Multi-dimensional error measurement device and method based on porous diffraction interference

Abstract

The invention discloses a multi-dimensional error measurement device and method based on porous diffraction interference, and relates to the field of equipment measurement, and a measurement assembly in the multi-dimensional error measurement device based on porous diffraction interference comprises a diffraction pore plate; the measuring assembly is fixed on target equipment and is parallel to the axis direction of the target equipment; the target equipment drives the measuring assembly to move when moving at multiple degrees of freedom in the space; the laser source and the measuring assembly are in the same axis direction. A plurality of diffraction pores distributed in the horizontal and vertical directions of a plane orthogonal to the optical axis are formed in the diffraction pore plate; laser emitted by the laser source passes through a plurality of diffraction pores in the diffraction pore plate to form a porous diffraction interference image; the calculation module obtains the multi-degree-of-freedom pose change according to the pose recognition model and the acquired porous diffraction interference image in the multi-degree-of-freedom movement process, and compensates the multi-degree-of-freedom pose change to obtain the multi-dimensional error of the target equipment, and the measurement of the multi-dimensional error of the equipment can be realized with high precision and low cost.

Description

Technical Field

[0001] The present application relates to the field of equipment measurement, and in particular to a multi-dimensional error measurement device and method based on multi-aperture diffraction interferometry. Background Art

[0002] With the development of modern manufacturing, high-end CNC machine tools, high-precision coordinate measuring machines and other precision equipment with multi-axis motion mechanisms are becoming increasingly important in the precision manufacturing industry. Spatial geometric position accuracy is one of the key indicators for evaluating such precision equipment.

[0003] At present, for high-end CNC machine tools, high-precision three-dimensional coordinate measuring machines and other precision equipment, the machining errors of the motion axis structure, friction effects and other factors will cause the axis to jump left and right and up and down when moving, resulting in multi-dimensional deviations between the target posture and actual posture of key components. These spatial geometric position errors will lead to serious characteristic posture errors or motion profile deviations, which will cause multi-dimensional errors in the size and shape of the machine tool workpiece or the measurement results of the measuring machine. Therefore, high-precision measurement and compensation of multi-dimensional errors of motion units are of great scientific significance and have very broad application prospects in the field of precision instruments and equipment.

[0004] Among the error measurement methods for precision equipment, laser trackers have the advantages of simple operation, fast measurement speed, and large measurement range. However, most existing trackers can only measure single target balls, which inevitably amplifies the influence of human errors and repeatability errors when measuring spatial errors. Although laser interferometers can measure errors in each degree of freedom with high precision, laser interferometers can only measure one direction error at a time and cannot measure roll (rotation angle around the linear platform axis) errors. Renishaw's multi-beam interferometer has significant advantages in six-degree-of-freedom error measurement and can measure multiple degrees of freedom errors at a time, but it is expensive and sensitive to light source stability and environmental interference. The software system is not open and is only compatible with some control systems.

[0005] Therefore, how to achieve multi-dimensional error measurement of equipment with high precision and low cost has become a problem that needs to be solved urgently. Summary of the invention

[0006] The purpose of this application is to provide a multi-dimensional error measurement device and method based on multi-aperture diffraction interferometry, which can measure the multi-dimensional error (i.e., multi-degree-of-freedom posture change) of equipment with high precision and low cost.

[0007] To achieve the above objectives, this application provides the following solutions:

[0008] In a first aspect, the present application provides a multi-dimensional error measurement device based on multi-aperture diffraction interferometry, comprising: a laser source, a measurement component and a calculation module; the measurement component comprises: a diffraction aperture plate;

[0009] The measuring component is fixed on the target equipment; the axial direction of the measuring component is parallel to the axial direction of the target equipment; the target equipment is used to drive the measuring component to move with multiple degrees of freedom when moving with multiple degrees of freedom in space; the laser source and the measuring component are in the same axial direction; a plurality of diffraction apertures are arranged on the diffraction aperture plate; the plurality of diffraction apertures are distributed in the horizontal direction and the vertical direction of the plane orthogonal to the optical axis;

[0010] The laser source is used to emit laser light; the laser light is used to form a multi-aperture diffraction interference image after passing through a plurality of diffraction apertures on the diffraction aperture plate;

[0011] The calculation module is used to obtain the multi-degree-of-freedom diffraction interference image of the target equipment during the multi-degree-of-freedom movement in space, obtain the multi-degree-of-freedom posture change of the target equipment according to the posture recognition model and the collected multi-degree-of-freedom diffraction interference image, and compensate for the multi-degree-of-freedom posture change, and determine the compensated multi-degree-of-freedom posture change as the multi-dimensional error of the target equipment; the posture recognition model is constructed based on the machine learning model.

[0012] Optionally, in terms of obtaining the multi-degree-of-freedom posture change of the target equipment according to the posture recognition model and the collected multi-aperture diffraction interference image, the calculation module is specifically used to:

[0013] A training data set is used to train a machine learning model, and the trained machine learning model is determined as a posture recognition model; the training data set includes: multi-aperture diffraction interference images at different times when the equipment used for training moves in multiple degrees of freedom in space and corresponding real multi-degree-of-freedom posture changes;

[0014] The collected multi-aperture diffraction interference image of the target equipment at the current moment is input into the posture recognition model to obtain the multi-degree-of-freedom posture change of the target equipment at the current moment.

[0015] Optionally, the diffraction apertures on the diffraction aperture plate are diffraction holes or diffraction slits of different shapes and sizes.

[0016] Optionally, in inputting the collected multi-aperture diffraction interference image of the target equipment at the current moment into the posture recognition model to obtain the multi-degree-of-freedom posture change of the target equipment at the current moment, the calculation module is specifically used to:

[0017] The collected porous diffraction interference image of the target equipment at the current moment is input into the posture recognition model, and the posture recognition model extracts the key features of the porous diffraction interference image at the current moment, and recognizes the multi-degree-of-freedom posture change at the current moment according to the key features at the current moment; the key features include: one or more of the stripe distribution state, the number of stripes, the stripe form, the stripe spacing, the phase contained in the stripes, the gradient of various stripe changes and the distribution information of various stripe change gradients.

[0018] Optionally, the multi-dimensional error measurement device based on multi-aperture diffraction interferometry further includes: an adjustment device; the measurement component is fixed to the end of the target equipment through the adjustment device; the adjustment device is used to adjust the relative multi-dimensional position of the axis of the laser source and the axis of the measurement component.

[0019] Optionally, the measuring component further includes: an image acquisition device; the image acquisition device is connected to the computing module; the image acquisition device is used to acquire the porous diffraction interference image and send the acquired porous diffraction interference image to the computing module.

[0020] Optionally, the measuring component further includes: an optical lens group; the laser irradiates the diffraction aperture plate after passing through the optical lens group; the axial distance between the optical lens group, the diffraction aperture plate and the image acquisition device is at an optimal imaging position; the optimal imaging position is determined based on the clarity of the multi-aperture diffraction interference image and the sensitivity of the fringe edges of the multi-aperture diffraction interference image to the change of the diffraction aperture plate position.

[0021] Optionally, the target equipment moves in six degrees of freedom in space; the six degrees of freedom include: linearity, horizontal straightness, vertical straightness, pitch, yaw and roll.

[0022] Optionally, the image acquisition device is an image sensor.

[0023] In a second aspect, the present application provides a multi-dimensional error measurement method based on multi-aperture diffraction interferometry, which is used in the above-mentioned multi-dimensional error measurement device based on multi-aperture diffraction interferometry; the multi-dimensional error measurement method based on multi-aperture diffraction interferometry comprises:

[0024] Acquire multi-aperture diffraction interference images of target equipment during multi-degree-of-freedom movement in space;

[0025] Obtaining multi-degree-of-freedom posture changes of the target equipment according to the posture recognition model and the collected multi-aperture diffraction interference image;

[0026] The multi-degree-of-freedom posture change is compensated, and the compensated multi-degree-of-freedom posture change is determined as a multi-dimensional error of the target equipment; the posture recognition model is constructed based on a machine learning model.

[0027] According to the specific embodiments provided in this application, this application has the following technical effects:

[0028] The present application provides a multi-dimensional error measurement device and method based on multi-aperture diffraction interference, which includes a measuring component including a diffraction aperture plate. A plurality of diffraction apertures distributed in horizontal and vertical directions on a plane orthogonal to the optical axis are arranged on the diffraction aperture plate, so that a complex diffraction interference image formed by the diffraction aperture plate includes the postures of the diffraction aperture plate in multiple degrees of freedom in space. When the target equipment drives the diffraction aperture plate to move, the multi-degree-of-freedom posture changes are obtained by identifying the multi-aperture diffraction interference image, that is, the multi-dimensional error is obtained. In this way, multi-dimensional error measurement is performed based on the diffraction interference principle, which can reduce costs. The calculation module identifies the multi-degree-of-freedom posture changes according to a posture recognition model constructed based on a machine learning model, which can improve the measurement accuracy. Therefore, the present application realizes the measurement of multi-dimensional errors of equipment with high precision and low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 A schematic structural diagram of a multi-dimensional error measurement device based on multi-aperture diffraction interferometry provided in one embodiment of the present application;

[0031] Figure 2 A schematic diagram of the structure of a measurement component provided in one embodiment of the present application;

[0032] Figure 3 A flowchart of a multi-dimensional error measurement method based on multi-aperture diffraction interferometry provided in one embodiment of the present application.

[0033] Figure numerals: laser source - 1, measuring component - 2, adjustment device - 3, calculation module - 4, optical lens group - 201, diffraction aperture plate - 202, image acquisition device - 203, measuring tool - 204. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0035] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0036] This application provides a low-cost measurement solution with a novel principle, which uses the sensitivity of the diffraction result after the light beam passes through the small hole to the position of the small hole, and further enhances the sensitivity and correlation by adding the multi-hole diffraction interference information after the multi-hole distribution, establishes a mapping model between the diffraction interference image and the multi-degree-of-freedom posture, and solves the multi-degree-of-freedom posture change by obtaining the diffraction interference image. This application is novel in principle, low in cost, has low requirements for light source stability and environmental conditions, and has high measurement accuracy.

[0037] In an exemplary embodiment, Figure 1 As shown, a multi-dimensional error measurement device based on multi-aperture diffraction interferometry is provided, comprising: a laser source 1, a measurement component 2 and a calculation module 4, see Figure 2 , the measuring component 2 includes: a diffraction aperture plate 202. Figure 1 In the figure, Δx, Δy, and Δz are translational motions in three directions; Δα, Δβ, and Δγ are rotational yaws in three directions.

[0038] The measuring component 2 is fixed on the target equipment; the axial direction of the measuring component 2 is parallel to the axial direction of the target equipment; the target equipment is used to drive the measuring component 2 to move with multiple degrees of freedom when moving with multiple degrees of freedom in space; the laser source 1 and the measuring component 2 are in the same axial direction; a plurality of diffraction apertures are arranged on the diffraction aperture plate 202; the plurality of diffraction apertures are distributed in the horizontal direction and the vertical direction of the plane orthogonal to the optical axis. The target equipment can be a linear platform to be measured.

[0039] The laser source 1 is used to emit laser light; the laser light is used to form a multi-aperture diffraction interference image after passing through a plurality of diffraction apertures on the diffraction aperture plate 202 .

[0040] The calculation module 4 is used to obtain the multi-degree-of-freedom diffraction interference image of the target equipment during the multi-degree-of-freedom movement in space, obtain the multi-degree-of-freedom posture change of the target equipment according to the posture recognition model and the collected multi-degree-of-freedom diffraction interference image, and compensate for the multi-degree-of-freedom posture change, and determine the compensated multi-degree-of-freedom posture change as the multi-dimensional error of the target equipment; the posture recognition model is constructed based on the machine learning model.

[0041] In another exemplary embodiment of the present application, in terms of obtaining the multi-degree-of-freedom posture change of the target equipment according to the posture recognition model and the collected multi-aperture diffraction interference image, the calculation module 4 is specifically used to:

[0042] The machine learning model is trained using a training data set, and the trained machine learning model is determined as a posture recognition model; the training data set includes: a multi-aperture diffraction interference image and a corresponding real multi-degree-of-freedom posture change at different times when the equipment used for training moves in space with multiple degrees of freedom. Among them, the real multi-degree-of-freedom posture change can be obtained in the following two ways: one is to perform multi-dimensional error measurement on some linear platforms through the currently leading high-precision multi-dimensional error measurement device, and then use the proposed method to perform corresponding multi-aperture diffraction interference image measurement and collection, so that the multi-aperture diffraction interference image corresponds to the measured multi-degree-of-freedom posture change; the second is to install a high-precision five-axis electric translation stage on the high-precision linear translation stage, install the measurement component 2 on this five-axis translation stage, and record the diffraction interference image while the five-axis translation stage and the linear platform move, and take the movement of the five-axis translation stage and the linear platform as the reference, repeat multiple groups, and obtain the real multi-degree-of-freedom posture change.

[0043] The collected multi-aperture diffraction interference image of the target equipment at the current moment is input into the posture recognition model to obtain the multi-degree-of-freedom posture change of the target equipment at the current moment.

[0044] In another exemplary embodiment of the present application, the diffraction apertures on the diffraction aperture plate 202 are diffraction holes or diffraction slits of different shapes and sizes. The shape and size of the diffraction hole or the diffraction slit are determined according to actual needs and are not limited here. For example, the shape of the diffraction hole can be a circular hole, a square hole, a triangular hole, or an oblong hole.

[0045] In another exemplary embodiment of the present application, in terms of compensating the multi-degree-of-freedom posture change and determining the compensated multi-degree-of-freedom posture change as the multi-dimensional error of the target equipment, the calculation module 4 is specifically used to:

[0046] Using the formula Y = (1-k)y pre-b Compensate the multi-degree-of-freedom posture change, and determine the compensated multi-degree-of-freedom posture change as the multi-dimensional error of the target equipment; wherein, y pre represents the multi-degree-of-freedom posture change of the target equipment; Y represents the multi-degree-of-freedom posture change after compensation; k and b represent fitting parameters; k and b are obtained by fitting the predicted multi-degree-of-freedom posture change and the actual multi-degree-of-freedom posture change using a fitting method (for example, the least squares method); the predicted multi-degree-of-freedom posture change is determined based on the multi-degree-of-freedom posture output by the posture recognition model.

[0047] In another exemplary embodiment of the present application, in inputting the collected multi-aperture diffraction interference image of the target equipment at the current moment into the posture recognition model to obtain the multi-degree-of-freedom posture change of the target equipment at the current moment, the calculation module 4 is specifically used to:

[0048] The collected porous diffraction interference image of the target equipment at the current moment is input into the posture recognition model, and the posture recognition model extracts the key features of the porous diffraction interference image at the current moment, and recognizes the multi-degree-of-freedom posture change at the current moment according to the key features at the current moment; the key features include but are not limited to: one or more of the stripe distribution state, the number of stripes, the stripe form, the stripe spacing, the phase contained in the stripes, the gradient of various stripe changes and the distribution information of various stripe change gradients.

[0049] In another exemplary embodiment of the present application, please refer to Figure 1 The multi-dimensional error measurement device based on multi-aperture diffraction interferometry also includes: an adjustment device 3; the measurement component 2 is fixed to the end of the target equipment through the adjustment device 3; the adjustment device 3 is used to adjust the relative multi-dimensional position of the axis of the laser source 1 and the axis of the measurement component 2.

[0050] The adjustment device 3 may be a five-axis adjustment mechanism having two translations and three rotations.

[0051] In another exemplary embodiment of the present application, the computing module 4 is mainly used for data processing and deployment of pre-trained machine learning models.

[0052] In another exemplary embodiment of the present application, please refer to Figure 2 The measuring component 2 also includes: an image acquisition device 203; the image acquisition device 203 is connected to the calculation module 4; the image acquisition device 203 is used to acquire the porous diffraction interference image and send the acquired porous diffraction interference image to the calculation module 4.

[0053] The image acquisition device 203 may be an image sensor.

[0054] In another exemplary embodiment of the present application, the computing module 4 may be a general-purpose computer or a dedicated data analysis device.

[0055] In another exemplary embodiment of the present application, please refer to Figure 2 The measuring assembly 2 further includes: an optical lens group 201; the laser beam passes through the optical lens group 201 and is irradiated onto the diffraction aperture plate 202. The axial distance between the optical lens group 201, the diffraction aperture plate 202 and the image acquisition device 203 is at an optimal imaging position; the optimal imaging position is determined according to the clarity of the multi-aperture diffraction interference image and the sensitivity of the fringe edge of the multi-aperture diffraction interference image to the position change of the diffraction aperture plate 202.

[0056] In practical applications, the laser is transformed by the optical lens group 201 (for example, the optical path is adjusted) and then irradiated onto the diffraction aperture plate 202 .

[0057] In another exemplary embodiment of the present application, please refer to Figure 2 The measuring assembly 2 further includes a measuring fixture 204, which is used to install the optical lens group 201, the diffraction aperture plate 202 and the image acquisition device 203. The relative positions of the optical lens group 201, the diffraction aperture plate 202 and the image acquisition device 203 can be adjusted according to the measurement requirements.

[0058] In another exemplary embodiment of the present application, the shape, number, size, thickness and distribution of the diffraction apertures are determined according to measurement requirements to determine different types of diffraction aperture plates 202 for replacement; the position information of the diffraction aperture plate 202 in multiple degrees of freedom in space is obtained by identifying the complex multi-hole diffraction interference image formed by the diffraction apertures.

[0059] In another exemplary embodiment of the present application, the target equipment moves in space with six degrees of freedom; the six degrees of freedom include: linearity, horizontal straightness, vertical straightness, pitch, yaw and roll.

[0060] An overall posture relationship of the multi-dimensional error measurement device based on multi-aperture diffraction interferometry of this embodiment can be described as follows: the measuring component 2 is installed on the adjusting device 3, and the spatial posture of the measuring component 2 is adjusted by the adjusting device 3 to achieve calibration; the computing module 4 is connected to the measuring end component through a cable; the optical lens group 201, the diffraction aperture plate 202, and the image acquisition device 203 are installed on the measuring fixture 204 in sequence to ensure that the light emitted by the laser source 1 passes through in sequence and is finally imaged by the image acquisition device 203; the optical lens group 201, the diffraction aperture plate 202, and the image acquisition device 203 are installed on the measuring fixture 204 to determine their relative postures, and after calibration before use, the posture is no longer adjusted and the measuring component 2 is packaged as a whole.

[0061] Taking six degrees of freedom as an example, the implementation process of the multi-dimensional error measurement device based on multi-aperture diffraction interferometry in practical applications is further described in detail below.

[0062] Six-dimensional error means that in the motion of the linear module, the slider does not simply translate along the track direction, but it will undergo posture jumps in six degrees of freedom (linearity, horizontal straightness, vertical straightness, pitch, yaw, and roll) in space, that is, six-degree-of-freedom posture changes.

[0063] First, a multi-dimensional error measurement device based on multi-aperture diffraction interferometry is installed. The measuring component 2 consists of an optical lens group 201, a diffraction aperture plate 202, an image acquisition device 203 and a measuring end fixture 204. The measuring component 2 is installed on an adjusting device 3, and the spatial posture of the measuring component 2 is adjusted by the adjusting device 3 to achieve calibration. The computing module 4 is connected to the measuring component 2 through a cable. The optical lens group 201, the diffraction aperture plate 202 and the image acquisition device 203 are sequentially installed on the measuring end fixture 204 to ensure that the light emitted by the laser source 1 passes through in sequence and is finally imaged in the image acquisition device 203. After installation, the relative postures are determined, and no adjustment is performed after calibration before use, and the measuring component 2 is packaged as a whole.

[0064] The optical lens group 201 is a combination of optical elements that adjusts the axial range (i.e., the range along the axial motion direction of the linear platform) and the radial range (i.e., the range along the radial motion direction of the linear platform). Corresponding adjustments can be made for measurements of different axial and radial ranges (i.e., the elements in the optical lens group 201 are replaced or combined). Since the axial range along the motion of the linear platform is large, the laser emitted by the laser source 1 adjusts the optical path through the optical lens group 201 at the front end, and the axial and radial resolution and sensitivity of the image can also be improved by adjusting some optical elements.

[0065] A plurality of diffraction apertures are arranged on the diffraction aperture plate 202 in the horizontal and vertical directions along a plane orthogonal to the optical axis. The shape, quantity, size, thickness and distribution of the diffraction apertures are determined according to measurement requirements, and different types of diffraction aperture plates 202 are replaced. The position information of the diffraction aperture plate 202 in six degrees of freedom in space is obtained by identifying the complex diffraction interference image formed by the diffraction apertures.

[0066] The image acquisition device 203 uses an image sensor to receive the multi-aperture diffraction interference image, and other types of image sensors may also be used instead.

[0067] The relative positions of the optical lens assembly 201 , the diffraction aperture plate 202 , and the image acquisition device 203 can be adjusted according to measurement requirements.

[0068] Based on the above multi-dimensional error measurement device based on multi-aperture diffraction interferometry, the process of realizing six-dimensional error measurement is as follows:

[0069] Step 1: Build the measurement system device.

[0070] Fix the measuring component 2 to the adjusting device 3, fix the adjusting component 3 to the end of the linear platform to be measured, and adjust the spatial posture of the measuring component 2 by adjusting the adjusting device 3 so that the optical axis direction of the measuring component 2 is parallel to the axial direction of the linear platform.

[0071] Step 2: Calibrate the measuring device.

[0072] Arrange the laser source 1, and arrange the laser source 1 in the same axial direction as the measuring component 2, so that the axis of the laser source 1 is parallel to the optical path direction of the measuring system; then adjust the relative position of the axis of the laser source 1 and the axis of the measuring component 2 by adjusting the adjustment device 3 to complete the calibration of the measuring system.

[0073] Step 3: Six-dimensional error measurement.

[0074] When the linear platform moves, the image acquisition device 203 collects the porous diffraction interference image and transmits it to the calculation module 4. After normalization and preprocessing, it is input into the pre-trained machine learning model (posture recognition model) for recognition. By identifying the key features of the porous diffraction interference image, the six-dimensional error generated by the movement of the adjustment device 3 and the measuring component 2 driven by the end of the linear axis is obtained.

[0075] In this step, in order to improve the generalization and robustness of the machine learning model and reduce the instability caused by the difference in order of magnitude, the sample value of each feature is normalized to [0, 1] using the Min-Max method. The normalization equation is as follows:

[0076]

[0077] Among them, x n is the normalized eigenvalue, x is the original sample eigenvalue, x min is the minimum eigenvalue of the sample data, x max is the maximum eigenvalue of the sample data.

[0078] The correspondence between the image and the pose in this step can be explained by the Fresnel-Kirchhoff diffraction formula:

[0079]

[0080] The spherical wave emitted by a monochromatic point light source irradiates the aperture ∑, and the complex amplitude of light vibration is generated at any point P after ∑. In the formula, l is the distance from the point light source S to any point Q on ∑, r is the distance from point Q to point P, (n, l) and (n, r) are the angles between the normal n of the aperture surface ∑ and the directions of l and r, respectively. is the diffraction result, i is the imaginary unit, k is the wave number, A is the amplitude at a unit distance from the light source, λ is the wavelength, and dσ is the integral element of the surface integral. Therefore, changes in the position and posture of the diffraction pinhole will affect the distribution of the complex amplitude of the light source on the diffraction pinhole, the angle between the normal of the aperture surface ∑ and the l and r directions, and ultimately cause the diffraction image to change. There is obviously a mapping relationship between the diffraction result and the pinhole position and posture, but it is obviously very difficult to reversely calculate the change in the pinhole position and posture through the diffraction result. In addition, it is difficult to obtain accurate diffraction results in actual experiments, which further increases the difficulty of analytical calculations.

[0081] Furthermore, this embodiment increases the distribution of holes, adds interference information based on the diffraction image, enhances the mapping relationship between the diffraction interference image and the six-dimensional error, and improves the measurement performance.

[0082] The machine learning model uses a large amount of data collected previously as a training data set to train the model and obtain model parameters that can accurately derive the six-dimensional posture according to the diffraction interference image. The training data set includes: the porous diffraction interference images and the corresponding real six-degree-of-freedom posture changes at different times when the equipment used for training moves in space with six degrees of freedom. Among them, the acquisition method of the real six-degree-of-freedom posture change can be adopted in the following two ways: one is to use the high-precision multi-dimensional error measurement device that currently occupies a leading advantage to measure the six-dimensional error of some linear platforms, and then use the proposed method to measure and collect the corresponding porous diffraction interference images, so that the porous diffraction interference images correspond to the measured six-degree-of-freedom posture changes; the second is to install a high-precision five-axis electric displacement stage on the high-precision linear displacement stage, install the measurement component 2 on this five-axis displacement stage, and record the diffraction interference images while the five-axis displacement stage and the linear platform move, and repeat multiple groups based on the movement amount of the five-axis displacement stage and the linear platform. The five-axis precision displacement stage here is different from the aforementioned adjustment device 3.

[0083] During training, the training data set can be composed of a large number of spatial six-degree-of-freedom posture changes and multi-aperture diffraction interference images under corresponding postures. The multi-aperture diffraction interference images are model inputs and the six-degree-of-freedom posture changes are outputs. The multi-aperture diffraction interference images are input into the machine learning model (such as a neural network) in the form of a digital matrix. During the entire model training process, the machine learning model adjusts the parameters in each layer through optimization methods such as backpropagation algorithms and gradient descent to minimize the prediction error. As the training progresses, the model gradually learns to extract key features that are useful for the task from the original image data and make accurate predictions based on these features. The machine learning model obtains six-dimensional posture changes by extracting key features.

[0084] Step 4: Compensation of measurement results.

[0085] The six-dimensional posture change obtained in step 3 is subjected to system error compensation to obtain the final six-dimensional posture change (six-dimensional error).

[0086] After the model training is completed, the model is tested with a sufficient amount of test data, and the actual values ​​of the six degrees of freedom and the model prediction errors are linearly fitted using the least squares method. The calculation formula is y=kx+b; where x i is the model prediction value of the i-th test data, y i is the model prediction error of the i-th test data (the difference between the actual value and the model prediction value), x and y are the means of the corresponding samples. The final measurement result during use is Y = y pre -y=y pre -(ky pre +b)=(1-k)y pre -b.

[0087] In the above-mentioned embodiment, a plurality of diffraction apertures are arranged on the diffraction aperture plate 202 along a plane orthogonal to the optical axis; the complex optical diffraction interference image formed after the light emitted by the laser source 1 passes through the plurality of diffraction apertures on the diffraction aperture plate 202 includes the position information of the diffraction aperture plate 202 in six degrees of freedom in space; when the diffraction aperture plate 202 moves, the nonlinear mapping model established by the changing relationship between the position information of the diffraction aperture plate 202 and the diffraction interference image can be used as a high-precision recognition model of the six-degree-of-freedom position of the diffraction aperture plate 202, that is, the measurement of the six-degree-of-freedom position change (i.e., six-dimensional error) of the diffraction aperture plate 202 is realized by recognizing the multi-aperture diffraction interference image.

[0088] The advantages of this embodiment are described below based on the current problems.

[0089] High-precision three-dimensional coordinate measuring machines and high-end CNC machine tools are key equipment in the fields of national defense, military industry, and civilian precision manufacturing. The level of spatial geometric position accuracy during their movement directly affects the development of products towards high stability and high precision. There are many reasons for the low manufacturing technology of high-precision three-dimensional coordinate measuring machines and high-end CNC machine tools, but the lack of high-precision measurement technology and equipment is one of the main reasons.

[0090] For high-end CNC machine tools and three-coordinate measuring machines, the six-dimensional errors of the linear motion units such as modules and guide rail sliders will cause spatial geometric position errors in the machine tool end and three-coordinate measuring head, and ultimately errors will occur in the size and shape of the workpiece processed by the machine tool or the measurement results of the measuring machine.

[0091] The multi-dimensional error measurement device based on multi-aperture diffraction interferometry in this embodiment can measure six-dimensional errors, analyze the real error sources of the current mobile platform, and then make targeted adjustments and compensate for its spatial six-dimensional errors to improve the spatial geometric position accuracy of the equipment.

[0092] The multi-dimensional error measurement device based on multi-aperture diffraction interference of this embodiment performs multi-dimensional error measurement through the diffraction interference principle, is insensitive to the original error in the optical path system, and is beneficial to reducing manufacturing costs; a plurality of diffraction apertures are arranged on the diffraction aperture plate 202 in the horizontal and vertical directions along a plane orthogonal to the optical axis, and a complex diffraction interference image formed by these diffraction apertures contains the position information of the diffraction aperture plate 202 in multiple degrees of freedom in space. When the diffraction aperture plate 202 moves, the position change information of the diffraction aperture plate 202 in multiple degrees of freedom in space, i.e., the multi-dimensional error, is obtained by identifying the complex diffraction interference image formed by the diffraction apertures; feature recognition is performed on the complex multi-aperture diffraction interference image through a pre-trained machine learning model, and effective information is extracted from the massive features contained in the diffraction interference image to improve performance indicators such as measurement resolution and measurement accuracy.

[0093] Based on the same inventive concept, the embodiment of the present application also provides a multi-dimensional error measurement method based on multi-aperture diffraction interferometry implemented by the multi-dimensional error measurement device based on multi-aperture diffraction interferometry mentioned above. The implementation scheme for solving the problem provided by the method is similar to the implementation scheme recorded in the above-mentioned device, so the specific limitations of one or more multi-dimensional error measurement method embodiments based on multi-aperture diffraction interferometry provided below can refer to the limitations of the multi-dimensional error measurement device based on multi-aperture diffraction interferometry above, and will not be repeated here.

[0094] In an exemplary embodiment, Figure 3 As shown, a multi-dimensional error measurement method based on multi-aperture diffraction interferometry is provided, including:

[0095] Step 301 : acquiring a multi-aperture diffraction interference image of a target device during multi-degree-of-freedom movement in space.

[0096] Step 302 , obtaining multi-degree-of-freedom posture changes of the target equipment according to the posture recognition model and the collected multi-aperture diffraction interference image.

[0097] Step 303, compensating the multi-degree-of-freedom posture change, and determining the compensated multi-degree-of-freedom posture change as the multi-dimensional error of the target equipment; the posture recognition model is constructed based on a machine learning model.

[0098] This application only needs one setup to accurately measure the posture changes of multiple degrees of freedom along the linear platform at the same time, which is beneficial to measure and compensate for the multi-dimensional geometric posture changes in the terminal space of precision equipment. The measurement method has a novel principle, low cost and high measurement accuracy.

[0099] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A multi-dimensional error measurement device based on multi-aperture diffraction interferometry, characterized in that: The multi-dimensional error measurement device based on multi-aperture diffraction interference includes: a laser source, a measurement component and a calculation module; the measurement component includes: a diffraction aperture plate; The measuring component is fixed on the target equipment; the axial direction of the measuring component is parallel to the axial direction of the target equipment; the target equipment is used to drive the measuring component to move with multiple degrees of freedom when moving with multiple degrees of freedom in space; the laser source and the measuring component are in the same axial direction; a plurality of diffraction apertures are arranged on the diffraction aperture plate; the plurality of diffraction apertures are distributed in the horizontal direction and the vertical direction of the plane orthogonal to the optical axis; The laser source is used to emit laser light; the laser light is used to form a multi-aperture diffraction interference image after passing through a plurality of diffraction apertures on the diffraction aperture plate; The calculation module is used to obtain the multi-degree-of-freedom diffraction interference image of the target equipment during the multi-degree-of-freedom movement in space, obtain the multi-degree-of-freedom posture change of the target equipment according to the posture recognition model and the collected multi-degree-of-freedom diffraction interference image, and compensate for the multi-degree-of-freedom posture change, and determine the compensated multi-degree-of-freedom posture change as the multi-dimensional error of the target equipment; the posture recognition model is constructed based on the machine learning model.

2. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 1, characterized in that: In terms of obtaining the multi-degree-of-freedom posture change of the target equipment according to the posture recognition model and the collected multi-aperture diffraction interference image, the calculation module is specifically used for: A training data set is used to train a machine learning model, and the trained machine learning model is determined as a posture recognition model; the training data set includes: multi-aperture diffraction interference images at different times when the equipment used for training moves in multiple degrees of freedom in space and corresponding real multi-degree-of-freedom posture changes; The collected multi-aperture diffraction interference image of the target equipment at the current moment is input into the posture recognition model to obtain the multi-degree-of-freedom posture change of the target equipment at the current moment.

3. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 1, characterized in that: The diffraction apertures on the diffraction aperture plate are diffraction holes or diffraction slits of different shapes and sizes.

4. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 2, characterized in that: In terms of inputting the collected multi-aperture diffraction interference image of the target equipment at the current moment into the posture recognition model to obtain the multi-degree-of-freedom posture change of the target equipment at the current moment, the calculation module is specifically used for: The collected porous diffraction interference image of the target equipment at the current moment is input into the posture recognition model, and the posture recognition model extracts the key features of the porous diffraction interference image at the current moment, and recognizes the multi-degree-of-freedom posture change at the current moment according to the key features at the current moment; the key features include: one or more of the stripe distribution state, the number of stripes, the stripe form, the stripe spacing, the phase contained in the stripes, the gradient of various stripe changes and the distribution information of various stripe change gradients.

5. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 1, characterized in that: The multi-dimensional error measurement device based on multi-aperture diffraction interferometry also includes: an adjustment device; the measurement component is fixed to the end of the target equipment through the adjustment device; the adjustment device is used to adjust the relative multi-dimensional position of the axis of the laser source and the axis of the measurement component.

6. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 1, characterized in that: The measuring component further includes: an image acquisition device; the image acquisition device is connected to the calculation module; the image acquisition device is used to acquire the porous diffraction interference image and send the acquired porous diffraction interference image to the calculation module.

7. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 6, characterized in that: The measuring component also includes: an optical lens group; the laser is irradiated onto the diffraction aperture plate after passing through the optical lens group; the axial distance between the optical lens group, the diffraction aperture plate and the image acquisition device is at an optimal imaging position; the optimal imaging position is determined based on the clarity of the multi-aperture diffraction interference image and the sensitivity of the fringe edges of the multi-aperture diffraction interference image to the change of the diffraction aperture plate position.

8. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 1, characterized in that: The target equipment moves in space with six degrees of freedom; the six degrees of freedom include: linearity, horizontal straightness, vertical straightness, pitch, yaw and roll.

9. The multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to claim 6, characterized in that: The image acquisition device is an image sensor.

10. A multi-dimensional error measurement method based on multi-aperture diffraction interferometry, characterized in that: The multi-dimensional error measurement method based on multi-aperture diffraction interferometry is used for the multi-dimensional error measurement device based on multi-aperture diffraction interferometry according to any one of claims 1 to 9; The multi-dimensional error measurement method based on multi-aperture diffraction interferometry comprises: Acquire multi-aperture diffraction interference images of target equipment during multi-degree-of-freedom movement in space; Obtaining multi-degree-of-freedom posture changes of the target equipment according to the posture recognition model and the collected multi-aperture diffraction interference image; The multi-degree-of-freedom posture change is compensated, and the compensated multi-degree-of-freedom posture change is determined as a multi-dimensional error of the target equipment; the posture recognition model is constructed based on a machine learning model.

Images