WO2021186545A1

WO2021186545A1 - Wire electrical discharge machining machine and machine learning apparatus

Info

Publication number: WO2021186545A1
Application number: PCT/JP2020/011699
Authority: WO
Inventors: 信行太田; 大介関本
Original assignee: 三菱電機株式会社
Priority date: 2020-03-17
Filing date: 2020-03-17
Publication date: 2021-09-23
Also published as: CN115243817A; CN115243817B; JPWO2021186545A1; JP6775719B1

Abstract

A wire electrical discharge machining machine (100) is provided with a wire electrode (2) discharging electricity between the wire electrode (2) and a workpiece (7) as the wire electrode (2) is torsionally rotating, to machine the workpiece (7); an upper guide (6) supporting the wire electrode (2) above the workpiece (7); a lower guide (11) supporting the wire electrode (2) below the workpiece (7); and a controller (14). The controller (14) is for adjusting at least one of the offset amount of the wire electrode (2) with respect to a face to be machined of the workpiece (7), and the inclination of the wire electrode (2), by correction of the positions of the upper guide (6) and the lower guide (11) individually with values respectively determined for a case where part of the workpiece (7) which is to remain is positioned on the right side with respect to a machining direction and for a case where the part is positioned on the left side with respect thereto, based on the positional relationship between the machining direction and the part, and a machining program; and compensating a machining dimensional error that is an error in the dimensions of an actual shape of a machined object to be obtained by machining the workpiece (7) compared to the dimensions of a target shape thereof.

Description

Wire EDM and Machine Learning Equipment

The present disclosure relates to a wire electric discharge machine and a machine learning device that machine a workpiece using a wire electrode.

EDM is to melt and remove the workpiece by the heat of the discharge generated between the electrode and the workpiece. When a workpiece is machined by wire electric discharge machining, the machined surface of the workpiece after machining, that is, the surface after machining is separated from the wire electrode by the discharge gap generated between the workpiece and the wire electrode. Further, since the wire electrode is conveyed from the upper surface side of the workpiece toward the lower surface of the workpiece while generating an electric discharge facing the workpiece, it is consumed before being conveyed to the lower surface of the workpiece. In wire electric discharge machining, an offset is provided in the wire electrode locus with respect to a desired machined surface in consideration of these discharge gaps and the amount of wire electrode consumption (for example, Patent Document 1). In the invention described in Patent Document 1, by performing machining while changing the offset amount set in the wire electrode locus, it is possible to improve the machining accuracy when machining a workpiece (workpiece) whose plate thickness changes. do.

Japanese Unexamined Patent Publication No. 53-83192

In the wire electric discharge machine, the wire electrode is conveyed while rotating in a certain direction due to the bobbin winding habit and friction with the roller in the transfer path. This rotation of the wire electrode causes deterioration of processing accuracy. When the wire electrode is conveyed while rotating, the amount of wear on the side surface of the wire electrode differs between the right side and the left side with respect to the machining direction on the lower surface of the workpiece. Therefore, on the lower surface of the workpiece, the position of the center of the machined groove after machining is displaced with respect to the wire electrode locus by the difference in the amount of wear on the side surface of the wire electrode, and as a result, the position of the machined surface is displaced and a machining error occurs. ..

In conventional wire electric discharge machining such as the invention described in Patent Document 1, since the rotation of the wire electrode during machining is not considered, it is not possible to compensate for the machining error generated by the rotation of the wire electrode.

The present disclosure has been made in view of the above, and an object of the present disclosure is to obtain a wire electric discharge machining apparatus capable of improving machining accuracy.

In order to solve the above-mentioned problems and achieve the object, the wire discharge processing apparatus according to the present disclosure includes a wire electrode for processing a work piece by discharging electricity from the work piece while rotating in the twisting direction, and a work piece. It is provided with an upper guide that supports the wire electrode above and a lower guide that supports the wire electrode below the workpiece. Further, the wire discharge processing apparatus is based on the positional relationship between the machining direction and the workpiece to be left and the machining program, depending on whether the workpiece to be left is located on the right side or the left side with respect to the machining direction. By correcting the positions of the upper guide and the lower guide with individually determined values, at least one of the offset amount of the wire electrode with respect to the machined surface and the inclination of the wire electrode is adjusted, and the work is machined. A control device for compensating for a machining dimensional error, which is an error between the dimension of the target shape of the work piece obtained and the dimension of the actual shape, is provided.

The wire electric discharge machining apparatus according to the present disclosure has the effect of being able to improve the machining accuracy.

The figure which shows the structural example of the wire electric discharge machining apparatus which concerns on Embodiment 1. The figure which shows the structural example of the control device of the wire electric discharge machine which concerns on Embodiment 1. The figure which shows the other configuration example of the wire electric discharge machining apparatus which concerns on Embodiment 1. A bird's-eye view of the machining groove formed as a result of wire electric discharge machining of a workpiece The first figure for demonstrating the problem that a wire electric discharge machine tries to solve. The second figure for demonstrating the problem that a wire electric discharge machine tries to solve. FIG. 3 for explaining the problem to be solved by the wire electric discharge machine. FIG. 4 for explaining the problem to be solved by the wire electric discharge machine. The figure which shows the outline of the processing operation by the wire electric discharge machining apparatus which concerns on Embodiment 1. The figure which shows another example of the correction method by the wire electric discharge machining apparatus which concerns on Embodiment 1. The figure which shows the coordinate axis used when the wire electric discharge machine makes correction. A first flowchart showing an example of an operation in which the wire electric discharge machining apparatus according to the first embodiment corrects a machining dimensional error in wire electric discharge machining. A second flowchart showing an example of an operation in which the wire electric discharge machining apparatus according to the first embodiment corrects a machining dimensional error in wire electric discharge machining. The figure which shows the connection relation of each component which realizes the operation example shown in FIG. 13 of the wire electric discharge machining apparatus which concerns on Embodiment 1. FIG. A third flowchart showing an example of an operation in which the wire electric discharge machining apparatus according to the first embodiment corrects a machining dimensional error in wire electric discharge machining. The figure which shows the connection relation of each component which realizes the operation example shown in FIG. 15 of the wire electric discharge machining apparatus which concerns on Embodiment 1. FIG. The figure which shows the relation example of the processing direction of wire electric discharge machining, and the consumption amount of a wire electrode. The figure which shows the relation example of the processing direction of wire electric discharge machining, the rotation direction of a wire electrode, and the consumption amount of a wire electrode. The figure which shows the structural example of the control apparatus included in the wire electric discharge machining apparatus which concerns on Embodiment 2.

The wire electric discharge machine and the machine learning device according to the embodiment of the present disclosure will be described in detail below with reference to the drawings.

Embodiment 1.
FIG. 1 is a diagram showing a configuration example of the wire electric discharge machine according to the first embodiment. The wire discharge processing apparatus 100 includes a wire electrode bobbin 1, a wire electrode 2 drawn from the wire electrode bobbin 1, a wire electrode transfer roller 3 for conveying the wire electrode 2, and a wire electrode 2 during processing of the workpiece 7. The upper guide 6 and the lower guide 11 that support the position and inclination of the above, the table 8 for installing the workpiece 7, the lower roller 13 that conveys the wire electrode 2 after being machined by the machined portion 16, and the wire electrode 2 A recovery roller 9 for generating a driving force for carrying the wire electrode 2 and a wire electrode recovery box 10 for discarding the processed wire electrode 2 are provided. The processed portion 16 is composed of an upper guide 6 and a lower guide 11, and a wire electrode 2 supported by the upper guide 6 and the lower guide 11.

Further, the wire discharge processing apparatus 100 includes a plate thickness detector 4 for detecting the plate thickness of the workpiece 7, a wire electrode rotation detector 5 for detecting the rotation direction, rotation speed or rotation angle of the wire electrode 2, and a wire electrode. The wire electrode consumption measuring instrument 12 for measuring the consumption of 2 is provided. The plate thickness detector 4 detects the plate thickness of the workpiece 7 in the vertical direction, that is, in the transport direction of the wire electrode 2, and generates plate thickness information. The "rotation angle of the wire electrode 2" in the present embodiment is the angle at which the wire electrode 2 rotates while machining the workpiece 7, that is, the arbitrary position of the wire electrode 2 is the upper guide 6 of the workpiece 7. The angle is set so that the workpiece 7 is rotated while being electric-discharged while being conveyed from the side to the lower guide 11 side. The rotation speed information is configured to include information on the rotation speed of the wire electrode 2 and information on the rotation direction. Further, the rotation angle information is configured to include the above-mentioned information on the rotation angle of the wire electrode 2 while processing the workpiece 7 and information on the rotation direction. The "consumable amount of the wire electrode 2" is the amount of the wire electrode 2 consumed while the workpiece 7 is being machined, that is, the wire electrode 2 is moved from the upper guide 6 side to the lower guide 11 side of the workpiece 7. The amount of decrease in the diameter of the wire electrode 2 during transportation.

Further, the wire discharge processing apparatus 100 includes plate thickness information indicating the plate thickness detected by the plate thickness detector 4, rotation information of the wire electrode 2 indicating the result detected by the wire electrode rotation detector 5, and wire electrode. Based on the consumption information indicating the consumption amount measured by the consumption amount measuring device 12, it is estimated by the machining dimension displacement estimator 15 and the machining dimension displacement estimator 15 for estimating the machining groove center displacement amount or the machining surface position displacement amount. It is provided with a control device 14 that controls the positions of the upper guide 6 and the lower guide 11 based on the displacement amount of the center of the machined groove or the displacement amount of the machined surface position and the machining program.

FIG. 2 is a diagram showing a configuration example of the control device 14 of the wire electric discharge machining device 100 according to the first embodiment. As shown in FIG. 2, the control device 14 includes an NC (Numerical Control) device 141 and a drive unit 142. Here, the NC device 141 and the drive unit 142 are realized by, for example, a control circuit in which a processor and a memory are combined. That is, the control device 14 includes a processor and a memory that realizes the NC device 141 and the drive unit 142. The memory corresponds to, for example, a non-volatile or volatile semiconductor memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), and a flash memory. The NC device 141 and the drive unit 142 are realized by the processor reading and executing a program stored in a memory for operating as the NC device 141 and the drive unit 142. The NC device 141 and the drive unit 142 may be realized by an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a circuit combining these.

The NC device 141 analyzes the machining program and corrects the wire electrode position command, which is a position command for the wire electrode 2, among the commands described in the machining program, based on the machining groove center displacement amount or the machining surface position displacement amount. do. The NC device 141 outputs the corrected wire electrode position command to the drive unit 142. The drive unit 142 obtains the command positions of the upper guide 6 and the lower guide 11 based on the corrected wire electrode position command received from the NC device 141, and the upper guide 6 and the lower guide 11 move to the obtained command positions. The upper guide 6 and the lower guide 11 are driven so as to do so. The command position is a position commanded by a position command described in the machining program. Here, when the drive unit 142 obtains the command positions of the upper guide 6 and the lower guide 11, for example, the drive unit 142 first analyzes the corrected wire electrode position command to obtain the command position of the wire electrode 2, and obtains the command position of the wire electrode. The command positions of the upper guide 6 and the lower guide 11 are calculated from the command position and the positional relationship between the wire electrode 2, the upper guide 6 and the lower guide 11. Here, the position commanded by the command position of the wire electrode 2 means the position of the wire electrode 2 relative to the workpiece 7. Therefore, the command position of the upper guide 6 indicates the relative position of the upper guide 6 with respect to the workpiece 7, and the command position of the lower guide 11 indicates the relative position of the lower guide 11 with respect to the workpiece 7. The positional relationship between the wire electrode 2, the upper guide 6 and the lower guide 11 is fixed if there is no change in the inclination of the wire electrode 2, and the command positions of the upper guide 6 and the lower guide 11 are calculated based on the command position of the wire electrode 2. It is possible to do. The drive unit 142 holds information on the default positional relationship between the wire electrode 2, the upper guide 6, and the lower guide 11, and is based on this information, the command position of the wire electrode 2, and the inclination of the wire electrode 2. , The command positions of the upper guide 6 and the lower guide 11 are calculated. Information on the inclination of the wire electrode 2 is obtained by analyzing the machining program. The NC device 141 may calculate the command positions of the upper guide 6 and the lower guide 11. Further, it is assumed that the wire electrode position command is described in the machining program, but the position command for each of the upper guide 6 and the lower guide 11 is described in the machining program, and the NC device 141 describes the upper guide 6 and the lower guide 11. The position command for each of the above may be corrected, and the corrected position command may be output to the drive unit 142.

In the configuration example shown in FIG. 1, the machining dimension displacement estimator 15 and the control device 14 are separately configured, but as shown in FIG. 3, these may be integrated. FIG. 3 is a diagram showing another configuration example of the wire electric discharge machining apparatus according to the first embodiment. The wire electric discharge machining device 101 shown in FIG. 3 includes a control device 18 in which these are integrated, instead of the machining dimension displacement estimator 15 and the control device 14 of the wire electric discharge machining device 100 shown in FIG. In the following description, the configuration shown in FIG. 1 is assumed.

Here, the problem to be solved by the wire electric discharge machine 100 according to the present embodiment will be described.

FIG. 4 is a bird's-eye view of a machined groove formed as a result of wire electric discharge machining of a workpiece. In FIG. 4, the center line of the machining groove 21 formed after the wire electrode 2 has passed is defined as the machining groove center 20, and the wall surface of the workpiece forming the machining groove 21 is defined as the machining surface 19.

FIG. 5 is a first diagram for explaining a problem to be solved by the wire electric discharge machining apparatus 100. FIG. 6 is a second diagram for explaining a problem to be solved by the wire electric discharge machine 100. In FIG. 5, it is assumed that the wire electrode 2 moves in the machining direction 22 and is transported in the wire electrode transport direction 26 while machining the workpiece, and the wire electrode 2 is rotating in the rotation direction 24. .. In this case, the wire electrode 2 has a different wire cross-sectional shape between the upper part 23 of the workpiece and the lower part 25 of the workpiece due to wear due to machining. Specifically, the wire electrode 2 wears only the surface in the machining direction 22 at the upper part 23 of the workpiece. On the other hand, in the lower part 25 of the workpiece, the wire electrode consumable surface 27 and the wire electrode non-consumable surface 28 are present on the left and right sides of the wire electrode 2 with respect to the machining direction 22 as the wire electrode 2 is conveyed and rotated. In the lower part 25 of the workpiece, the wire electrode is actually worn and the shape is changed with respect to the shape 39 of the wire electrode 2 when there is no wear. Therefore, considering the discharge gap 31, the machining direction of the workpiece The position of the machined surface deviates from the ideal target position between the work piece 32 on the left side and the work piece 33 on the right side in the machining direction. As shown in FIG. 6, the actual machined surface position 37 at the lower part 25 of the workpiece is the amount of displacement of the machined surface position on the left side of the wire electrode with respect to the machined direction 22 with respect to the target machined surface which is the ideal machined surface position 38. There is a deviation between 34 and the machined surface position displacement 35 on the right side of the wire electrode. Further, the actual machined groove center 29 at the lower part 25 of the workpiece also deviates from the machine-commanded wire electrode center 30 by the amount of displacement 36 at the center of the machined groove.

That is, as shown in FIG. 7, in the lower part 25 of the workpiece, the actual machined groove center 29 deviates from the machine-commanded wire electrode center 30 by the machined groove center displacement amount 36. As a result, the position and inclination of the machined surface of the workpiece 32 on the left side of the machining direction and the position and tilt of the machined surface of the workpiece 33 on the right side of the machining direction are different from the ideal ones. FIG. 7 is a third diagram for explaining a problem to be solved by the wire electric discharge machine 100.

Further, as shown in FIG. 8, if the machining direction 22 is different, the direction in which the actual machining groove center 29 in the lower part 25 of the workpiece deviates from the wire electrode center 30 of the machine command is different. That is, the direction in which the actual machining groove center 29 deviates from the machine-commanded wire electrode center 30 in the lower part 25 of the workpiece differs depending on the machining direction 22. This means that the inclination and position of the machined surface caused by the actual machined groove center 29 deviating from the machine-commanded wire electrode center 30 differ depending on the machined direction 22. FIG. 8 is a fourth diagram for explaining a problem to be solved by the wire electric discharge machine 100. Similarly, when the rotation direction 24 of the wire electrode 2 is reversed from the direction shown in the figure, the direction in which the actual machining groove center 29 deviates from the machine-commanded wire electrode center 30 in the lower part 25 of the workpiece is in the machining direction 22. It depends.

In the conventional wire electric discharge machining, the radius of the wire electrode and the electric discharge gap are set as the offset of the center locus of the wire electrode of the machine command in consideration of the direction in which the workpiece to be left exists. The offset here is the distance between the machined surface of the workpiece and the center of the machined groove. However, the machining direction is not taken into consideration when setting this offset amount. Therefore, the problem described with reference to FIGS. 5 to 8 above is not addressed.

On the other hand, in the wire electric discharge machining apparatus 100 according to the present embodiment, the actual machining groove center 29 at the lower part 25 of the workpiece shown in FIGS. 5 to 8 and the wire electrode center 30 of the machine command are deviated from each other, or It has a function of correcting the position and inclination of the wire electrode 2 based on the deviation between the actual machined surface position 37 and the ideal machined surface position 38 at the lower part 25 of the workpiece shown in FIG. Specifically, as shown in FIG. 9, the wire electric discharge machining apparatus 100 has a function of correcting the central locus of the wire electrode 2 according to the machining direction. FIG. 9 is a diagram showing an outline of a machining operation by the wire electric discharge machining apparatus 100 according to the first embodiment. FIG. 9A shows a wire electrode center locus represented by a machine command of wire electric discharge machining (wire electrode command position locus of a machine command) and a target shape of a workpiece created by wire electric discharge machining. FIG. 9B shows an actual wire electrode center locus when processing is performed by moving the wire electrode clockwise (CW: clockwise) along the wire electrode command position locus of the machine command shown in FIG. 9A. And the shape of the work piece to be created. FIG. 9 (c) shows the actual wire electrode center when the wire electrode is moved counterclockwise (CCW: counterclockwise) along the wire electrode command position locus of the machine command shown in FIG. 9 (a) for processing. The trajectory and the shape of the work piece to be created are shown. As shown in FIGS. 9 (a) to 9 (c), the wire electrode center locus of the machine command and the actual wire electrode center locus are different. Further, even if the wire electrode center locus of the machine command is the same, if the machining direction, which is the direction in which the wire electrode moves, is different, the actual wire electrode center locus does not match, and the shape of the created workpiece also matches. do not. Therefore, as shown in FIG. 9D, the wire electric discharge machining apparatus 100 according to the present embodiment corrects the position of the center of the wire electrode with a different correction amount for each machining direction, and performs machining in any machining direction. Even if this is done, a work piece with a target shape can be obtained. The “wire electrode command position locus before correction” in FIG. 9 (d) coincides with the “machine command wire electrode command position locus” in FIG. 9 (a). That is, the wire electric discharge machining apparatus 100 corrects the position of the wire electrode commanded by the machine command with a correction amount determined in consideration of the machining direction, and controls the position of the wire electrode so as to be the corrected position. ..

In the following description, the deviation between the actual machined groove center 29 and the machine-commanded wire electrode center 30 shown in FIG. 6 or the difference between the actual machined surface position 37 and the ideal machined surface position 38. The correction is made based on the above, but as shown in FIG. 10, the correction is based on the average dimension of the shape obtained by processing the workpiece, that is, the correction so that the shape obtained by processing becomes the average dimension. May be done. FIG. 10 is a diagram showing another example of the correction method by the wire electric discharge machine 100 according to the first embodiment. Further, in the wire electric discharge machine 100, the command value of the UV axis or the command value of the XY axis shown in FIG. 11 is used to correct the offset amount of the wire electrode 2 and the inclination of the wire electrode 2. To correct. That is, when correcting the inclination of the wire electrode 2, the command value of the UV axis is corrected, and when correcting the offset amount of the wire electrode 2, the command value of the XY axis is corrected. The command value of the UV axis is defined by the position in the UV coordinates defined by the orthogonal U-axis and the V-axis, that is, the command value of commanding the coordinates, and the command value of the XY axis is defined by the orthogonal X-axis and the Y-axis. It is a command value that commands a position in XY coordinates, that is, coordinates. The correction of the offset amount of the wire electrode 2 means the correction of the relative position of the wire electrode 2 with respect to the workpiece 7. FIG. 11 is a diagram showing coordinate axes used when the wire electric discharge machine 100 performs correction.

Next, refer to FIGS. 12 to 16 for the operation of the wire electric discharge machining apparatus 100 according to the present embodiment, specifically, the operation of the wire electric discharge machining apparatus 100 for correcting the machining dimensional error generated in the wire electric discharge machining. I will explain while. The machining dimensional error is an error between the dimension of the target shape of the workpiece obtained by wire electric discharge machining of the workpiece and the dimension of the actual shape. The machining dimensional error includes the displacement of the center of the machining groove, which is the amount of deviation between the actual machining groove center 29 and the wire electrode center 30 of the machine command, or the actual machining surface position 37 and the ideal machining surface position. The machined

surface position displacements

34 and 35, which are the amount of deviation from 38, correspond (see FIG. 6).

FIG. 12 is a first flowchart showing an example of an operation in which the wire electric discharge machining apparatus 100 according to the first embodiment corrects a machining dimensional error in wire electric discharge machining.

In the operation example shown in FIG. 12, first, the control device 14 of the wire electric discharge machining device 100 acquires a machining dimensional error (step S11). That is, the control device 14 acquires the machined groove center displacement amount 36 or the machined surface position displacement amounts 34 and 35. The control device 14 acquires this information from the machined dimensional displacement estimator 15. The wire electrode 2 is consumed as the electric discharge machining progresses, and the actual machined groove center 29 gradually deviates from the machine-commanded wire electrode center 30. However, if the wire electrode transport speed is constant, the wire electrode 2 is consumed. The amount of deviation per hour is almost constant. Similarly, the amount of deviation between the actual machined surface position 37 and the ideal machined surface position 38 per hour is also substantially constant. Therefore, the wire discharge processing apparatus 100 holds the machining dimensional error (the amount of displacement at the center of the machining groove 36 or the amount of displacement of the machining surface position 34 and 35) that has been converted into data in advance in the memory, and the control device 14 holds the memory. It is also possible to obtain the machining dimensional error from. The machining dimensional errors digitized in advance include, for example, the distance from the reference position on the wire electrode 2 in the wire electrode transport direction 26 (for example, the position where the discharge to the workpiece 7 is started), the machining direction, and the machining. The data is tabulated with the dimensional error associated with it. In addition to this information, information such as the thickness of the workpiece, the material of the wire electrode, the material of the workpiece, the transport speed of the wire electrode, the thickness of the wire electrode, the machining speed, and the machining voltage are stored in a table in the memory. You may. Various information registered in the table may be acquired, for example, by repeatedly performing a simulation under various conditions.

After acquiring the machining dimensional error, the control device 14 corrects the wire electrode position command based on the positional relationship between the machining direction and the workpiece to be left and the machining dimensional error (step S12). Specifically, the NC device 141 of the control device 14 determines on which side of the machining direction the workpiece is desired to be left with respect to the machining direction, and compensates for the machining dimensional error based on the judgment result. Determine the amount of correction for. That is, the NC device 141 determines the correction amount individually in each case by dividing into the case where the workpiece to be left exists on the right side with respect to the machining direction and the case where the workpiece to be left exists on the left side. .. Then, the NC device 141 corrects the wire electrode position command with the determined correction amount. That is, in the NC device 141, the work piece that the wire electrode 2 wants to leave is due to the machining dimensional error (machine groove center displacement amount 36, machined surface position displacement amount 34, 35) that occurs on the side where the work piece to be left is located. Correct the wire electrode position command to move to a certain side. From the analysis result of the machining program, the NC apparatus 141 determines on which side the machining direction and the workpiece to be left exist with respect to the machining direction.

Next, the control device 14 moves the wire electrode 2 according to the corrected wire electrode position command (step S13). Specifically, the drive unit 142 of the control device 14 drives the upper guide 6 and the lower guide 11 so that the wire electrode 2 moves to the position commanded by the corrected wire electrode position command.

The control device 14 of the wire electric discharge machining device 100 reads and executes each command described in the machining program in order, and when the wire electrode position command is read, executes each of the above steps S11 to S13.

FIG. 13 is a second flowchart showing an example of an operation in which the wire electric discharge machining apparatus 100 according to the first embodiment corrects a machining dimensional error in wire electric discharge machining. FIG. 14 is a diagram showing a connection relationship of each component that realizes an operation example shown in FIG. 13 of the wire electric discharge machining apparatus 100 according to the first embodiment.

The second flowchart shown in FIG. 13 is such that steps S21 to S23 are executed instead of step S11 of the first flowchart shown in FIG. Since steps S12 and S13 in FIG. 13 are the same as steps S12 and S13 in FIG. 12, detailed description thereof will be omitted.

First, the machining dimension displacement estimator 15 of the wire electric discharge machining device 100 acquires information regarding the rotation of the wire electrode 2 (step S21). The information regarding the rotation of the wire electrode 2 acquired in step S21 is any one of the rotation direction, the rotation angle, and the rotation speed of the wire electrode 2 during transportation. The machined dimension displacement estimator 15 acquires rotation information, which is information on the rotation of the wire electrode 2, from the wire electrode rotation detector 5. It is considered that the rotation direction, the rotation angle, and the rotation speed of the wire electrode 2 are constant in the state where the processing is normally performed. Therefore, the wire discharge processing apparatus 100 holds in advance data on the rotation of the wire electrode 2 (rotation direction, rotation speed, or rotation angle of the wire electrode 2) in the memory, and the processing dimension displacement estimator 15 is used. , It is also possible to acquire information on the rotation of the wire electrode 2 from the memory. It should be noted that there is no problem even if the order of execution of this step S21 and step S22 described below is changed.

The machined dimension displacement estimator 15 then acquires the amount of wear of the wire electrode 2 (step S22). The machining dimension displacement estimator 15 acquires the wear amount of the wire electrode 2 from the wire electrode wear amount measuring device 12. The wire electric discharge machining device 100 holds the consumption amount of the wire electrode 2 that has been converted into data in advance in a memory, and the machining dimension displacement estimator 15 is configured to acquire the consumption amount of the wire electrode 2 from the memory. It is also possible. The pre-datad consumption amount of the wire electrode 2 is associated with, for example, the distance from the reference position on the wire electrode 2 in the wire electrode transport direction 26, the processing direction, and the consumption amount of the wire electrode 2. The data tabulated in the state, that is, the data showing how much the wire electrode 2 is consumed while the wire electrode 2 is transported from the reference position on the wire electrode 2 in the wire electrode transport direction 26 to a certain position. .. In addition to these information, information such as the material of the wire electrode 2, the material of the workpiece 7, the transport speed of the wire electrode 2, the thickness of the wire electrode 2 and the like may be tabulated and stored in the memory.

After executing steps S21 and S22, the machining dimensional displacement estimator 15 calculates the machining dimensional error based on the information regarding the rotation of the wire electrode 2 and the consumption amount of the wire electrode 2 (step S23). The method by which the machining dimensional displacement estimator 15 calculates the machining dimensional error will be described later.

The machining dimension displacement estimator 15 outputs the calculated machining dimension error to the control device 14. The control device 14 corrects the wire electrode position command based on the machining dimensional error received from the machining dimensional displacement estimator 15, and moves the wire electrode 2 (steps S12 and S13). That is, as shown in FIG. 14, in the control device 14, the wire electrode position command extraction unit 141A of the NC device 141 extracts the wire electrode position command from the machining program, and the command correction unit 141B issues the wire electrode position command. Correct based on the machining dimensional error. The drive unit 142 drives the upper guide 6 and the lower guide 11 of the processing unit 16 to move the wire electrode 2 in accordance with the corrected wire electrode position command.

FIG. 15 is a third flowchart showing an example of an operation in which the wire electric discharge machining apparatus 100 according to the first embodiment corrects a machining dimensional error in wire electric discharge machining. FIG. 16 is a diagram showing a connection relationship of each component that realizes an operation example shown in FIG. 15 of the wire electric discharge machining apparatus according to the first embodiment.

The third flowchart shown in FIG. 15 is such that steps S31 to S33 are executed instead of step S11 of the first flowchart shown in FIG. Since steps S12 and S13 in FIG. 15 are the same as steps S12 and S13 in FIG. 12, detailed description thereof will be omitted.

First, the machining dimension displacement estimator 15 of the wire electric discharge machining device 100 acquires the plate thickness information of the workpiece 7 (step S31). The machining dimension displacement estimator 15 acquires plate thickness information from the plate thickness detector 4. The wire electric discharge machine 100 may receive input of the plate thickness information of the workpiece 7 from the user in advance and hold the input plate thickness information in the memory. That is, the machining dimension displacement estimator 15 may acquire the plate thickness information of the workpiece 7 that has been input by the user from the memory.

Next, the machining dimension displacement estimator 15 calculates the rotation angle of the wire electrode 2 and the amount of wear of the wire electrode 2 based on the plate thickness information (step S32). The transport speed of the wire electrode 2 during machining of the workpiece 7 is constant, and similarly, the rotation speed of the wire electrode 2 is also constant. Therefore, the machining dimension displacement estimator 15 holds, for example, the transport speed and the rotation speed of the wire electrode 2 in advance, and the rotation angle of the wire electrode 2 is based on these information and the plate thickness information of the workpiece 7. That is, the angle of rotation is calculated while the workpiece 7 is being machined. Further, the consumption of the wire electrode 2 at an arbitrary position is proportional to the number of discharges, and the number of discharges during machining depends on the plate thickness of the workpiece 7. Therefore, the machining dimension displacement estimator 15 holds, for example, the transport speed of the wire electrode 2 and the number of discharges per hour in advance, and calculates the consumption amount of the wire electrode 2 based on these information and the plate thickness information. do.

Next, the machining dimensional displacement estimator 15 calculates the machining dimensional error based on the rotation angle of the wire electrode 2 and the consumption amount of the wire electrode 2 (step S33).

A method of calculating the machining dimensional error based on the rotation angle of the wire electrode 2 and the consumption amount of the wire electrode 2 by the machining dimensional displacement estimator 15 will be described with reference to FIGS. 17 and 18. Here, a method for calculating the machining dimensional error will be described when the machining dimensional error is set to the machining surface position displacement amounts 34 and 35 shown in FIG. 6 and the machining groove center displacement amount 36. Further, the explanation of the machining dimensional error is divided into two stages, specifically, a method of calculating the consumption amount of the wire electrode 2, a method of calculating the correction amount for correcting the displacement of the machined surface position, and a wire electrode center. The method of calculating the correction amount for correcting the locus will be described separately. FIG. 17 is a diagram showing an example of the relationship between the processing direction of wire electric discharge machining and the amount of wear of the wire electrode 2. FIG. 18 is a diagram showing an example of the relationship between the processing direction of wire electric discharge machining, the rotation direction of the wire electrode 2, and the amount of wear of the wire electrode 2.

As shown in FIG. 17, the wear of the wire electrode 2 mainly occurs in the machining direction, and the amount of wear in the portion where the angle formed with the machining direction is θ is calculated by the following equation (1).
(Consumed amount of wire electrode in the portion where the angle formed by the machining direction is θ) = f (θ)… (1)

In equation (1), f (θ) is a consumption function whose variable is the angle θ formed with the machining direction. This consumption function f (θ) is obtained by simulation or experiment.

Considering the rotation of the wire electrode 2 in the equation (1), the correction amount of the left and right machining surface positions in the machining direction and the correction amount of the wire electrode center locus are as shown in the equations (2), (3), and (4), respectively. Is calculated to.
(Correction amount of the machining surface position on the left side in the machining direction) = f (-π / 2) ... (2)
(Correction amount of machining surface position on the right side of machining direction) = f (π / 2) ... (3)
(Correction amount of wire electrode center locus)
= {F (-π / 2) -f (π / 2)} / 2 ... (4)

As described above, the wire discharge processing apparatus 100 according to the present embodiment detects the processing dimensional error based on the rotation direction, rotation speed or rotation angle of the wire electrode 2 during transportation and the consumption amount of the wire electrode 2. However, the machining dimensional error is compensated by using different correction amounts depending on whether the workpiece to be left is on the right side or the left side with respect to the machining direction. Thereby, the processing accuracy can be improved.

Embodiment 2.
Next, the wire electric discharge machining apparatus according to the second embodiment will be described. The configuration of the wire electric discharge machine according to the second embodiment is the same as that of the wire electric discharge machine 100 according to the first embodiment (see FIG. 1). In the present embodiment, the description of the portion overlapping with the wire electric discharge machine 100 according to the first embodiment will be omitted. Further, in the following description, for convenience, the wire electric discharge machining apparatus according to the second embodiment will be referred to as a wire electric discharge machining apparatus 100a.

In the wire electric discharge machining apparatus 100a according to the second embodiment, the correction amount for compensating for the machining dimensional error described in the first embodiment is determined by using machine learning. FIG. 19 is a diagram showing a configuration example of a control device 14a included in the wire electric discharge machining device 100a according to the second embodiment. The control device 14a is a device in which the machine learning device 50 is added to the control device 14 included in the wire electric discharge machining device 100. In the example shown in FIG. 19, the machine learning device 50 is included in the control device 14a, but the machine learning device 50 exists outside the control device 14a, that is, the machine learning device 50 and the control device 14a. May exist as a separate device.

In the control device 14a, the machine learning device 50 machine-learns the correction amount for compensating for the machining dimensional error described in the first embodiment. The NC device 141 determines a correction amount for compensating for the machining dimensional error based on the learning result of the machine learning device 50, and compensates for the machining dimensional error, that is, corrects the wire electrode position command with the determined correction amount. ..

As shown in FIG. 19, the machine learning device 50 includes a state observation unit 51 and a learning unit 52.

The state observing unit 51 states the rotation information including the rotation direction, rotation angle, or rotation speed of the wire electrode 2, the consumption amount of the wire electrode 2, the machining dimensional error, and the correction amount for compensating for the machining dimensional error. Observe as a variable.

The learning unit 52 machine-learns the correction amount (correction amount for compensating the machining dimensional error) used for compensating the machining dimensional error according to the data set created based on the state variable observed by the state observing section 51. Any learning algorithm may be used as the learning algorithm used by the learning unit 52. As an example, the case where reinforcement learning is applied will be described. Reinforcement learning is that an agent (behavior) in a certain environment observes the current state and decides the action to be taken. Agents get rewarded from the environment by choosing an action and learn how to get the most reward through a series of actions. Q-learning and TD-learning are known as typical methods of reinforcement learning. For example, in the case of Q-learning, the general update equation (behavior value table) of the action value function Q (s, a) is expressed by the following equation (5).

In the formula (5), s _t represents the environment at time t, a _t represents the behavior in time t. By the action a _t, the environment is changed to s _{t + 1.} rt _{+ 1} represents the reward received by the change of the environment, γ represents the discount rate, and α represents the learning coefficient. Note that γ is in the range of 0 <γ ≦ 1 and α is in the range of 0 <α ≦ 1. When applying the Q-learning, it calculates a correction amount to be used for compensation of the processing dimension error becomes action a _t.

In the update formula represented by the equation (5), if _{the action value of the best action a t + 1} at time t + 1 is larger than the action value Q of _{the action a t} executed at time t, the action value Q is increased. However, in the opposite case, the action value Q is reduced. In other words, the action value Q action a _t at time t, so as to approach the best action value at time t + 1, and updates the action value function Q (s, a). As a result, the best behavioral value in a certain environment is sequentially propagated to the behavioral value in the previous environment.

The learning unit 52 includes a reward calculation unit 521 and a function update unit 522 for machine learning the correction amount used for compensating for the machining dimensional error.

The reward calculation unit 521 calculates the reward based on the state variables observed by the state observation unit 51. The reward calculation unit 521 calculates the reward r based on the machining dimensional error. For example, when the machining dimensional error is smaller than a predetermined reference value (for example, 1 μm), the reward r is increased (for example, a reward of “1” is given). On the other hand, when the machining dimensional error is larger than the reference value, the reward r is reduced (for example, a reward of "-1" is given). The reference value of the machining dimensional error may be appropriately changed by the user according to the desired machining dimensional error.

The function update unit 522 updates the function for determining the correction amount used for compensating the machining dimensional error according to the reward calculated by the reward calculation unit 521. For example, in the case of Q-learning, it is used as a function for calculating a correction amount used action value represented by the formula (5) function Q (s _t, a _t) to compensate for the processing dimension error.

In the present embodiment, the case where reinforcement learning is applied to the learning algorithm used by the learning unit 52 has been described, but the present invention is not limited to this. As for the learning algorithm, in addition to reinforcement learning, supervised learning, unsupervised learning, semi-supervised learning, and the like can also be applied.

Further, as the learning algorithm described above, deep learning, which learns the extraction of the feature amount itself, can also be used, and other known methods such as neural networks, genetic programming, functional logic programming, and support vectors can be used. Machine learning may be executed according to the machine or the like.

The machine learning device 50 is used to learn the correction amount used for compensating for the machining dimensional error of the wire electric discharge machine 100a. For example, the machine learning device 50 is connected to the wire electric discharge machine 100a via a network and the wire electric discharge machining is performed. It may be a device separate from the device 100a. Further, the machine learning device 50 may be built in the wire electric discharge machining device 100a as shown in FIG. Further, the machine learning device 50 may exist on the cloud server.

Further, the machine learning device 50 may learn the correction amount used for compensating for the machining dimensional error according to the data set created for the plurality of wire electric discharge machining devices 100a. The machine learning device 50 may acquire a data set from a plurality of wire electric discharge machines 100a used at the same site, or a plurality of wire electric discharge machines 100a operating independently at different sites. The amount of correction used to compensate for machining dimensional errors may be learned using the data set collected from. Further, the wire electric discharge machine 100a for collecting the data set can be added to the target on the way, or conversely, can be removed from the target. Further, a machine learning device 50 that has learned the correction amount used for compensating the machining dimensional error for a certain wire electric discharge machining device 100a is attached to another wire electric discharge machining device 100a, and machining is performed for the other wire electric discharge machining device 100a. The correction amount used for compensating for the dimensional error may be relearned and updated.

As described above, the wire discharge processing apparatus 100a according to the present embodiment includes rotation information including the rotation direction, rotation angle or rotation speed of the wire electrode 2, consumption amount of the wire electrode 2, processing dimension error, and processing dimension. A machine learning device 50 for observing a correction amount for compensating for an error and learning a correction amount for compensating for a machining dimensional error is provided, and the machining dimensional error is compensated based on the learning result by the machine learning device 50. .. As a result, it is possible to improve the processing accuracy as in the first embodiment.

In the present embodiment, the case where the machine learning device 50 is added to the wire discharge processing device 100 shown in FIG. 1 to use machine learning has been described, but the machine learning device 50 is added to the wire discharge processing device 101 shown in FIG. It is also possible to add.

The configuration shown in the above embodiment is an example of the content, can be combined with another known technique, and a part of the configuration is omitted or changed without departing from the gist. It is also possible.

1 wire electrode bobbin, 2 wire electrode, 3 wire electrode transfer roller, 4 plate thickness detector, 5 wire electrode rotation detector, 6 upper guide, 7 workpiece, 8 table, 9 recovery roller, 10 wire electrode recovery box, 11 Lower guide, 12 Wire electrode wear measuring instrument, 13 Lower roller, 14, 14a, 18 Control device, 15 Machining dimension displacement estimator, 16 Machining part, 19 Machining surface, 20 Machining groove center, 21 Machining groove, 22 Machining direction , 23 Upper part of workpiece, 24 Rotation direction, 25 Lower part of workpiece, 26 Wire electrode transfer direction, 27 Wire electrode consumable surface, 28 Wire electrode non-consumable surface, 29 Actual machined groove center, 30 Machine command wire electrode center, 31 Discharge gap, 32 Work piece on the left side in the machining direction, 33 Work piece on the right side in the machining direction, 34, 35 Machining surface position displacement amount, 36 Machining groove center displacement amount, 37 Actual machining surface position, 38 Ideal machining surface position, 39 Wire electrode shape when there is no wear, 50 Machine learning device, 51 State observation unit, 52 Learning unit, 100, 100a, 101 Wire discharge processing device, 141 NC device, 141A Wire electrode position command extraction unit, 141B Command correction Department, 142 drive unit, 521 reward calculation unit, 522 function update unit.

Claims

A wire electrode that discharges from the workpiece while rotating in the twisting direction to process the workpiece, and
An upper guide that supports the wire electrode above the workpiece and
A lower guide that supports the wire electrode below the workpiece,
Based on the positional relationship between the machining direction and the workpiece to be left, and the machining program, the upper guide and the above-mentioned upper guide and the said By correcting the position of the lower guide with a value determined individually, at least one of the offset amount of the wire electrode with respect to the machined surface of the work and the inclination of the wire electrode is adjusted, and the work is machined. A control device that compensates for the machining dimension error, which is the error between the target shape dimension of the workpiece and the actual shape dimension.
A wire electric discharge machine characterized by being provided with.
The control device corrects the position command for the wire electrode described in the machining program based on the machining dimensional error and the positional relationship, and moves the upper guide and the lower guide according to the corrected position command. Let,
The wire electric discharge machining apparatus according to claim 1.
The machining dimensional error is defined as the amount of displacement of the center of the machining groove, which is the amount of deviation between the locus of the center of the machining groove formed in the workpiece and the locus of the center of the wire electrode represented by the machining program.
The wire electric discharge machining apparatus according to claim 1 or 2.
The machining dimensional error is the amount of displacement of the machining surface position, which is the amount of deviation between the actual machining surface of the workpiece obtained by machining the workpiece and the target machining surface of the workpiece represented by the machining program. ,
The wire electric discharge machining apparatus according to claim 1 or 2.
The control device relates to a machining direction and a workpiece to be retained, which is estimated based on rotation information which is information on the rotation of the wire electrode during transportation and the amount of wear of the wire electrode during machining of the workpiece. The positions of the upper guide and the lower guide are controlled based on the positional relationship and the machining program to compensate for the machining dimensional error.
The wire electric discharge machining apparatus according to any one of claims 1 to 4.
A wire electrode rotation detector that detects the rotation direction, rotation speed, or rotation angle of the wire electrode during transportation and generates the rotation information indicating the detection result.
A wire electrode consumption measuring instrument that measures the consumption, and
A machining dimensional displacement estimator that estimates the machining dimensional error based on the rotation information and the consumption amount, and
The wire electric discharge machining apparatus according to claim 5, further comprising.
A plate thickness detector that detects the plate thickness in the vertical direction of the workpiece and generates plate thickness information indicating the detection result, and
A machining dimensional displacement estimator that estimates the machining dimensional error based on the plate thickness information,
The wire electric discharge machining apparatus according to any one of claims 1 to 5, wherein the wire electric discharge machining apparatus is provided.
The machined dimensional displacement estimator exists inside the control device.
The wire electric discharge machining apparatus according to claim 6 or 7.
The rotation information, which is information on the rotation of the wire electrode during transportation, the amount of wear of the wire electrode during machining of the workpiece, the machining dimensional error, and the correction amount used to compensate for the machining dimensional error are observed. A machine learning device that performs machine learning using the state variables obtained from the above observations.
With
The control device compensates for the machining dimensional error based on the learning result by the machine learning device and the machining program.
The wire electric discharge machining apparatus according to any one of claims 1 to 8.
It is a machine learning device that machine-learns the correction amount to compensate for the machining dimensional error, which is the error between the dimension of the target shape of the workpiece obtained by machining the workpiece by the wire discharge machining device and the dimension of the actual shape. hand,
A state observing unit that observes rotation information, which is information on the rotation of the wire electrode during transportation, the amount of wear of the wire electrode during processing of the workpiece, the processing dimensional error, and the correction amount as state variables. ,
A learning unit that machine-learns the correction amount according to the state variable,
A machine learning device characterized by being equipped with.
The learning unit
A reward calculation unit that calculates rewards based on the state variables,
A function update unit that updates a function for determining the correction amount based on the reward,
10. The machine learning apparatus according to claim 10.
The reward calculation unit
The reward is increased when the machining dimensional error is smaller than a predetermined reference value, and the reward is decreased when the machining dimensional error is larger than the reference value.
The machine learning device according to claim 11.