Machining apparatus

Abstract

To provide a technique capable of estimating a part to be preferentially processed with higher accuracy.SOLUTION: Acquiring binary data for classifying whether an arbitrary position of a processing target surface is a contact position or a non-contact position with respect to a reference surface when the processing target surface is rubbed against the reference surface with a workpiece in a predetermined assembled state in a processing apparatus that automatically performs scraping; A processor that acquires measured height data in which height information at an arbitrary position on a machining target surface of a workpiece that is not in a predetermined assembled state is a point group, acquires correction amount data in which a correction amount of the height information at the arbitrary position on the machining target surface is a point group based on the binary data and the measured height data, acquires assumed height data in the predetermined assembled state by superimposing the correction amount data on the measured height data, specifies a convex portion of the machining target surface based on the assumed height data, and generates machining instruction data for cutting the convex portion.SELECTED DRAWING: Figure 9

Description

[Technical Field]

[0001] The present invention relates to a processing device that automatically performs scraping processing. [Background technology]

[0002] Patent Document 1 discloses an automatic scraping device that uses a robot to automatically perform scraping. Scraping is a process that finishes the sliding surface of a workpiece (surface to be machined) to a flat state so that the entire sliding surface comes into uniform contact with the mating surface (reference surface) while leaving extremely minute irregularities on the micron level in order to form a reservoir of lubricating oil on the surface. Patent Document 1 discloses a technology that identifies, based on surface height data and sliding image data, the areas on the surface to be machined that should be scraped with priority, something that was previously done by a craftsman based on feel and experience. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2023-151401 Summary of the Invention [Problem to be solved by the invention]

[0004] For example, if the workpiece is deformed by screw fastening or by contact with the mating surface, the sliding surface of the workpiece may be in a slightly inclined contact state with the mating surface, and the desired sliding properties may not be obtained. Therefore, it may be difficult to accurately estimate the area to be machined with priority based only on the planar height information of the workpiece, which does not take into account the above-mentioned deformation state.

[0005] An object of the present invention is to provide a technique that takes into account the deformation state of a workpiece and enables more accurate estimation of the area to be machined with priority. [Means for solving the problem]

[0006] In order to solve the above problems, the processing device of the present invention comprises: A processing device having a processor that generates processing instruction data, and automatically scraping a surface to be processed of a workpiece based on the processing instruction data, The processor: When the workpiece is in a predetermined assembly state and the processing target surface is rubbed against a reference surface, binary data is acquired which classifies an arbitrary position of the processing target surface as either a contact position or a non-contact position with respect to the reference surface; acquiring measured height data in the form of a point cloud of height information at any position on the surface of the workpiece not in the predetermined assembled state; acquiring correction amount data representing a point cloud of correction amounts for the height information at any position on the processing object surface based on the binary data and the measured height data; by superimposing the correction amount data on the measured height data, estimated height data in the predetermined assembly state is obtained; The method is characterized in that convex portions on the surface to be machined are identified based on the assumed height data, and machining instruction data for cutting the convex portions is generated. [Effects of the Invention]

[0007] According to the present invention, it is possible to estimate with higher accuracy the portion that should be machined with priority, taking into account the deformation state of the workpiece. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a diagram showing a schematic configuration of an automatic scraping device according to a first embodiment. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of a control device. [Figure 3] FIG. 2 is a block diagram illustrating an example of a functional configuration of a control device. [Figure 4] FIG. 10 is a diagram illustrating surface height data of a surface to be processed. [Figure 5] FIG. 10 is a diagram illustrating aligned image data. [Figure 6]10A and 10B are schematic diagrams showing deformations that occur in a workpiece during a predetermined assembly process. [Figure 7] FIG. 4 is an explanatory diagram of a method for correcting surface height data in the first embodiment. [Figure 8] FIG. 10 is a schematic diagram showing the distribution of contact positions / non-contact positions based on surface height data. [Figure 9] FIG. 4 is an explanatory diagram of a method for correcting surface height data in the first embodiment. [Figure 10] FIG. 3 is an image diagram of a correction amount curved surface in the first embodiment. [Figure 11] 10A and 10B are diagrams illustrating surface height data before and after correction. [Figure 12] 10 is an explanatory diagram of identifying a contact point on a measured surface Hm in the second embodiment. FIG. [Figure 13] 10 is an explanatory diagram for adjusting the amount of deformation of the acquired deformed reproduction surface f. FIG. [Figure 14] FIG. 10 is an explanatory diagram for acquiring an assumed surface (correcting surface height data). [Figure 15] 10 is an explanatory diagram showing the results of applying the surface height data correction method according to the second embodiment. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that each configuration and their combinations in the embodiments are merely examples, and additions, omissions, substitutions, and other modifications of the configurations are possible as appropriate within the scope of the gist of the present invention. The present invention is not limited to the embodiments, but is limited only by the claims.

[0010] <Embodiment 1> (Schematic configuration of processing device) Fig. 1 is a diagram showing a schematic configuration of an automatic scraping device 1 according to embodiment 1. As shown in Fig. 1, the automatic scraping device 1 includes a control device 100, a robot arm 200, a three-dimensional shape measuring device 300, a camera 400, and the like.

[0011] The automatic scraping device 1 is a device that automatically performs scraping on a surface 11 to be processed of a workpiece 10, which is a workpiece to be processed. The scraping is a type of metal processing, and uses a scraper, which is a cutting tool, to scrape off convex portions of the surface 11 to be processed, thereby increasing the flatness of the surface 11 to be processed (processing). (Reduce unevenness of the work surface 11).

[0012] The robot arm 200 is an articulated robot arm, and is controlled by the control device 100. The robot arm 200 has a robot hand 210 at its tip end, which is capable of detachably holding (grasping) the scraper unit 20 and the hand chuck 30. The scraper unit 20 has a support part 23 and a scraper 22 which is composed of a cutting blade 24 attached to the tip end of the support part 23, etc.

[0013] The scraping process for the processing target surface 11 of the workpiece 10 is performed by fixing the workpiece 10 to the processing stand C1 shown in Fig. 1, controlling the robot arm 200 with the scraper unit 20 held by the robot hand 210, and stroking the scraper unit 20 along the processing target surface 11. The surface of the processing stand C1 is formed in a flat shape parallel to the XY plane.

[0014] The measurement stand C2 is provided with a three-dimensional shape measuring instrument 300 and a camera 400, and when measuring the surface 11 to be machined using the three-dimensional shape measuring instrument 300 and the camera 400, the workpiece 10 is The surface of the measurement stage C2 is also formed as a flat surface parallel to the XY plane.

[0015] The three-dimensional shape measuring instrument 300 is a measuring device for measuring the three-dimensional shape of the processing target surface 11 of the workpiece 10, and may be, for example, a white light interference measuring instrument or a three-dimensional laser scanner. The measured data can be acquired as coordinates in the XYZ axis space.

[0016] The camera 400 is an imaging device that captures an image of the surface 11 to be machined of the workpiece 10.

[0017] The lapping stand C3 is a stand on which a mating member 12 (for example, a guide gib or guide rail of a machine tool) can be placed, along which the work surface 11 of the workpiece 10 slides. In this embodiment, the work surface 11 of the workpiece 10 is scraped to match the shape of the reference surface 13 of the mating member 12, so the work surface 11 of the workpiece 10 and the reference surface 13 of the mating member 12 are lapping together on the lapping stand C3. In addition, the automatic scraping device 1 may also include a tool mounting stand C4 for mounting the scraper unit 20, a hand chuck stand C5 for mounting the hand chuck 30, etc.

[0018] 2 is a block diagram showing an example of the configuration of the control device 100. The control device 100 is a control device that controls the robot arm 200, and is, for example, a general-purpose computer. As shown in FIG. 2, the control device 100 includes a communication interface (communication I / F) 101, a storage device 102, an input / output device 103, and a processor (control unit) 104, which are electrically connected to the robot arm 200 via a communication bus 105.

[0019] The processor 104 is an arithmetic processing device, and performs each process according to this embodiment by executing a computer program. The processor 104 loads a program stored in the auxiliary storage device of the storage device 102 into the main storage device and executes it, and processes information input via the communication I / F 101 and the input / output device 103 and information read from the storage device 102, thereby realizing various processes such as a processing instruction data generation process for generating processing instruction data. Information such as the processing instruction data used to control the robot arm 200 is transmitted to the robot arm 200 via the communication I / F 101 and the input / output device 103. Note that the configuration and method by which the control device 100 controls the robot arm 200 to perform scraping can be existing ones, and therefore detailed description thereof will be omitted.

[0020] 3 is a block diagram showing an example of the functional configuration of the control device 100. The control device 100 has a processing instruction data generation unit 110 and a control unit 111 as functional units. The processor 104 of the control device 100 loads a program stored in the auxiliary storage device of the storage device 102 into the main storage device and executes it, thereby realizing each of the above-mentioned functional units. The signal processing unit 100 may be formed of hardware such as a circuit or other digital circuit.

[0021] The processing instruction data generating unit 110 executes a processing instruction data generating process for generating processing instruction data. The control unit 111 acquires the processing instruction data generated by the processing instruction data generating unit 110 and controls the robot arm 200 in accordance with the processing instruction data.

[0022] The method for generating processing instruction data according to the present invention is particularly suitable for generating processing instruction data used in flattening processing. In flattening processing of the processing target surface 11, convex parts of the processing target surface 11 to be cut are acquired based on surface height data, lapping image data, etc., and the flatness of the processing target surface 11 is improved by cutting the convex parts. Hereinafter, the control device 100 will be described in detail when executing flattening processing in the processing device 1 according to this embodiment. Next, a process for generating the processing instruction data will be described.

[0023] In generating the processing instruction data in this embodiment, first, the surface height data of the processing target surface 11 and the fitting image data are acquired.

[0024] FIG. 4 is a diagram illustrating the surface height data of the processing target surface 11. The surface height data of the processing target surface 11 can be acquired based on measurement data from the three-dimensional shape measuring instrument 300. The surface height data is information indicating the height (Z coordinate) corresponding to each coordinate (each measurement point) in the planar direction (XY plane direction) of the processing target surface 11. In the surface height data in FIG. 4, the darker the color (closer to black), the higher the height, and the lighter the color (closer to white), the lower the height. The surface height data is acquired by measuring the processing target surface 11 of the workpiece 10 placed on the measurement stand C2 with the three-dimensional shape measuring instrument 300. In other words, the surface height data can be said to be data (measured height data) that represents a point cloud of height information at any position on the processing target surface 11 of the workpiece 10, without taking into account the deformation state when the workpiece 10 is in a specified assembly state.

[0025] FIG. 5 is a diagram illustrating the lapping image data. The lapping image data can be acquired based on measurement data from the camera 400. A surface inspection agent (such as red rouge or a pigment) is applied to either the reference surface 13 or the workpiece surface 11 of the mating member 12 shown in FIG. 1, and the two are then lapping together. This forms areas where the applied surface inspection agent is peeled off (removed) by the lapping of the reference surface 13 and the workpiece surface 11, and areas where the applied surface inspection agent remains. The workpiece surface 11 after lapping is imaged with the camera 400, and the shading of the surface inspection agent in the obtained image data is binarized to acquire lapping image data (binary data) in which the workpiece surface 11 is divided into areas that are in contact with the reference surface 13 (also referred to as contact areas, black areas in FIG. 5) and areas that are not in contact with the reference surface 13 (also referred to as non-contact areas, white areas in FIG. 5).

[0026] FIG. 6 illustrates the deformation of the workpiece 10 when it is assembled in a predetermined state. As shown in FIG. 6(a), when the workpiece 10 is fastened with a screw S during the lapping operation, the workpiece 10 may be deformed to some extent. Also, as shown in FIG. 6(b), the workpiece 10 may be deformed due to the action of a force F applied when it comes into contact with the mating member 12 and a moment M generated by unevenness on the contact surface. These deformations cause changes in the surface shape of the workpiece surface 11 (changes in surface height). By performing the lapping operation on the workpiece 10 in this state, a lapping image reflecting the shape change of the workpiece surface 11 can be acquired. Meanwhile, the measurement data of the three-dimensional shape measuring instrument 300 is usually measured after the workpiece is released from the predetermined assembly state described above and without contact with the workpiece. Therefore, the surface height data does not reflect the effects of the deformation, making it difficult to accurately estimate the convex portions to be cut as priority areas for machining.

[0027] Therefore, in this embodiment, the surface height data is corrected so that the state of the surface shape of the processing target surface 11 of the workpiece 10 when it is in a predetermined assembly state is reflected, and processing instruction data for flattening is generated in which the convex parts to be cut are more accurately estimated. Details of this will be explained below.

[0028] 7 is a schematic diagram for explaining a method for correcting surface height data in this embodiment, showing how the processing target surface 11 is divided into mesh coordinates mc. As shown in FIG. 7, the processing target surface 11 is divided into meshes at regular intervals, and each mesh coordinate mc is classified as a contact area or a non-contact area with respect to the reference surface 13 in the surface height data and the sliding image data. In the example shown in FIG. 7, the processing target surface 11 is divided into 25 points at 100 pixel intervals in the X direction and 7 points at 20 pixel intervals in the Y direction, and therefore 2 The mesh coordinates mc of 5×7=175 points are classified as contact / non-contact in the above-mentioned aligned image data.

[0029] Since the surface height data does not take into account the deformation state of the machining surface 11 when the workpiece 10 is in a specified assembly state, each mesh coordinate mc may include points that are contact positions in the surface height data but are non-contact positions in the fitting image data, and conversely, points that are non-contact positions in the former data but are contact positions in the latter data.

[0030] FIG. 8 is a schematic diagram showing the binarized surface height data of FIG. 4 and the distribution of contact and non-contact regions on the workpiece surface 11 as predicted by the surface height data. Black areas in FIG. 8 indicate regions predicted to be in contact, and white areas indicate regions predicted not to be in contact. The surface height data may be divided into contact and non-contact positions by, for example, calculating a representative value AS (see FIG. 9 ) from point cloud data of height information included in the surface height data, and dividing areas higher than the representative value AS into contact regions and areas lower than the representative value AS into non-contact regions. In this embodiment, the representative value is the average value of the height information, but this is not limited thereto. For example, the representative value may be the median or the mode. As can be seen from a comparison with FIG. 5 , which shows the rubbing image data, the distribution of contact and non-contact regions shown in FIG. 8 differs from the distribution of contact and non-contact regions shown in FIG. 5 because it does not take into account the deformation state of the workpiece surface 11.

[0031] Fig. 9 is a diagram for explaining a method for correcting surface height data in this embodiment, and is a schematic diagram showing the unevenness of the processing target surface 11 based on the surface height data as viewed from the side. The solid line Hm in Fig. 9 indicates the uneven shape (actual measured height) of the processing target surface 11 based on the surface height data, and the dashed line Ha indicates the uneven shape (estimated height) of the processing target surface 11 that is expected when the workpiece 10 is in a specified assembly state. AS in Fig. 9 indicates the representative value (average value) of the height information included in the surface height data described above.

[0032] Point A in FIG. 9 indicates a point that is lower than the representative value AS in the surface height data and is a non-contact position, but is a contact position in the aligned image data. For such points, a correction is made by adding the height information (measured height Hm) in the surface height data to approach the assumed height Ha indicated by the dashed line (correction is made using a second correction amount, which is an additive correction amount). The addition amount using the second correction amount may be, for example, an addition amount corresponding to the difference between the measured height Hm and the assumed height Ha. That is, the second correction amount may be a correction amount that increases as the absolute value of the difference increases. Point B in FIG. 9 indicates a point that is higher than the representative value AS in the surface height data and is a contact position, but is a non-contact position in the aligned image data. For such points, a correction is made by subtracting the measured height Hm to approach the assumed height Ha (correction is made using a first correction amount, which is a subtraction correction amount). The subtraction amount using the first correction amount may be, for example, an addition amount corresponding to the difference between the measured height Hm and the assumed height Ha. That is, the first correction amount may be a correction amount in which the subtraction amount increases as the absolute value of the difference increases. Note that no correction is performed at points where both the surface height data and the aligned image data are in contact areas or where both are in non-contact areas. That is, the correction amount at these points is zero.

[0033] In the surface height data shown in Fig. 4, there are 99 mesh coordinates mc that are in the contact area, but there are 24 points where a subtraction correction is performed, such as point B. Also, there are 76 mesh coordinates mc that are in the non-contact area, but there are 32 points where an addition correction is performed, such as point A.

[0034] At each point of all mesh coordinates mc shown in FIG. 7, the deviation amount between the surface height data and the representative value AS is calculated, and the deviation amount of all mesh coordinates mc obtained is plotted three-dimensionally. A correction amount surface (correction amount data) is obtained, which is surface approximation data in which the correction amount at any position is represented as a point cloud. An example of an image of this correction amount surface (a quartic approximation surface of ordinate coordinates X and Y) is shown in Figure 10. By superimposing this correction amount surface on the surface height data, post-correction surface height data is obtained.

[0035] 11(a) shows the surface height data before the above-described correction, and FIG. 11(b) shows the surface height data after the above-described correction. In particular, in the region enclosed by the dashed circle in FIG. 11(b), i.e., the region of the longitudinal end of the workpiece surface 11, the height information has been corrected to one in which high-height regions are concentrated, similar to the concentration of contact regions at the longitudinal end in FIG. 5. The surface height data based on this corrected height information is assumed height data that assumes the deformation state of the workpiece surface 11 when the workpiece 10 is in a specified assembly state. This makes it possible to bring the height information of the surface height data closer to the height in the specified assembly state.

[0036] Based on the corrected surface height data obtained as described above, the convex portions of the workpiece surface 11 to be cut can be appropriately obtained, and more appropriate processing instruction data can be created. For example, based on the corrected surface height data and the fitted image data, the positions of the convex portions to be cut can be identified, and processing instruction data can be created so that the positions are included in the cutting target area.

[0037] <Embodiment 2> An automatic scraping device according to a second embodiment of the present invention will be described. Here, differences between the second embodiment and the first embodiment will be mainly described, and explanations of commonalities with the first embodiment will be omitted. Items in the second embodiment that are not specifically described here are the same as those in the first embodiment.

[0038] 12 to 14, the algorithm for correcting the surface height data in embodiment 2 will be described. In this embodiment, an identification surface fa is obtained from the height information of the fitted image data and the surface height data, and a deformation reproduction surface f (FIG. 14), which is a virtual surface that reproduces the deformation of the processing surface 11 when the workpiece 10 is in a predetermined assembly state using the identification surface fa, is obtained as correction amount data.

[0039] 12 is a diagram showing the relationship between the measured surface (measured height) Hm and the identification surface fa in embodiment 2. The measured surface Hm shown by the solid line in Fig. 12 indicates the uneven shape of the surface 11 to be machined based on the surface height data, and the contact area CP shown by the diagonal lines corresponds to the contact area in the lapping image data. It is assumed that the area of ​​the measured surface Hm that overlaps with the contact area CP will form a convex shape on the surface 11 to be machined when the workpiece 10 is in a predetermined assembled state.

[0040] The discrimination surface fa is generated so that it passes below the measured surface Hm in the contact region CP and passes above the measured surface Hm in the non-contact region outside the contact region CP. The discrimination surface fa, which has a curved surface shape that satisfies these conditions, can be found using polynomial feature values ​​for the x, y coordinates of the surface to be machined, which are generated from the surface height data. The fitted image data is used as training data for contact / non-contact, and a linear classifier is trained so that the polynomial feature values ​​and surface height data match the output of the training data. The discrimination surface fa can then be obtained by linearly combining the polynomial feature values ​​using the coefficients of the trained linear classifier.

[0041] For example, in the case of a two-dimensional surface (curve), the discrimination surface fa can be expressed by the following equation: TIFF2026003488000002.tif9170 In the case of a three-dimensional curved surface, it can be expressed by the following formula: TIFF2026003488000003.tif9125The coefficient vector W of the polynomial that determines the discrimination surface fa can be obtained using a linear discriminator such as a linear support vector machine, stochastic gradient descent, or logistic regression.

[0042] 13A and 13B are schematic diagrams illustrating adjustment of the deformation amount of the acquired discrimination surface fa. As shown in FIG. 13A, the fitted image data only reveals the positions of the contact and non-contact areas on the workpiece surface 11, but it is not possible to determine the type of deformation occurring on the workpiece surface 11 in the non-contact areas. In other words, it is not possible to completely estimate the degree of curvature of the workpiece surface 11. Although multiple discrimination surfaces can be assumed, such as the discrimination surface fa1 and the discrimination surface fa2 shown in FIG. 13A, it is necessary to adjust (limit) the deformation amount of the discrimination surface fa so as not to include a discrimination surface fa that represents a state in which the workpiece 10 has been deformed into a shape that deviates from reality.

[0043] To adjust the amount of deformation, a scaling parameter s is calculated, and if s≦1, scaling is performed by setting fb=sfa. TIFF2026003488000004.tif27170Here, Δfa is the deformation amount of the curved surface, σ is the standard deviation of the height of the surface 11 to be processed, and K is the gain (fixed value).

[0044] 13(b) is calculated from a constant multiple (=Kσ) of the standard deviation of the height of the processing target surface 11. In other words, the acquired identification surface fa is corrected so that it falls within a height range obtained by multiplying the standard deviation of the processing target surface 11 by a constant multiple, and the corrected identification surface fb is acquired.

[0045] 14 is a conceptual diagram of the acquisition of the deformed reproduction surface f and the correction process in which the deformed reproduction surface f is superimposed on the measured surface Hm of the processing surface 11 based on the surface height data to acquire the assumed surface Hc, which is the corrected surface height data. In generating the identification surface fa, the magnitude relationship between the binary values ​​(contact area / non-contact area) of the fitted image data is set so that the measured surface Hm in the contact area CP is higher than the identification surface fa. Therefore, the surface fc obtained by flipping the corrected identification surface fb upside down can be used as the deformed reproduction surface f. Note that the method for generating the identification surface fa is not particularly limited. For example, if the magnitude relationship between the binary values ​​(contact area / non-contact area) of the fitted image data is set inversely to that described above, the corrected identification surface fb can be acquired as is as the deformed reproduction surface f.

[0046] The assumed surface Hc obtained by superimposing the deformed reproduction surface f on the measured surface Hm becomes assumed height data (corrected surface height data) that assumes the deformation state of the processing target surface 11 when the workpiece 10 is in a specified assembly state. Using the assumed height data obtained in this way, the convex portion of the processing target surface 11 to be cut can be estimated, and more appropriate processing instruction data can be created. For example, the position TP of the convex portion to be cut can be identified based on the assumed height data based on the assumed surface Hc and the mated image data, and processing instruction data can be created so that the position TP is included in the cutting target area.

[0047] FIG. 15 shows an example of the results when the surface height data correction method according to this embodiment is applied (a specific example that can confirm the validity of the correction algorithm of this embodiment). FIG. 15(a) is an example of fitted image data that serves as training data for generating a deformed reproduction surface. Light-colored areas are contact areas, and dark-colored areas are non-contact areas. FIG. 15(b) shows binarized image data obtained by surface approximation correction, which uses FIG. 15(a) as training data and trains a linear classifier to learn polynomial features and surface height data so that they match the output of the training data. In contrast, FIG. 15(c) shows planar binarized image data obtained by planar approximation correction as a comparative example. As shown in the areas surrounded by dashed circles in FIGS. 15(b) and 15(c), it can be seen that the contact / non-contact approximation that could not be reproduced by the planar approximation correction in FIG. 15(c) is reproduced by the surface approximation in FIG. 16(b). Figure 15(d) shows a classification surface generated by linearly combining polynomial feature values ​​from the coefficients of the trained linear classifier. It can be seen that the estimated deformation is represented as a curved surface. Figure 15(e) shows the corrected surface height data, in which the deformation reproduction surface after scaling transformation is superimposed on the surface height data. As can be seen from a comparison with the fitted image data in Figure 15(a), the contact area in the corrected surface height data has been corrected to take into account the deformation state of the workpiece. [Explanation of symbols]

[0048] 1...Automatic scraping processing equipment 10. Work 11. Machining surface 100 Control device 104 Processor 110 Processing instruction data generation unit 111 Control unit 200···Robot arm 300...3D shape measuring instrument 400···Camera

Claims

1. A processing device having a processor that generates processing instruction data, and automatically scraping a surface to be processed of a workpiece based on the processing instruction data, The processor: When the workpiece is in a predetermined assembly state and the processing target surface is rubbed against a reference surface, binary data is acquired which classifies an arbitrary position of the processing target surface as either a contact position or a non-contact position with respect to the reference surface; acquiring measured height data in the form of a point cloud of height information at any position on the surface of the workpiece not in the predetermined assembled state; acquiring correction amount data representing a point cloud of correction amounts for the height information at any position on the processing object surface based on the binary data and the measured height data; by superimposing the correction amount data on the measured height data, estimated height data in the predetermined assembly state is obtained; A convex portion of the surface to be machined is identified based on the assumed height data, and processing instruction data for cutting the convex portion is generated. A processing device characterized by:

2. The correction amount is a first correction amount by which the height information is subtracted when the arbitrary position is the non-contact position in the binary data and is equal to or greater than a predetermined height in the measured height data; a second correction amount to be added to the height information when the arbitrary position corresponds to the contact position in the binary data and is less than the predetermined height in the measured height data; Including, The processing device according to claim 1 .

3. The predetermined height is a representative value of the height information. The processing device according to claim 2 .

4. the first correction amount is a subtraction amount based on a difference between a height in the binary data and a height in the measured height data, the second correction amount is an addition amount based on a difference between a height in the binary data and a height in the measured height data. The processing device according to claim 2 .

5. the correction amount is acquired based on the height of a virtual curved surface representing the machining target surface predicted when the workpiece is in the predetermined assembled state and the measured height data. The processing device according to claim 1 .

6. The virtual curved surface is adjusted so that the height range is within a constant multiple of the standard deviation of the measured height data. The processing device according to claim 5.

7. the binary data is acquired by placing the workpiece in the predetermined assembled state, applying a surface inspection agent to either the processing target surface or a reference surface of the processing target surface, and then rubbing the processing target surface against the reference surface, and then imaging the processing target surface. The processing device according to claim 1 .

8. the measured height data is acquired by measuring the processing target surface of the workpiece not in the predetermined assembled state with a three-dimensional shape measuring device; The processing device according to claim 1 .

Images