, training a convolutional neural network structure together, obtaining a curve measurement error prediction model after training, and measuring and predicting a curve to be measured by adopting the curve measurement error prediction model to obtain measurement errors delta X, delta Y and delta Z of the curve to be measured.
8. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 1, wherein the method comprises the following steps: the step S4 specifically comprises the following steps:
and obtaining the sum of error prediction results DeltaX, deltaY and DeltaZ in the sampling point actual coordinate values X, Y, Z and S3 obtained by in-situ measurement, namely the final measurement coordinate value.
9. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 1, wherein the method comprises the following steps: the simulation measurement of the S5 is specifically as follows:
The method comprises the steps of respectively obtaining dense point cloud and sparse point cloud by adopting an equal-divided curved surface parameter domain method and an adaptive method, taking a result of sampling by adopting an equal-parameter method as a training set, taking a sample obtained by adopting the adaptive sampling method as a verification set, obtaining sampling points by adopting the equal-parameter sampling method, obtaining the sampling points by adopting the adaptive sampling method, and using the sampling points obtained by adopting the equal-parameter sampling method for convolutional neural network structure training so as to generate error distribution experience, wherein the training process obtains measurement experience of in-place detection in a working space of a machine tool, and the sampling points obtained by adopting the characteristic-based sampling method are used as the verification set for testing the prediction effect of the error compensation method on measurement errors.
10. The free-form surface machine tool contact type in-situ detection error compensation method according to claim 1, wherein the method comprises the following steps: the actual measurement of S5 is specifically:
Measuring the actual curved surface by adopting a three-coordinate measuring machine and in-situ measurement respectively to obtain measurement results, and comparing the curved surface measured in-situ measurement with the curved surface measured by the three-coordinate measuring machine to obtain in-situ detection errors, thereby obtaining maximum errors and average errors before the in-situ measurement system compensates; and calculating the maximum error and average error of the in-situ measurement according to the measurement error after the prediction and compensation of the trained convolutional neural network structure.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410569195.5A CN118288107A (en) | 2024-05-09 | 2024-05-09 | Free-form surface machine tool contact type in-situ detection error compensation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410569195.5A CN118288107A (en) | 2024-05-09 | 2024-05-09 | Free-form surface machine tool contact type in-situ detection error compensation method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN118288107A true CN118288107A (en) | 2024-07-05 |
Family
ID=91678139
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410569195.5A Pending CN118288107A (en) | 2024-05-09 | 2024-05-09 | Free-form surface machine tool contact type in-situ detection error compensation method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118288107A (en) |
-
2024
- 2024-05-09 CN CN202410569195.5A patent/CN118288107A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5244786B2 (en) | Differential calibration | |
US7079969B2 (en) | Dynamic artefact comparison | |
CN109493389B (en) | Camera calibration method and system based on deep learning | |
JP2011528829A (en) | Prototype tool manufacturing method and apparatus | |
CN107315391A (en) | A kind of pretravel error compensating method of Digit Control Machine Tool on-line checking | |
CN109579733B (en) | Method for rapidly measuring and calculating laser 3D printing forming dimensional precision | |
CN112926152B (en) | Digital twin-driven thin-wall part clamping force precise control and optimization method | |
CN107063060A (en) | A kind of method and device for determining surface planarity | |
JP2000046543A (en) | Three-dimensional profile measuring equipment | |
TW202128378A (en) | Calibrating method and calibrating system | |
Ostrowska et al. | Comparison of accuracy of virtual articulated arm coordinate measuring machine based on different metrological models | |
JP5183884B2 (en) | Correction method and measuring apparatus | |
CN109202539B (en) | Online detection method for composite material weak-rigidity special-shaped structure | |
Moona et al. | Measurement uncertainty assessment of articulated arm coordinate measuring machine for length measurement errors using Monte Carlo simulation | |
Cappetti et al. | Fuzzy approach to measures correction on Coordinate Measuring Machines: The case of hole-diameter verification | |
JP2012145551A (en) | R-shape measuring device, r-shape measuring method, and r-shape measuring program | |
JPH10301609A (en) | Position error detection method for robot and device therefor | |
Acero et al. | Evaluation of a metrology platform for an articulated arm coordinate measuring machine verification under the ASME B89. 4.22-2004 and VDI 2617_9-2009 standards | |
Van Gestel et al. | Determining measurement uncertainties of feature measurements on CMMs | |
CN118288107A (en) | Free-form surface machine tool contact type in-situ detection error compensation method | |
CN115358435A (en) | Five-axis adaptive scanning measurement path planning method for model unknown workpiece | |
CN113362456A (en) | Measuring instrument construction method and system based on digital twinning technology | |
KR101525677B1 (en) | Virtual error estimation simulating method for circular test | |
ElMaraghy et al. | Evaluation of actual geometric tolerances using coordinate measuring machine data | |
US20170067729A1 (en) | Method and apparatus for characerising instrument error |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination |