Control device for wire electrical discharge machining machine and control method for wire electrical discharge machining machine

Abstract

A control device for a wire electrical discharge machining machine, which controls, in the wire electrical discharge machining machine, the driving trajectory of an electrode relative to a workpiece, the machining speed between the workpiece and the electrode, and the discharge frequency of the voltage applied between the electrode and the workpiece, and has a storage device and an arithmetic device. The storage device sets n as an integer greater than or equal to 2, and stores at least one piece of data selected from the inter-electrode average voltage, the discharge frequency, the machining speed, and the driving trajectory, which are the voltage applied between the workpiece and the electrode in the (n - 1)-th machining in the case of machining the machining section n times. The arithmetic device calculates the inter-electrode distance in the n-th machining using a calculation model representing the relationship between the data in the (n - 1)-th machining and the machining shape of the workpiece, and calculates at least one command value selected from the machining speed, the discharge frequency, and the driving trajectory in the n-th machining according to the inter-electrode distance in the n-th machining, corresponding to the machining amount required for the desired shape of the machining program for machining the machining section.

Description

Technical Field

[0001] The present invention relates to a control device and a control method for a wire electrical discharge machining machine that controls a wire electrical discharge machining machine that machines a workpiece by applying a voltage between the workpiece and an electrode to generate a discharge. Background Art

[0002] An electrical discharge machining device is a device that machines a workpiece by causing an arc discharge between the machining electrode and the workpiece. In an electrical discharge machining device, a power source for causing a discharge between the electrodes is required. If a high voltage is applied between the electrodes, or the distance between the electrodes is shortened to increase the electric field strength, a discharge occurs due to dielectric breakdown, and the workpiece is removed by machining. After the discharge ends and the insulation recovers, when discharging again, the distance between the electrodes increases, so it is necessary to apply a high voltage between the electrodes or narrow the distance between the electrodes to increase the electric field strength. In addition, in machining using an electrical discharge machining device, machining is repeated multiple times while changing the machining conditions according to the accuracy of the target dimensions and surface roughness. That is, first, a rough machining process of machining the target shape from the workpiece is performed, and then, the shape accuracy is improved to match the target shape, and a shape correction machining process of continuously making the surface roughness finer is performed.

[0003] In shape correction machining, it is required to correct the shape deviation generated in the machining up to the number of machining times including rough machining in the previous machining while improving the accuracy of the surface roughness. In addition, in which direction and to what extent the shape has shifted in the machining up to the number of machining times in the previous machining with respect to the machining progress direction varies depending on the shape of the machining sample, the machining progress direction, the machining conditions, and the like.

[0004] In shape correction machining, the ability to machine according to the target dimensions is required even if the machining amount to be corrected changes due to the machining location and the direction with respect to the machining progress direction. If the shape cannot be corrected according to the target dimensions in shape correction machining, the possibility of changes in the distance between the electrodes, surface roughness, and machining dimensions due to the machining location increases. Therefore, Patent Document 1 discloses the following technique: the average machining voltage between the electrodes is monitored to detect the machining state including the discharge frequency and machining amount in electrical discharge machining, and the relative movement speed between the machining electrode and the workpiece is controlled so that the average machining voltage between the electrodes becomes a set voltage.

[0005] Patent Document 1: Japanese Unexamined Patent Application Publication No. 2020-146788 Summary of the Invention

[0006] When estimating the inter-electrode distance based on the voltage or discharge frequency, the voltage or discharge frequency fluctuates at a high frequency. If the voltage or discharge frequency is directly monitored for feedback control, the fluctuations of the controlled object will become excessive. Therefore, as shown in the above prior art, a control system that averages the voltage or discharge frequency is used. However, in the control system as described above, the change in the relative movement speed becomes slow. Therefore, in the case where there is a shape in which the inter-electrode distance changes sharply as a result of the number of machining operations up to the previous machining, shape correction cannot be sufficiently performed due to the delay of the control system, and there is a problem that the shape fluctuations cannot be eliminated.

[0007] The present invention has been made in view of the above circumstances, and an object thereof is to obtain a control device for a wire electrical discharge machining machine that can perform high-precision shape correction machining compared to the past even when there is a sharp shape change as a result of the previous machining.

[0008] In order to solve the above problems and achieve the object, the present invention is a control device for a wire electrical discharge machining machine, which controls the drive locus of the electrode relative to the workpiece, the relative machining speed between the workpiece and the electrode, and the discharge frequency of the voltage periodically applied between the electrode and the workpiece in a wire electrical discharge machining machine that machines the workpiece by applying a voltage between the workpiece and the electrode to generate a discharge, and has a storage device and an arithmetic device. The storage device sets n as an integer greater than or equal to 2, and stores at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive locus in the (n - 1)-th machining when the workpiece is machined n times in a specified machining section of the workpiece. The arithmetic device calculates the inter-electrode distance in the n-th machining using a calculation model representing the relationship between the data in the (n - 1)-th machining and the machining shape of the workpiece, and calculates at least one command value selected from the machining speed, discharge frequency, and drive locus in the n-th machining corresponding to the machining amount required for the desired shape based on the machining program for machining the machining section according to the inter-electrode distance in the n-th machining.

[0009] Effects of the Invention

[0010] The control device for a wire electrical discharge machining machine according to the present invention has the following effect, that is, even when there is a sharp shape change as a result of the previous machining, high-precision shape correction machining can be performed compared to the past. Description of the Drawings

[0011] Figure 1 It is a block diagram showing an example of the structure of a wire electrical discharge machining machine according to Embodiment 1.

[0012] Figure 2 It is a block diagram showing an example of a PI control system.

[0013] Figure 3 It is a diagram showing an example of the relationship between the shape of the machining part and the machining speed in an existing PI control system.

[0014] Figure 4 It is a diagram showing an example of the relationship between the shape of the machining part and the machining speed in the control system of Embodiment 1.

[0015] Figure 5 It is a diagram schematically showing the machining amount of the workpiece at the corner of the curve shape.

[0016] Figure 6 It is a diagram showing an example of the machining shape of the workpiece obtained from the previous machining result.

[0017] Figure 7 It is a diagram showing an example of the machining shape of the workpiece obtained from the previous machining result.

[0018] Figure 8 It is a flowchart showing an example of the sequence of the control method of the wire electrical discharge machining machine according to Embodiment 1.

[0019] Figure 9 It is a flowchart showing an example of the sequence of the control method of the wire electrical discharge machining machine according to Embodiment 1.

[0020] Figure 10 It is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 2.

[0021] Figure 11 It is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 3.

[0022] Figure 12 It is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 5.

[0023] Figure 13 It is a model diagram showing an example of the outline of machine learning according to Embodiment 6.

[0024] Figure 14 It is a diagram schematically showing an example of the structure of the learning device used in the control device of the wire electrical discharge machining machine according to Embodiment 6.

[0025] Figure 15 It is a diagram schematically showing an example of the neural network used by the model generation unit.

[0026] Figure 16It is a flowchart showing an example of the order of learning processing performed by the learning device according to Embodiment 6.

[0027] Figure 17 It is a model diagram showing an example of the outline of machine learning according to Embodiment 6.

[0028] Figure 18 It is a diagram schematically showing an example of the structure of an inference device used by the control device of the wire electrical discharge machining machine according to Embodiment 6.

[0029] Figure 19 It is a flowchart showing an example of the order of inference processing performed by the inference device according to Embodiment 6.

[0030] Figure 20 It is a diagram showing an example of the hardware structure of the control device of the wire electrical discharge machining machine according to Embodiments 1 to 6. Detailed Embodiment

[0031] Hereinafter, the control device of the wire electrical discharge machining machine and the control method of the wire electrical discharge machining machine according to the embodiments of the present invention will be described in detail based on the drawings.

[0032] Embodiment 1.

[0033] Figure 1 It is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 1. The wire electrical discharge machining machine 1 has a machining electrode 10, a power supply unit 20, and a control unit 30. The power supply unit 20 and the control unit 30 correspond to the control device of the wire electrical discharge machining machine 1.

[0034] The wire electrical discharge machining machine 1 is a machining device that repeats machining of a specified machining section of the workpiece 11 n times. n is an integer greater than or equal to 2. At this time, in each machining, the distance and electric energy between the workpiece 11 and the machining electrode 10 as machining conditions are changed. In one example, machining is performed such that the electric energy becomes smaller as the number of machining times increases. Moreover, in the nth machining, shape correction machining for improving surface roughness and shape accuracy is performed. Among the n times of machining, the first to the (n - 1)th machinings are called rough machining, and the nth machining is called shape correction machining.

[0035] The machining electrode 10 is a wire electrode as an electrode made of a linear conductive material. The machining electrode 10 has a structure capable of machining the workpiece 11 with a wire material, but Figure 1Among them, the structure of the machining electrode 10 is shown in a simplified manner. In one example, the machining electrode 10 is fed out from a spool, and its direction is changed by a supply roller so that the machining electrode 10 is arranged in the vertical direction. During the machining electrode 10 passing through the holes in the upper die and the holes in the lower die, electrical discharge machining is performed on the workpiece 11. After the machining electrode 10 passes through the lower die, its direction is changed by a lower roller and is recovered into a recovery box by a recovery roller.

[0036] The power supply unit 20 includes a machining power supply 21 and a machining power supply control unit 22. The machining power supply 21 applies a voltage between the machining electrode 10 and the workpiece 11. The machining power supply control unit 22 controls the on and off of the machining power supply 21. The wire electrical discharge machining machine 1 applies a voltage between the workpiece 11 and the machining electrode 10 to generate an electrical discharge, thereby performing electrical discharge machining on the workpiece 11. Here, the detailed description of the power supply unit 20 including the mechanical structure is omitted because it is different from the gist of the present invention.

[0037] The control unit 30 controls the machining speed according to the average voltage between the machining electrode 10 and the workpiece 11. The machining speed is the relative speed between the machining electrode 10 and the workpiece 11.

[0038] The control unit 30 includes a drive trajectory control unit 31, an average inter-electrode voltage detection unit 32, a voltage calculation unit 33, a machining speed control unit 34, a drive control unit 35, an average inter-electrode voltage storage unit 36, a discharge frequency storage unit 37, a machining speed storage unit 38, and a drive trajectory storage unit 39.

[0039] The drive trajectory control unit 31 controls the movement of the axes of the wire electrical discharge machining machine 1 according to the machining program. That is, the drive trajectory control unit 31 calculates the command value of the drive trajectory of the machining electrode 10 relative to the workpiece 11. In one example, the drive trajectory control unit 31 discriminates whether the machining shape of the machining section to be machined is a straight shape or a curved shape according to the machining program, and calculates the drive trajectory as the path of the machining electrode 10. In the case of a curved shape, the drive trajectory is calculated using the corner diameter and the opening angle of the corner obtained by considering the radius of the corner forming the curved shape, the diameter of the machining electrode 10, and the offset. The drive trajectory control unit 31 outputs the calculated command value of the drive trajectory to the drive control unit 35. In addition, the drive trajectory control unit 31 stores the drive trajectory in the drive trajectory storage unit 39.

[0040] The inter-electrode average voltage detection unit 32 detects the average value of the voltage between the machining electrode 10 and the workpiece 11 during a specified time, i.e., the inter-electrode average voltage. The inter-electrode average voltage detection unit 32 stores the detected inter-electrode average voltage in the inter-electrode average voltage storage unit 36. In addition, the inter-electrode average voltage detection unit 32 estimates the discharge conditions such as the discharge frequency and the machining amount based on the inter-electrode average voltage for each specified time, and stores the estimated discharge frequency in the discharge frequency storage unit 37.

[0041] The voltage operation unit 33 operates the difference between the detected inter-electrode average voltage and the set voltage. In the machining using the wire electrical discharge machining machine 1, it is considered that the appropriate discharge conditions such as the discharge frequency and the machining amount vary according to the machining conditions, and the appropriate inter-electrode average voltage is determined according to the purpose. Therefore, the set voltage is preset as an appropriate voltage according to the purpose.

[0042] The machining speed control unit 34 includes: a machining speed calculation unit 341 that calculates the machining speed; and an arithmetic device 342 that calculates a correction value for correcting the machining speed calculated by the machining speed calculation unit 341. The machining speed calculation unit 341 calculates the machining speed such that the inter-electrode average voltage during the measured specified time becomes the set voltage, i.e., the difference calculated by the voltage operation unit 33 becomes 0. The machining speed can be calculated by the machining speed calculation unit 341 using a known method. In one example, the machining speed calculation unit 341 can operate the command value of the machining speed using at least one of the inter-electrode average voltage, the discharge frequency, and the drive locus in the current machining as the nth machining. In addition, the machining speed calculation unit 341 corresponds to the command value calculation unit.

[0043] When the machining section being machined is linear, the machining speed calculation unit 341 calculates the machining speed by PI (Proportional-Integral) control that makes the difference between the inter-electrode average voltage and the set voltage 0. In addition, when the machining section is curved, the machining speed calculation unit 341 calculates the machining speed based on the inter-electrode average voltage and the set voltage, considering the angular diameter and the opening angle of the corner forming the curved shape at the drive position.

[0044] In addition, the machining speed calculation unit 341 corrects the calculated machining speed using the correction value calculated by the arithmetic device 342 in the shape correction machining for improving the surface roughness and the shape accuracy. Hereinafter, the machining speed corrected by the correction value is referred to as the corrected machining speed when distinguished from the uncorrected machining speed. In addition, the uncorrected machining speed and the corrected machining speed are referred to as the machining speed when not distinguished from each other.

[0045] In the case of shape correction machining for improving surface roughness and shape accuracy, the arithmetic unit 342 calculates the inter-electrode distance in the current machining using a calculation model that represents the relationship between at least one piece of data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining and the machining shape of the workpiece 11. In addition, the arithmetic unit 342 calculates a command value for the machining speed in the current machining corresponding to the machining amount required to make the machining shape of the workpiece 11 in the current machining the desired shape based on the machining program for machining the machining section. When machining is performed n times on the machining section, the previous machining corresponds to the (n - 1)-th machining, and the current machining corresponds to the n-th machining. At least one piece of data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining is data representing an overview in one example. The calculation model represents the relationship between the above data and the machining shape of the workpiece 11, but may also represent the relationship between the above data and the inter-electrode distance in the current machining.

[0046] The machining speed calculated by the machining speed calculation unit 341 is used to make the calculated machining shape of the workpiece 11 obtained when the (n - 1)-th machining is performed based on the machining program the desired shape. The calculated machining shape of the workpiece 11 often does not correspond to the actual machining shape of the workpiece 11. The command value for the machining speed calculated by the arithmetic unit 342 is a command value for eliminating the difference between the actual machining shape and the desired machining shape of the workpiece 11 obtained as a result of the previous machining, and is also a correction value for correcting the machining speed calculated by the machining speed calculation unit 341. Therefore, the command value for the machining speed calculated by the arithmetic unit 342 is also referred to as a correction value. The actual machining shape of the workpiece 11 obtained as a result of the previous machining is also a machining shape estimated through strict calculation. However, since the estimation is performed using at least one piece of data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining, the machining shape estimated using the above data is closer to the actual machining shape than the calculated machining shape of the workpiece 11 obtained when the (n - 1)-th machining is performed based on the machining program.

[0047] When estimating the inter-electrode distance, a calculation model is first obtained that describes the relationship between at least one of the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory selection during the previous machining and the machining shape of the workpiece 11. Then, by using a function that includes this calculation model, the inter-electrode distance during the current machining is calculated, and the command value for the machining speed to achieve the desired shape is also calculated. In addition, the accuracy of the estimated inter-electrode distance improves as the number of data items related to the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory selection during the previous machining increases. Therefore, it is preferable to use as many types of data as possible, such as the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory during the previous machining. The machining speed control unit 34 outputs the calculated command value for the machining speed to the drive control unit 35.

[0048] Based on the command value for the drive trajectory from the drive trajectory control unit 31 and the command value for the machining speed from the machining speed control unit 34, the drive control unit 35 controls the movement of the machining electrode 10 relative to the workpiece 11. That is, based on the command value for the drive trajectory, the position of the machining electrode 10 relative to the workpiece 11 is controlled, and based on the command value for the machining speed, the relative speed of the machining electrode 10 relative to the workpiece 11 is controlled. The drive control unit 35 can move the workpiece 11 based on the command value, move the machining electrode 10 based on the command value, or move both the workpiece 11 and the machining electrode 10 based on the command value.

[0049] The average inter-electrode voltage storage unit 36 stores the value of the average inter-electrode voltage detected by the average inter-electrode voltage detection unit 32 at each specified time. As described above, in the on-line electric discharge machining machine 1, the machining area is machined continuously n times. Moreover, in the last nth machining, shape correction machining is performed to improve the shape accuracy to match the desired shape and make the surface roughness finer. That is, machining is performed repeatedly while changing the machining conditions. In Embodiment 1, the average inter-electrode voltage storage unit 36 only needs to be able to store the average inter-electrode voltage during the machining immediately before the shape correction machining, that is, the (n - 1)th machining. In one example, the average inter-electrode voltage storage unit 36 stores the value of the average inter-electrode voltage in association with the coordinate values of the drive trajectory during each machining. In addition, the average inter-electrode voltage storage unit 36 can store the time from voltage application to discharge occurrence, that is, the no-load time, but it is described as the average inter-electrode voltage here. The average inter-electrode voltage can be measured by the machining power supply control unit 22.

[0050] The discharge frequency storage unit 37 stores the value of the discharge frequency estimated by the inter-electrode average voltage detection unit 32. The discharge frequency storage unit 37 only needs to be able to store at least the discharge frequency during the processing immediately before the shape correction processing, that is, the (n - 1)-th processing. In one example, the discharge frequency storage unit 37 stores the discharge frequency in association with the coordinate values of the drive trajectory during each processing. The discharge frequency can be measured by the processing power supply control unit 22.

[0051] The machining speed storage unit 38 stores the command value of the machining speed calculated by the machining speed control unit 34. The machining speed storage unit 38 only needs to be able to store at least the machining speed during the processing immediately before the shape correction processing, that is, the (n - 1)-th processing. In one example, the machining speed storage unit 38 stores the command value of the machining speed in association with the coordinate values of the drive trajectory during each processing.

[0052] The drive trajectory storage unit 39 stores the drive trajectory from the drive trajectory control unit 31. Regarding the drive trajectory, when the machining section is in a straight line shape, the coordinate values based on a specified position are stored, but when the machining section is at a corner, in addition to the coordinate values, the radius of the drive trajectory at the corner, the opening angle, etc. of the corner are also stored. The drive trajectory storage unit 39 only needs to be able to store at least the drive trajectory during the processing immediately before the shape correction processing, that is, the (n - 1)-th processing.

[0053] The inter-electrode average voltage storage unit 36, the discharge frequency storage unit 37, the machining speed storage unit 38, and the drive trajectory storage unit 39 correspond to the storage device.

[0054] Here, the control of the machining speed in the straight line shape and the control of the machining speed in the curved shape during the shape correction processing will be described.

[0055] First, the control of the machining speed in the straight line shape will be described. In the wire electrical discharge machining machine 1, machining is performed by continuously generating discharges, so the power supply unit 20 applies a voltage between the electrodes. The inter-electrode average voltage detection unit 32 measures the inter-electrode average voltage at each specified time. In addition, the inter-electrode average voltage detection unit 32 can estimate the discharge conditions such as the discharge frequency and the machining amount based on the inter-electrode average voltage at each specified time. Depending on the machining conditions, the appropriate discharge frequency, machining amount, etc. will be different, and the appropriate inter-electrode average voltage is determined according to the purpose. That is, the set voltage is a voltage that is considered appropriate and is preset as the set voltage corresponding to the purpose. The voltage operation unit 33 obtains the inter-electrode average voltage measured by the inter-electrode average voltage detection unit 32 and calculates the difference between the inter-electrode average voltage and the set voltage. The voltage operation unit 33 outputs the calculated difference to the machining speed control unit 34.

[0056] If the machining speed calculation unit 341 of the machining speed control unit 34 obtains a difference from the voltage calculation unit 33, it calculates the machining speed so that the difference calculated by the voltage calculation unit 33 becomes zero. The drive trajectory control unit 31 calculates the command value of the drive trajectory according to the machining program. The drive control unit 35 controls the operation of the machining electrode 10 based on the command value of the machining speed calculated by the machining speed control unit 34 and the command value of the drive trajectory calculated by the drive trajectory control unit 31.

[0057] At this time, if the machining speed calculation unit 341 of the machining speed control unit 34 directly converts the voltage calculated based on the difference between the average inter-electrode voltage with large fluctuations and the set voltage into speed, there will be a sharp speed change, resulting in poor shape accuracy and surface roughness. Therefore, the machining speed calculation unit 341 uses Figure 2 the PI control system shown to obtain the machining speed. Figure 2 is a block diagram showing an example of the PI control system. As Figure 2 shown, in the PI control system, proportional control is performed by multiplying the difference between the average inter-electrode voltage and the set voltage by a specified proportional gain, integral control is performed by multiplying by the integral gain corresponding to the accumulated amount of the deviation, and the machining speed is determined by adding these results.

[0058] Figure 3 is a diagram showing an example of the relationship between the shape of the machining part and the machining speed in the existing PI control system. In Figure 3 is shown the case where there are small unevennesses in the shape of the machining part, which is the result shape of the previous machining. In Figure 2 the PI control system shown, if the gain of the integral term is small, even if there is a frequency component, the integral term reacts as a whole, so the integral term is insensitive and the response is slow. Therefore, when machining the workpiece 11 with the machining part shape shown in Figure 3 , it takes time until the machining speed calculated based on the average inter-electrode voltage reaches the speed capable of correcting the unevenness, and the shape cannot be corrected sufficiently. That is, due to the delay of the speed control system, the shape of the machining part cannot be corrected sufficiently. In the example of Figure 3 , even after passing through the vertex of the machining part shape, the machining speed will decrease. In addition, after passing through the vertex of the machining part shape, the machining speed will increase excessively, resulting in vibration. As described above, even if shape correction machining is performed only by PI control, the desired shape cannot be achieved.

[0059] Therefore, in the control system of Embodiment 1, the position in the previous machining performed before the shape correction machining is stored in the drive locus storage unit 39, the average interelectrode voltage is stored in the average interelectrode voltage storage unit 36, the discharge frequency is stored in the discharge frequency storage unit 37, and the machining speed is stored in the machining speed storage unit 38. Moreover, the arithmetic unit 342 of the machining speed control unit 34 uses a calculation model to predict in advance the interelectrode distance during the current machining of this position based on the general situation of the average interelectrode voltage, discharge frequency, and machining speed during the previous machining of this position, and calculates a correction value for the machining speed for setting the desired machining shape based on the predicted interelectrode distance. As described above, the calculation model includes a function for calculating the interelectrode distance between the machining shape of the workpiece 11 during the previous machining and the machining electrode 10 during the current machining based on the general situation of the average interelectrode voltage, discharge frequency, and machining speed during the previous machining. In addition, the calculation model includes the following function, that is, using the predicted interelectrode distance and the distance between the desired machining shape obtained based on the machining program and the machining electrode 10 during the current machining, the amount of the workpiece 11 to be removed during the current machining is determined, and the machining speed of the machining electrode 10 for removing this amount of the workpiece 11 is calculated.

[0060] Figure 4 FIG. is an example showing the relationship between the machining part shape and the machining speed in the control system of Embodiment 1. In Embodiment 1, the interelectrode distance is calculated. However, in Figure 4 For the sake of easy understanding of the description, the machining part shape obtained when calculating the interelectrode distance is shown. In Embodiment 1, it is convex in advance in the machining section, but is predicted in the manner shown by the Figure 4 machining part shape. Therefore, the arithmetic unit 342 calculates the correction value of the machining speed in such a way that the command value of the machining speed decreases from the proximal end of the convex part, so as to reduce the corrected machining speed calculated by the machining speed calculation unit 341. In addition, the arithmetic unit 342 calculates the correction value of the machining speed in such a way that the machining speed does not vibrate after passing through the convex section. Thereby, it is possible to correct the unevenness on the machining surface of the workpiece 11 without reducing Figure 2 the integral gain of the PI control system or increasing the proportional gain. In addition, in Figure 4 , the machining speed also decreases after passing through the vertex of the convex part. This is related to the case where machining is performed not only by discharging in front of the machining electrode 10 but also by discharging backward. That is, the machining amounts before and after the machining electrode 10 passing through the vertex of the convex part are larger than those when the machining electrode 10 is at the vertex of the convex part, so the machining speed decreases. Figure 4 The point to be noted in Figure 3Compared with the case of , the decrease in the machining speed is advanced, and after the shape of the machining part passes through the convex part and returns to a straight line, the speed does not increase excessively, that is, there is no decaying vibration.

[0061] If the proportional gain is set to K1, the integral gain is set to K2, and the set voltage is set to V s , the average voltage between electrodes is set to V ave , and the correction value of the machining speed calculated based on the shape of the machining part in the previous machining is set to F comp , then the arithmetic expression of the PI control system for calculating the corrected machining speed F is represented by the following formula (1). The correction value F of the machining speed comp is calculated using the overview of the previous machining stored in the average voltage storage unit 36 between electrodes, the discharge frequency storage unit 37, and the machining speed storage unit 38. Specifically, if the average voltage between electrodes in the previous machining stored in the average voltage storage unit 36 between electrodes is set to V n-1 , the discharge frequency in the previous machining stored in the discharge frequency storage unit 37 is set to S n-1 , and the machining speed in the previous machining stored in the machining speed storage unit 38 is set to F n-1 , then the correction value F of the machining speed comp is represented by the following formula (2).

[0062] F = K1(V ave - V s ) + ∫K2(V ave - V s )dt + F comp ··· (1)

[0063] F comp = ∫f(V n-1 , F n-1 , S n-1 )dt··· (2)

[0064] As described above, formula (2) includes calculating the overview of the average voltage V n-1 , the discharge frequency S n-1 , and the machining speed F n-1 in the previous machining, and calculating the interelectrode distance in the current machining according to the overview of the average voltage V n-1 , the discharge frequency S n-1 , and the machining speed F n-1 . It is a calculation model that is a function for calculating the machining speed according to the interelectrode distance.

[0065] Next, the control of the machining speed in the curve shape will be described. Figure 5It is a diagram schematically showing the machining amount of the workpiece in the corner of a curved shape. Here, the case of machining the inner corner, i.e., the inner angle, of the corner is shown. It is known that starting from the center of curvature of the corner, the machining amount at the inner angle when the machining electrode 10 rotates by θ at a position on the circumference with an angular diameter R from the center of curvature is represented by the following formula (3). Here, R is the angular diameter obtained by considering the diameter and offset of the machining electrode 10, and θ is the opening angle of the corner. In addition, the distance between the center of the machining electrode 10 and the surface of the machined workpiece 11 after machining is set as GAP.

[0066] Machining amount at the inner angle = R * GAP * θ + GAP 2 *θ ···(3)

[0067] In formula (3), the machining electrode 10 moves by R * θ. When moving by R * θ in linear shape machining, the machining amount is represented by the following formula (4).

[0068] Machining amount in linear shape = R * GAP * θ ···(4)

[0069] According to formula (3) and formula (4), the machining amount at the inner angle is more than that in the linear shape. That is, in the angular shape as Figure 5 shown, compared with the linear shape, the required machining amount, i.e., the area opposite to the machining electrode 10, changes. Therefore, it is necessary to change the average voltage between electrodes, discharge frequency, machining speed, etc. according to the change in the machining amount. In Patent Document 1, a calculation method of machining speed and discharge frequency corresponding to the change in the machining amount in the corner is disclosed. However, as described above, when controlling in a feedforward manner for the change in the machining amount, if the result of the previous machining does not become the expected angular shape, the desired angular shape cannot be obtained through the current shape correction machining.

[0070] Therefore, in the control system of Embodiment 1, in the same way as in the case of the linear shape, the position in the previous machining is stored in the drive locus storage unit 39, the average voltage between electrodes is stored in the average voltage between electrodes storage unit 36, the discharge frequency is stored in the discharge frequency storage unit 37, and the machining speed is stored in the machining speed storage unit 38. In addition, in the drive locus storage unit 39, during the machining of the corner, both the position in the previous machining and the information of the corner are stored in the drive locus storage unit 39. The information of the corner includes the drive locus in the corner, i.e., the radius, opening angle, etc. of the corner. The arithmetic device 342 of the machining speed control unit 34 calculates a correction value based on the overview of the discharge frequency, average voltage between electrodes, and machining speed in the previous machining and according to the information of the corner stored in the drive locus storage unit 39, and this correction value corrects the machining speed at this position of the corner in the current machining.

[0071] Figure 6 and Figure 7 is a diagram showing an example of the processed shape of the workpiece obtained from the result of the previous processing. In these diagrams, the dashed line represents the desired shape, and the solid line represents the shape actually obtained through the previous processing. Figure 6 It shows a case where the actual corner diameter obtained is smaller than the shape of the desired corner. Figure 7 It shows a case where there are irregularities at the entrance and exit portions of the corner. In Embodiment 1, the arithmetic unit 342 calculates, through a calculation model, the inter-electrode distance in the current processing corresponding to the actually obtained shape of Figure 6 and Figure 7 and uses this inter-electrode distance to calculate the processing speed required to remove the region corresponding to the difference between the actually obtained shape and the desired shape of Figure 6 and Figure 7 and sets this processing speed as the correction value. By performing processing at the corrected processing speed corrected by this correction value, it is possible to perform processing so as to obtain the desired shape. As shown in Figure 6 and Figure 7 , in the case where the result of the previous processing is that the shape of the corner is smaller than the desired shape, or in the case where there are irregularities at the entrance and exit portions of the corner, etc., by optimizing the processing speed in accordance with these shapes, it is possible to improve the shape accuracy after processing.

[0072] The corrected processing speed F at the corner cnr is represented by the following formula (5). In addition, let the position where processing is performed at the corner be x. Further, the function g is a processing speed obtained by considering the change in the processing amount at the corner as compared with the case of a linear shape, and can be obtained by a known method. The correction value F of the processing speed comp_cnr is calculated by the following formula (6) based on the outline of the previous processing stored in the inter-electrode average voltage storage unit 36, the discharge frequency storage unit 37, and the processing speed storage unit 38, using the corner diameter, the opening angle, and the position at the corner.

[0073] F cnr = g(R, θ, V s , V ave , x) + F comp_cnr ··· (5)

[0074] F comp_cnr = ∫h(R, θ, V n-1 , F n-1 , S n-1 , x)dt ··· (6)

[0075] In addition, as shown in equations (5) and (6), at the corner, corresponding to the position of the machining electrode 10 in the corner, the machining amount required as compared with the straight portion changes. Therefore, the position in the corner is included in the function. Further, in equation (5), a case is shown where the function g corrects the difference in the machining amounts of the corner and the straight portion by controlling the machining speed. However, the function g may also correct the difference in the machining amounts of the corner and the straight portion by controlling the discharge frequency or the trajectory. Also, in the above description, a case of machining an inner angle is given as an example. However, in the case of machining the outer side of the corner, i.e., the outer angle, the machining speed can also be controlled by the same concept. In the case of machining the outer angle, the machining amount becomes smaller compared to the case of machining a straight shape.

[0076] Next, a control method of the wire electrical discharge machining machine 1 as described above will be described. Figure 8 and Figure 9 is a flowchart showing an example of the sequence of the control method of the wire electrical discharge machining machine according to Embodiment 1. First, the drive trajectory control unit 31 calculates a drive trajectory based on the machining program and stores the drive trajectory in the drive trajectory storage unit 39 (step S11). The drive trajectory control unit 31 outputs the calculated drive trajectory to the drive control unit 35. The drive control unit 35 drives the machining electrode 10 based on the command value of the drive trajectory (step S12). Then, the machining power supply control unit 22 controls the machining power supply 21 to apply a voltage between the electrodes at a prescribed discharge frequency (step S13). Then, the inter-electrode average voltage detection unit 32 measures the inter-electrode average voltage for each prescribed time and stores the inter-electrode average voltage as the measurement result in the inter-electrode average voltage storage unit 36 (step S14). The inter-electrode average voltage detection unit 32 estimates the discharge frequency based on the inter-electrode average voltage for each prescribed time and stores the estimated discharge frequency in the discharge frequency storage unit 37 (step S15).

[0077] Next, it is determined whether the machining is of a straight shape (step S16). Whether the machining is of a straight shape or a curved shape is discriminated by the drive trajectory control unit 31 when determining the drive trajectory. In the case of machining a straight shape (when Yes in step S16), the voltage operation unit 33 calculates the difference between the inter-electrode average voltage obtained from the inter-electrode average voltage detection unit 32 and the set voltage (step S17). The voltage operation unit 33 outputs the calculated difference to the machining speed control unit 34. The machining speed calculation unit 341 of the machining speed control unit 34 calculates the machining speed that makes the difference zero by PI control (step S18).

[0078] On the other hand, in the case of machining that is not a linear shape in step S16, that is, in the case of machining a curved shape (when step S16 is No), the machining speed calculation unit 341 uses the corner diameter R, the opening angle θ of the corner, and the set voltage V s and the average inter - electrode voltage during the current machining and the position x in the corner to calculate the machining speed (step S19). In one example, the machining speed calculation unit 341 calculates the machining speed at which the difference between the average inter - electrode voltage and the set voltage becomes 0 at the position x during the current machining, considering the corner diameter R and the opening angle θ of the corner.

[0079] Then, or after step S18, the machining speed calculation unit 341 determines whether it is shape - correction machining (step S20). In the case of not being shape - correction machining (when step S20 is No), the machining speed calculation unit 341 stores the calculated machining speed in the machining speed storage unit 382 (step S21). In addition, the drive trajectory control unit 31 calculates the drive trajectory and stores the calculated drive trajectory in the drive trajectory storage unit 39 (step S22). Next, the drive control unit 35 drives the machining electrode 10 based on the drive trajectory and the command value of the machining speed (step S23). Then, the process returns to step S13 to perform rough machining that is not shape - correction machining.

[0080] On the other hand, in the case of shape - correction machining in step S20 (when step S20 is Yes), the arithmetic unit 342 of the machining speed control unit 34 uses formula (2) or formula (6) to calculate the inter - electrode distance during the current machining based on at least one data selected from the average inter - electrode voltage, discharge frequency, machining voltage, and drive trajectory during the previous machining, and calculates the correction value of the machining speed for setting the machining shape of the desired workpiece 11 according to the inter - electrode distance (step S24). Next, the machining speed calculation unit 341 uses formula (1) or formula (5) to calculate the corrected machining speed after correcting the calculated machining speed with the correction value (step S25). Then, the drive trajectory control unit 31 calculates the drive trajectory (step S26). Moreover, the drive control unit 35 performs shape - correction machining based on the drive trajectory and the corrected machining speed (step S27), and the process ends.

[0081] In Embodiment 1, n is set as an integer greater than or equal to 2, and at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive locus in the (n - 1)-th machining in the case of machining the machining section n times is stored. Using a calculation model representing the relationship between the data in the (n - 1)-th machining and the machining shape of the workpiece 11, the inter-electrode distance in the n-th machining is calculated, and a correction value of the machining speed in the n-th machining corresponding to the machining amount required to make the desired shape is calculated based on the n-th inter-electrode distance. That is, the machining shape of the workpiece 11 in the previous machining, i.e., the (n - 1)-th machining, can be estimated with high accuracy. Therefore, in shape correction machining, machining can be performed so that the workpiece 11 becomes the desired shape. As a result, even when there is a sharp shape change in the result of the previous machining, high-precision shape correction machining can be performed.

[0082] In particular, at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive locus in the (n - 1)-th machining is data reflecting the state of the (n - 1)-th machining. As described above, by using this data, the machining shape of the workpiece 11 in the (n - 1)-th machining can be estimated. That is, in Embodiment 1, in order to estimate the machining shape of the workpiece 11 in the (n - 1)-th machining with high accuracy, at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive locus in the (n - 1)-th machining is stored.

[0083] Embodiment 2.

[0084] Figure 10 FIG. is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 2. In addition, the same reference numerals are given to the same structural elements as in Embodiment 1, and their descriptions are omitted, and the differences from Embodiment 1 are described.

[0085] In the wire electrical discharge machining machine 1a of Embodiment 2, a machining speed control unit 34a is provided instead of the machining speed control unit 34, and a machining power supply control unit 22a is provided instead of the machining power supply control unit 22.

[0086] The machining speed control unit 34a corresponds to the machining speed calculation unit 341 described in Embodiment 1 and does not have an arithmetic device 342. That is, the machining speed control unit 34a calculates the machining speed so that the difference calculated by the voltage arithmetic unit 33 becomes 0.

[0087] The power supply control unit 22a for machining includes: a discharge frequency control unit 221 that controls the machining power supply 21 at a discharge frequency determined in advance by machining conditions; and an arithmetic unit 222 that calculates a correction value for correcting the discharge frequency determined in advance by machining conditions. The discharge frequency control unit 221 controls the on / off cycle of the machining power supply 21. In addition, the discharge frequency control unit 221 may calculate the command value of the discharge frequency using at least one piece of data selected from the average inter-electrode voltage, discharge frequency, and drive locus in the current machining. In the case of shape correction machining for improving surface roughness and shape accuracy, the discharge frequency control unit 221 corrects the pre-determined discharge frequency with the correction value calculated by the arithmetic unit 222 and controls the machining power supply 21 with the corrected discharge frequency. The discharge frequency control unit 221 corresponds to the command value calculation unit.

[0088] In the case of shape correction machining for improving surface roughness and shape accuracy, the arithmetic unit 222 calculates the inter-electrode distance in the current machining using a calculation model representing the relationship between at least one piece of data selected from the average inter-electrode voltage, discharge frequency, machining speed, and drive locus in the previous machining and the machining shape of the workpiece 11. In addition, the arithmetic unit 222 calculates the command value of the discharge frequency in the current machining corresponding to the machining amount required to set the machining shape of the workpiece 11 in the calculation to the desired shape based on the machining program for machining the machining section. When machining is performed n times for a machining section, the previous machining corresponds to the (n - 1)th machining, and the current machining corresponds to the nth machining. At least one piece of data selected from the average inter-electrode voltage, discharge frequency, machining speed, and drive locus in the previous machining is data representing an overview in one example. The discharge frequency is controlled by the discharge pause time of the power supply control unit 22a for machining. Here, the arithmetic unit 222 calculates the command value of the discharge pause time as the discharge frequency.

[0089] The pre-determined discharge frequency is used to set the machining shape of the workpiece 11 in the calculation obtained when performing the (n - 1)th machining based on the machining program to the desired shape. The machining shape of the workpiece 11 in the calculation often does not correspond to the machining shape of the actual workpiece 11. The command value of the discharge frequency calculated by the arithmetic unit 222 is a command value for eliminating the difference between the actual machining shape and the desired machining shape of the workpiece 11 obtained from the result of the previous machining, and is also a correction value for correcting the pre-determined discharge frequency. Therefore, the command value of the discharge frequency calculated by the arithmetic unit 222 is also referred to as a correction value. The arithmetic unit 222 outputs the correction value to the discharge frequency control unit 221.

[0090] When estimating the inter-electrode distance, a calculation model is obtained in advance that describes the relationship between at least one of the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory selected during the previous machining and the machining shape of the workpiece 11. Moreover, by using a function including this calculation model, the command value of the discharge frequency for setting the desired shape is calculated. In addition, the more data of the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory selected during the previous machining are included in the calculation model, the higher the accuracy of the estimated inter-electrode distance. Therefore, as data, it is preferable to use as many types of data as possible from the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory during the previous machining. The machining power supply control unit 22a controls the machining power supply 21 using the calculated command value of the discharge rest time.

[0091] Specifically, the arithmetic unit 222 estimates the inter-electrode distance at the current machining of the position that reflects the shape of the workpiece 11 during the previous machining using at least one of the average inter-electrode voltage, discharge frequency, machining speed, and drive trajectory selected during the previous machining. Based on the inter-electrode distance, the arithmetic unit 222 calculates a correction value of the discharge rest time on the basis of the estimation result of the inter-electrode distance so as to achieve the desired shape. In one example, when the result of the previous machining is that the position protrudes, that is, when the inter-electrode distance becomes smaller, the calculation is performed in such a way that the discharge rest time becomes shorter in order to increase the machining amount. In addition, when the result of the previous machining is that the position is concave, that is, when the inter-electrode distance becomes larger, the discharge rest time is calculated to be long so as to reduce the machining amount or not perform machining.

[0092] Under the control of the machining speed in the linear shape, the corrected discharge rest time OFF comp The calculation formula for calculation is represented by the following formula (7). Here, OFF is the set discharge rest amount during machining, and C1 is the conversion coefficient from the correction value of the machining speed to the correction value of the discharge rest amount. The discharge frequency control unit 221 adds the rest time OFF determined in advance by the machining conditions to the correction value, that is, the rest amount, calculated by the arithmetic unit 222, thereby calculating the command value of the corrected discharge rest time OFF comp As described above, by predicting the shape in advance and controlling the discharge frequency, it is possible to correct the unevenness on the machining surface without reducing the integral gain of the PI control system or increasing the proportional gain.

[0093] OFF comp = OFF + C1 * ∫f(V n-1 , F n-1 , S n-1 )dt ··· (7)

[0094] Under the control of the machining speed in the curve shape, the difference between the machined shape of the corner deduced from the result of the previous machining and the desired shape can also be eliminated by controlling the discharge frequency. Specifically, based on the overview of the previous machining stored in the inter-electrode average voltage storage unit 36, the discharge frequency storage unit 37, and the machining speed storage unit 38, the arithmetic unit 222 calculates the command value of the discharge frequency at the position of the corner according to the corner information stored in the drive locus storage unit 39. The corner information includes, as described above, the corner diameter, the opening angle of the corner, etc. The corrected discharge rest amount OFF comp_cnr in the corner is represented by the following formula (8). Here, the conversion coefficient from the correction value of the machining speed to the correction value of the discharge rest amount is set to C2.

[0095] OFF comp_cnr = OFF + C2 * ∫h(R, θ, V n-1 , F n-1 , S n-1 , x)dt ··

[0096] ·(8)

[0097] As Figure 6 and Figure 7 shown, when the shape of the corner in the result of the previous machining is smaller than the desired shape, or when there are irregularities such as concavities and convexities at the entrance and exit parts of the corner, the command value of the discharge frequency is calculated to match these shapes, thereby improving the shape accuracy after machining.

[0098] In Embodiment 2, n is set to an integer greater than or equal to 2, at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive locus in the (n - 1)-th machining in the case of machining the machining section n times is stored, and the inter-electrode distance in the n-th machining is calculated using a calculation model representing the relationship between the data in the (n - 1)-th machining and the machined shape of the workpiece 11. According to the inter-electrode distance in the n-th time, the command value of the discharge frequency in the n-th machining corresponding to the machining amount required to be the desired shape is calculated. That is, the machined shape of the workpiece 11 in the previous machining, i.e., the (n - 1)-th machining, can be estimated with high accuracy. Therefore, machining can be performed by shape correction machining so that the workpiece 11 becomes the desired shape. Thus, even when there is a sharp shape change in the result of the previous machining, high-precision shape correction machining can be performed.

[0099] Embodiment 3.

[0100] Figure 11It is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 3. In addition, the same reference numerals are given to the same structural elements as those in Embodiment 1, and their descriptions are omitted, and the differences from Embodiment 1 are described.

[0101] In the wire electrical discharge machining machine 1b according to Embodiment 3, a machining speed control unit 34b is provided instead of the machining speed control unit 34, and a drive locus control unit 31b is provided instead of the drive locus control unit 31.

[0102] The machining speed control unit 34b corresponds to the machining speed calculation unit 341 described in Embodiment 1 and does not have an arithmetic device 342. That is, the machining speed control unit 34b calculates the machining speed so that the difference calculated by the voltage arithmetic unit 33 becomes zero.

[0103] The drive locus control unit 31b includes: a drive locus calculation unit 311 that calculates a drive locus; and an arithmetic device 312 that calculates a correction value for correcting the drive locus calculated by the drive locus calculation unit 311. The drive locus calculation unit 311 calculates an instruction value of the drive locus of the machining electrode 10 according to the machining program. In addition, the drive locus calculation unit 311 can calculate the instruction value of the drive locus using at least one data selected from the average inter-electrode voltage, discharge frequency, and drive locus in the current machining. In the case of shape correction machining for improving surface roughness and shape accuracy, the drive locus calculation unit 311 corrects the calculated drive locus with the correction value calculated by the arithmetic device 312, and controls the machining electrode 10 with the corrected drive locus. The drive locus calculation unit 311 corresponds to the instruction value calculation unit.

[0104] In the case of shape correction machining for improving surface roughness and shape accuracy, the arithmetic unit 312 calculates the inter-electrode distance in the current machining using a calculation model representing the relationship between at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining and the machining shape of the workpiece 11. In addition, the arithmetic unit 312 calculates the command value of the drive trajectory in the current machining corresponding to the machining amount required to achieve the desired shape based on the machining program for machining the machining section. The drive trajectory calculated here is the axis movement trajectory, which is the trajectory for moving the axis of the wire electrical discharge machining machine 1b. When machining the machining section n times, the previous machining corresponds to the (n - 1)th machining, and the current machining corresponds to the nth machining. At least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining is, in one example, data representing an overview. Hereinafter, the command value of the drive trajectory in the current machining calculated by the arithmetic unit 312 is also referred to as the axis movement trajectory. The axis movement trajectory is a trajectory calculated in the direction of the normal vector perpendicular to the machining direction vector with respect to the machining direction vector obtained from the drive trajectory calculated by the drive trajectory calculation unit 311. In addition, the direction of the normal vector is a direction within the direction perpendicular to the machining direction vector and not the extending direction of the machining electrode 10.

[0105] The drive trajectory calculated by the drive trajectory calculation unit 311 is a trajectory for setting the machining shape of the computational workpiece 11 obtained during the (n - 1)th machining based on the machining program to the desired shape. The machining shape of the computational workpiece 11 often does not correspond to the actual machining shape of the workpiece 11. The command value of the axis movement trajectory calculated by the arithmetic unit 312 is a command value for eliminating the difference between the actual machining shape and the desired machining shape of the workpiece 11 resulting from the previous machining result, and is also a correction value for correcting the drive trajectory calculated by the drive trajectory calculation unit 311. Therefore, the command value of the axis movement trajectory calculated by the arithmetic unit 312 is also referred to as the correction value of the drive trajectory.

[0106] When estimating the inter - electrode distance, a calculation model is obtained in advance that describes the relationship between at least one of the average inter - electrode voltage, discharge frequency, machining speed, and drive trajectory selection during the previous machining and the machining shape of the workpiece 11. Moreover, by using a function including this calculation model, the command value for the axis movement trajectory to be set to the desired shape is calculated. In addition, the more the number of data of the average inter - electrode voltage, discharge frequency, machining speed, and drive trajectory selection during the previous machining increases in the calculation model, the higher the accuracy of the estimated inter - electrode distance. Therefore, as data, it is preferable to use as many types of data as possible from the average inter - electrode voltage, discharge frequency, machining speed, and drive trajectory during the previous machining. The drive trajectory control unit 31b outputs the calculated command value of the drive trajectory to the drive control unit 35.

[0107] The axis movement trajectory for the estimated inter - electrode distance is the trajectory calculated in the normal vector direction as described above. In one example, when the result of the previous machining is a protrusion at this position, that is, when the inter - electrode distance becomes smaller, in order to eliminate the convex shape, the axis movement trajectory in the direction towards the workpiece 11 is calculated. Additionally, when the result of the previous machining is a depression at this position, that is, when the inter - electrode distance becomes larger, in order to eliminate the concave shape, the correction value of the drive trajectory is calculated in the direction away from the workpiece 11.

[0108] Under the control of the machining speed in a linear shape, the axis movement trajectory ΔL in the normal direction with respect to the axis travel direction is calculated by the following formula (9). Here, C4 is the conversion coefficient of the machining speed correction value to the axis movement trajectory.

[0109] ΔL = C4 * ∫f(V n-1 , F n-1 , S n-1 )dt · · · (9)

[0110] Under the control of the machining speed in a curved shape, the difference from the desired shape in the machining shape of the corner estimated from the result of the previous machining can also be eliminated by controlling the axis movement trajectory. Specifically, based on the overview of the previous machining stored in the average inter - electrode voltage storage unit 36, discharge frequency storage unit 37, and machining speed storage unit 38, the arithmetic unit 312 calculates the correction value of the axis movement trajectory at this position of the corner according to the information of the corner stored in the drive trajectory storage unit 39. The information of the corner, as described above, includes the corner diameter, the opening angle of the corner, etc. The corrected axis movement trajectory ΔL in the corner comp_cnr is represented by the following formula (10). Here, the conversion coefficient of the machining speed correction value to the axis movement trajectory is set as C5.

[0111] ΔL comp_cnr= C5 * ∫h(R, θ, V n-1 , F n-1 , S n-1 , x)dt ··· (10)

[0112] As Figure 6 and Figure 7 shown, when the shape of the corner part in the result of the previous machining is smaller than the desired shape, or when there are irregularities such as concavities and convexities at the entrance and exit parts of the corner, the command value of the drive trajectory is calculated in accordance with these shapes, whereby the shape accuracy after machining can be improved.

[0113] In Embodiment 3, n is set to an integer greater than or equal to 2, at least one data among the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory selected in the (n - 1)-th machining in the case of machining the machining section n times is stored, the inter-electrode distance in the n-th machining is calculated using a calculation model representing the relationship between the data in the (n - 1)-th machining and the machining shape of the workpiece 11, and based on the inter-electrode distance in the n-th time, the axis movement trajectory in the n-th machining corresponding to the machining amount required to be the desired shape is calculated. That is, the machining shape of the workpiece 11 in the previous machining, i.e., the (n - 1)-th machining, can be estimated with high accuracy, and thus machining can be performed by shape correction machining so that the workpiece 11 becomes the desired shape. Thereby, even when there is a sharp shape change in the result of the previous machining, high-precision shape correction machining can be performed.

[0114] Embodiment 4.

[0115] In Embodiments 1 to 3, when performing shape correction machining, a calculation model that predicts in advance the inter-electrode distance at the time of the current machining at this position based on at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining is used. However, n may be set to an integer greater than or equal to 2. When the number of times of the shape correction machining, i.e., the current machining, is set to the n-th time, all data of the rough machining from the 1st to the (n - 1)-th times are stored in the inter-electrode average voltage storage unit 36, discharge frequency storage unit 37, machining speed storage unit 38, and drive trajectory storage unit 39, and the arithmetic units 342, 222, 312 can use all data from the 1st to the (n - 1)-th times to estimate the inter-electrode distance in the current machining at this position. Thereby, even when the data of the inter-electrode average voltage, discharge frequency, and machining speed at this part are disturbed due to interference in the previous machining, the inter-electrode distance can be stably estimated. That is, the accuracy of the machining shape of the workpiece 11 obtained from the results of the machining from the 1st to the (n - 1)-th times is improved compared with the case of using only the data of the (n - 1)-th time.

[0116] In addition, in the above description, a case where at least one data selected from the average inter-electrode voltage, discharge frequency, machining speed, and drive locus of the machining from the first time to the (n - 1)th time is shown, but at least one data selected from greater than or equal to two average inter-electrode voltages, discharge frequencies, machining speeds, and drive loci before the (n - 1)th time may also be used.

[0117] In Embodiment 4, in the average inter-electrode voltage storage unit 36, discharge frequency storage unit 37, machining speed storage unit 38, and drive locus storage unit 39, at least one data selected from the average inter-electrode voltage, discharge frequency, machining speed, and drive locus of a plurality of machinings before the (n - 1)th time is also stored. The inter-electrode distance in the current machining is estimated using a plurality of data in the machining before the (n - 1)th time, and at least one command value selected from the machining speed, discharge frequency, and drive locus is calculated based on the inter-electrode distance. Thus, even when the data of the average inter-electrode voltage, discharge frequency, and machining speed at this part is disturbed due to interference in the previous machining, the inter-electrode distance can be stably estimated.

[0118] Embodiment 5.

[0119] In Embodiments 1 to 4, in order to perform shape correction with higher accuracy, command values of a plurality of correction values selected from the correction value of the machining speed, correction value of the discharge frequency, and correction value of the drive locus in the current machining may be calculated.

[0120] Figure 12 FIG. is a block diagram showing an example of the structure of the wire electrical discharge machining machine according to Embodiment 5. In addition, the same reference numerals are given to the same structural elements as those in Embodiment 1, and the description thereof is omitted, and the differences from Embodiment 1 will be described.

[0121] In the wire electrical discharge machining apparatus 1c according to the fifth embodiment, the control unit 30 further includes an arithmetic unit 40. The arithmetic unit 40 combines the functions of the arithmetic units 342, 222, and 312 described in the first, second, and third embodiments. That is, in the case of shape correction machining for improving surface roughness and shape accuracy, the arithmetic unit 40 uses a calculation model representing the relationship between at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory selection in the previous machining and the machining shape of the workpiece 11 to calculate the inter-electrode distance in the current machining. Further, the arithmetic unit 40 calculates a plurality of correction values selected from the correction value of the machining speed, the correction value of the discharge frequency, and the correction value of the drive trajectory in the current machining corresponding to the machining amount required for the desired shape based on the machining program for machining the machining section. The arithmetic unit 40 outputs the calculation result to any one of the corresponding machining speed control unit 34c, the machining power supply control unit 22, and the drive trajectory control unit 31. Thus, the instruction value is corrected by at least two processing units among the machining speed control unit 34c, the machining power supply control unit 22, and the drive trajectory control unit 31. Alternatively, the arithmetic unit 40 can selectively use the instruction values for correction within the machining speed, discharge frequency, and drive trajectory according to the prediction result of the shape or inter-electrode distance obtained using the calculation model.

[0122] Further, in the wire electrical discharge machining apparatus 1c according to the fifth embodiment, instead of the machining speed control unit 34, a machining speed control unit 34c is provided. The machining speed control unit 34c corresponds to the machining speed calculation unit 341 described in the first embodiment and does not include the arithmetic unit 342. That is, the machining speed control unit 34c calculates the machining speed such that the difference calculated by the voltage arithmetic unit 33 becomes zero.

[0123] In addition, in the above description, the case where the arithmetic unit 40 performs calculations using at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the previous machining is shown. However, as in the fourth embodiment, it is also possible to perform calculations using at least one data selected from the inter-electrode average voltage, discharge frequency, machining speed, and drive trajectory in the machining up to the previous time.

[0124] In Embodiment 5, at least one piece of data selected from the inter-electrode average voltage, discharge frequency, machining speed, and driving trajectory in the previous machining is used to calculate the inter-electrode distance by a calculation model. Based on the calculated inter-electrode distance, a plurality of correction values selected from the correction value of the machining speed, the correction value of the discharge frequency, and the correction value of the driving trajectory in the current machining are calculated. Thus, compared with the cases of Embodiments 1 to 4, the machining accuracy of the workpiece 11 to the desired shape can be improved. In addition, sometimes the control of the machining speed, or the control of the discharge frequency, or the control of the axis movement trajectory may change according to the machining shape of the workpiece 11 estimated by the calculation model or the estimation result of the inter-electrode distance in the previous machining. In the above-described case, the arithmetic unit 40 can select a prescribed control method according to the estimation result of the machining shape or the inter-electrode distance.

[0125] Embodiment 6.

[0126] In Embodiments 1 to 5, the storage of the inter-electrode average voltage, machining speed, discharge frequency, and driving trajectory in the previous machining or up to the previous machining is described, and based on these results, in the current machining, at least one correction value selected from the machining speed, discharge frequency, and driving trajectory is calculated. When calculating the correction values of the machining speed, discharge frequency, and driving trajectory based on the stored overview, it is performed by an arithmetic expression that models the relationship between the inter-electrode distance or the machining shape of the workpiece 11 and the calculation object. However, in reality, due to the distance between the upper and lower nozzles during machining, the thickness of the workpiece 11, the deflection of the machining electrode 10 caused by machining hydraulic pressure, etc., and the discharge occurring behind the electrode traveling direction for surface finishing in a minute angle, sometimes the correction value of the calculation object cannot be calculated optimally. The calculation object here is at least one correction value selected from the machining speed, discharge frequency, and driving trajectory calculated based on the estimated inter-electrode distance. Therefore, in Embodiment 6, a method of optimizing the value of the calculation object, i.e., the object value, by machine learning based on the relationship between the overview of the inter-electrode average voltage, machining speed, discharge frequency, and driving trajectory obtained through pre-machining and the measurement result of the machining shape of the workpiece 11 after machining, i.e., the machining result of the workpiece 11, is described.

[0127] Next, the learning phase for generating the trained model and the effective use phase for estimating the command value information of the calculation object using the trained model generated through the learning phase are described in sequence.

[0128] <Learning Phase>

[0129] First, the pre-machining that becomes the learning phase is described. Figure 13This is a model diagram showing an example of the outline of the machine learning related to Embodiment 6. In preprocessing, teacher data is collected and stored. Specifically, through preprocessing, the processing specification values of the workpiece 11, the average inter-electrode voltage, processing speed, discharge frequency, drive trajectory during each processing, and the command value information and measurement results of the calculation object are obtained, and the relationship between the processing specification values, the average inter-electrode voltage, processing speed, discharge frequency, drive trajectory during each processing, and the command value information and measurement results of the calculation object is created as a database. An example of the processing specification value of the workpiece 11 is the thickness of the workpiece 11, the material of the workpiece 11, the positions of the upper and lower nozzles, and the diameter of the processing electrode 10, i.e., the wire diameter. The command value information of the calculation object is at least one data selected from the correction value of the processing speed, the correction value of the discharge frequency, and the correction value of the drive trajectory, and is the data actually used in preprocessing. The measurement result represents the error between the processed shape of the workpiece 11 after processing and the desired processing shape as a numerical value.

[0130] Next, the details of the learning stage will be described. Figure 14 This is a diagram schematically showing an example of the structure of the learning device used in the control device of the wire electrical discharge machining machine related to Embodiment 6. The learning device 50 includes a data acquisition unit 51, a teacher data storage unit 52, a model generation unit 53, and a trained model storage unit 54.

[0131] The data acquisition unit 51 acquires teacher data. The teacher data includes the processing specification values of the workpiece 11 in preprocessing, the average inter-electrode voltage, processing speed, discharge frequency, drive trajectory during each processing, and the command value information and measurement results of the calculation object. The data acquisition unit 51 outputs the processing specification values, average inter-electrode voltage, processing speed, discharge frequency, drive trajectory, and command value information of the calculation object as learning data to the model generation unit 53 within the acquired teacher data. The data acquisition unit 51 stores the acquired teacher data in the teacher data storage unit 52.

[0132] Preferably, all of the average inter-electrode voltage, processing speed, discharge frequency, and drive trajectory are included, but as long as at least one of them is included. In the following description, the case where all of the average inter-electrode voltage, processing speed, discharge frequency, and drive trajectory are included is taken as an example.

[0133] The teacher data storage unit 52 stores the relationship between the acquired teacher data, i.e., the processing specification values of the workpiece 11, the average inter-electrode voltage, processing speed, discharge frequency, drive trajectory during each processing time, and the command value information and measurement results of the calculation object, as a database. The teacher data stored in the teacher data storage unit 52 is used in the effective use stage described later.

[0134] The model generation unit 53 learns the first calculation object command value, which is the command value of the calculation object, based on the learning data created from the combination of the machining specification value, the average interelectrode voltage, the machining speed, the discharge frequency, the drive locus, and the command value information of the calculation object as the correct data. That is, a trained model is generated that infers the optimal first calculation object command value based on the machining specification value, the average interelectrode voltage, the machining speed, the discharge frequency, the drive locus, and the command value information of the calculation object. Here, the learning data is data that correlates the machining specification value, the average interelectrode voltage, the machining speed, the discharge frequency, the drive locus, and the command value information of the calculation object with each other. In addition, here, the learning data is a combination of the machining specification value, the average interelectrode voltage, the machining speed, the discharge frequency, the drive locus, and the command value information of the calculation object, but it may also be a combination of at least one data selected from the machining specification value, the average interelectrode voltage, the machining speed, the discharge frequency, and the drive locus and the command value information of the calculation object. In addition, the command value information of the calculation object used for the generation of the trained model is data that represents the desired shape as a command value.

[0135] The learning algorithm used by the model generation unit 53 can use known algorithms such as supervised learning. As an example, the case where a neural network is applied will be described.

[0136] In one example, the model generation unit 53 learns the first calculation object command value according to the neural network model by so-called supervised learning. Here, supervised learning refers to the following method, that is, by giving a data set of input and the label as the result to the learning device 50, the features existing in these learning data are learned, and the result is inferred based on the input.

[0137] The neural network is composed of an input layer composed of multiple neurons, an intermediate layer composed of multiple neurons, and an output layer composed of multiple neurons. The intermediate layer is also called the hidden layer, and it can be 1 layer, or 2 layers or more.

[0138] Figure 15 is a diagram schematically showing an example of the neural network used by the model generation unit. In one example, if it is Figure 15In the case of the three-layer neural network shown, if multiple inputs are input to input layers X1 to X3, these values are multiplied by the weights shown as w11 to w16 and input to intermediate layers Y1 to Y2. Without distinguishing from each other, the weights w11 to w16 are referred to as weight w1. Further, the results of intermediate layers Y1 to Y2 are multiplied by the weights shown as w21 to w26 and output from output layers Z1 to Z3. Without distinguishing from each other, the weights w21 to w26 are referred to as weight w2. The output results of output layers Z1 to Z3 change according to the values of weights w1 and w2.

[0139] In Embodiment 6, the neural network learns the first calculation object command value through so-called supervised learning in accordance with the learning data created based on the combination of the processing specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, the driving locus, and the command value information of the calculation object acquired by the data acquisition unit 51.

[0140] That is, the neural network learns by adjusting the weights w1 and w2 so that the result output from the output layer when the processing specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, and the driving locus are input to the input layer is close to the command value information of the calculation object.

[0141] The model generation unit 53 generates and outputs a trained model by performing the above learning.

[0142] The trained model storage unit 54 stores the trained model output from the model generation unit 53.

[0143] Next, the learning process implemented by the learning device 50 will be described. Figure 16 It is a flowchart showing an example of the order of the learning process implemented by the learning device according to Embodiment 6. First, the data acquisition unit 51 acquires the processing specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, the driving locus, the command value information of the calculation object as correct data, and the measurement result (step S51). In addition, it is assumed that the processing specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, the driving locus, the command value information of the calculation object, and the measurement result are acquired simultaneously, but as long as the processing specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, the driving locus, the command value information of the calculation object, and the measurement result can be input in association with each other, the data of the processing specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, the driving locus, the command value information of the calculation object, and the measurement result may also be acquired at different timings.

[0144] Next, the teacher data storage unit 52 stores the machining specification value, the inter-electrode average voltage, the machining speed, the discharge frequency and the drive trajectory, the command value information of the calculation object, and the measurement result (step S52).

[0145] Then, the model generation unit 53 creates learning data based on a combination of the processing specification values, inter-electrode average voltage, processing speed, discharge frequency and drive trajectory, and command value information of the calculation object obtained by the data acquisition unit 51, and learns the first calculation object command value through so-called taught learning to generate a trained model (step S53).

[0146] Then, the trained model storage unit 54 stores the trained model generated by the model generation unit 53 (step S54 ). The above processing ends.

[0147] <Effective use stage>

[0148] Next, the processing phase, which is the effective use phase, will be described. Figure 17 This is a model diagram showing an example of the outline of machine learning involved in Implementation Example 6. During processing, the overview of the previous processing or the processing until the previous processing is used as input data, the database is used as teacher data, and the instruction value of the calculation object, that is, the first calculation object instruction value and the second calculation object instruction value, are used as output data. The value of the object is calculated through machine learning so that the error relative to the desired shape becomes the minimum. Thus, the error in the actual processing and the formula that models the relationship between the inter-pole distance and the calculation object can be compensated by machine learning. The machine learning model at this time can use algorithms such as nearest neighbor regression, Bayesian optimization, etc.

[0149] Figure 18 1 is a diagram schematically showing an example of the configuration of an estimation device used in a control device for a wire electrical discharge machine according to Embodiment 6. The estimation device 60 includes a data acquisition unit 61 and an estimation unit 62 .

[0150] The data acquisition unit 61 acquires machining specification values, average inter-electrode voltage, machining speed, discharge frequency, and drive trajectory. In one example, the data acquisition unit 61 acquires machining specification values, average inter-electrode voltage, machining speed, discharge frequency, and drive trajectory in the n-1th, i.e., the previous machining.

[0151] The inference unit 62 infers the first calculation target command value using the trained model. That is, by inputting the processed specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, and the drive locus obtained by the data acquisition unit 61 into the trained model, the first calculation target command value can be inferred based on the processed specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, and the drive locus. In addition, here, an example is given of inferring the first calculation target command value based on the processed specification value, the average inter-electrode voltage, the processing speed, the discharge frequency, and the drive locus, but the first calculation target command value may also be inferred based on the processed specification value and at least one data selected from the average inter-electrode voltage, the processing speed, the discharge frequency, and the drive locus. In this case, a trained model that infers the first calculation target command value based on the processed specification value and at least one data selected from the average inter-electrode voltage, the processing speed, the discharge frequency, and the drive locus is used. Additionally, when inferring the first calculation target command value, the measurement result is not included in the input data. The reason is that it is not necessary to make the shape obtained from the processing result necessarily concave or convex, but to infer the command value with the smallest shape fluctuation.

[0152] In addition, if the inference unit 62 infers the first calculation target command value using the trained model, it uses the teacher data in the teacher data storage unit 52 to calculate the second calculation target command value, which is the command value of other calculation targets that minimizes the error with respect to the desired shape when the correction value of the calculation target is set to the first calculation target command value. The second calculation target command value of other calculation targets is within the correction value of the processing speed, the correction value of the discharge frequency, and the correction value of the drive locus, and is the command value of at least one calculation target remaining after removing the calculation target that becomes the first calculation target command value. The teacher data storage unit 52 stores the relationship between the processed specification value of the workpiece 11, the average inter-electrode voltage, the processing speed, the discharge frequency, the drive locus, the command value information of the calculation target, and the measurement result as a database. Therefore, in Embodiment 6, after the inference unit 62 estimates the first calculation target command value using the trained model, it uses the teacher data stored in the teacher data storage unit 52 to calculate the change in the error when changing any one of the command values of other calculation targets except the first calculation target command value, for example, the processing speed, the discharge frequency, and the drive locus, and obtains and outputs the second calculation target command value of the calculation target with the smallest error. That is, the inference unit 62 uses the teacher data in the teacher data storage unit 52 to learn and calculate the second calculation target command value of other calculation targets that minimizes the error with respect to the desired shape when the correction value of the calculation target is set to the first calculation target command value.

[0153] In addition, in Embodiment 6, it was described that the trained model learned by the model generation unit 53 of the control device of the wire electrical discharge machining machine 1 is used to infer the first calculation target command value, and the second calculation target command value of another calculation target at which the error becomes minimum when the correction value of the calculation target is set to the first calculation target command value is output. However, it is also possible to obtain a trained model from outside the control device of another wire electrical discharge machining machine 1 or the like, infer the first calculation target command value based on the trained model, and output the second calculation target command value of another calculation target at which the error becomes minimum when the correction value of the calculation target is set to the first calculation target command value.

[0154] Next, the inference process performed by the inference device 60 will be described. Figure 19 FIG. is a flowchart showing an example of the order of the inference process performed by the inference device according to Embodiment 6. First, the data acquisition unit 61 acquires machining specification values, interelectrode average voltage, machining speed, discharge frequency, and drive locus (step S71).

[0155] Next, the inference unit 62 inputs the machining specification values, interelectrode average voltage, machining speed, discharge frequency, and drive locus to the trained model stored in the trained model storage unit 54, and infers the first calculation target command value (step S72). Then, the inference unit 62 uses the teacher data in the teacher data storage unit 52 to calculate the second calculation target command value of another calculation target at which the error becomes minimum when the first calculation target command value obtained from the trained model is set as the correction value of the calculation target (step S73).

[0156] Then, the inference unit 62 outputs the calculated first calculation target command value and second calculation target command value to the corresponding processing units in the machining speed control unit 34c, the machining power supply control unit 22, and the drive locus control unit 31 (step S74).

[0157] Furthermore, each processing unit uses the output first calculation target command value and second calculation target command value to correct any command value among the machining speed, discharge frequency, and drive locus calculated by each processing unit (step S75). Thereby, machining can be performed to match the desired shape.

[0158] In addition, the model generation unit 53 can also learn the first calculation target command value according to the learning data created for the control devices of multiple wire electrical discharge machines 1. Additionally, the model generation unit 53 can obtain the learning data from the control devices of multiple wire electrical discharge machines 1 used in the same area, or can also learn the first calculation target command value by using the learning data collected from the control devices of multiple wire electrical discharge machines 1 that operate independently in different areas. Moreover, it is also possible to add the control device of the wire electrical discharge machine 1 that collects the learning data to the object midway or remove it from the object. Additionally, for the control device of a certain wire electrical discharge machine 1, the learning device 50 that has learned the first calculation target command value can be applied to the control device of another wire electrical discharge machine 1 different from it, and for the control device of the other wire electrical discharge machine 1, relearning is performed on the first calculation target command value for update.

[0159] In addition, the learning device 50 and the inference device 60 are used to learn the first calculation target command value of the control device of the wire electrical discharge machine 1, but in one example, they can also be connected to the control device of the wire electrical discharge machine 1 via a network and are devices separate from the control device of the wire electrical discharge machine 1. Additionally, the learning device 50 and the inference device 60 can also be built into the control device of the wire electrical discharge machine 1. Moreover, the learning device 50 and the inference device 60 can also exist on a cloud server.

[0160] In Embodiment 6, learning data is generated based on a combination of machining specification values, average inter-electrode voltage, machining speed, discharge frequency, drive locus, and instruction value information that is the calculation object of correct data, and a trained model that learns and infers the first calculation object instruction value is created. Then, the machining specification values, average inter-electrode voltage, machining speed, discharge frequency, and drive locus in the previous machining are input into the trained model to infer the first calculation object instruction value in the current machining. Additionally, when the correction value of the calculation object is set as the first calculation object instruction value, teacher data that stores the machining specification values, average inter-electrode voltage, machining speed, discharge frequency, drive locus, instruction value information of the calculation object, and measurement results in each machining in a database is used to calculate the second calculation object instruction value, which is the instruction value of other calculation objects that minimizes the error relative to the desired shape. In the equations that model the relationship between the inter-electrode distance and the calculation object as shown in Embodiments 1 to 5, in reality, due to factors such as the distance between the upper and lower nozzles during machining, the thickness of the workpiece 11, the deflection of the electrode caused by machining hydraulic pressure, etc., and discharges occurring behind the electrode traveling direction during surface finishing in a small angle, it is sometimes impossible to calculate the values optimally. However, in Embodiment 6, even due to factors such as the distance between the upper and lower nozzles during machining, the thickness of the workpiece 11, the deflection of the electrode caused by machining hydraulic pressure, etc., and discharges occurring behind the electrode traveling direction during surface finishing in a small angle, the object values can be optimized through machine learning based on the relationship between the average inter-electrode voltage, machining speed, discharge frequency, general situation of the drive locus, and measurement results obtained through pre-machining.

[0161] The power supply control unit 22 for machining and the control unit 30 described in Embodiments 1 to 6 correspond to the control device of the wire electrical discharge machining machine 1. Next, the hardware structure of the control device will be described. The control device is implemented by a processing circuit, which is a circuit that executes software by a processor. In one example, the processing circuit that executes the software is Figure 20 the control circuit shown. Figure 20 FIG. is an example of the hardware structure of the control device of the wire electrical discharge machining machine according to Embodiments 1 to 6. The control circuit 100 includes an input unit 101, a processor 102, a memory 103, and an output unit 104.

[0162] The input unit 101 is an interface circuit that receives data input from outside the control circuit 100 and supplies it to the processor 102. The output unit 104 is an interface circuit that sends data from the processor 102 or the memory 103 to the outside of the control circuit 100. When the processing circuit is Figure 20In the case of the control circuit 100 shown, each structural element of the processing power supply control unit 22 and the control unit 30 is implemented by the processor 102 reading and executing a program stored in the memory 103 corresponding to each structural element. The memory 103 is also used as a temporary memory in each process implemented by the processor 102. The processor 102 can output data such as calculation results to the memory 103 for storage, or can store data such as calculation results in the auxiliary storage device via the volatile memory of the memory 103.

[0163] The processor 102 is a CPU (also known as Central Processing Unit, central processing device, processing device, arithmetic device, microprocessor, microcomputer, processor, or DSP (Digital Signal Processor)). The memory 103 is, for example, a non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), EEPROM (registered trademark) (Electrically Erasable Programmable Read Only Memory), a magnetic disk, a floppy disk, an optical disk, a compact disc, a mini disc, or a DVD (Digital Versatile Disc).

[0164] Figure 20 This is an example of the hardware in the case where each structural element of the processing power supply control unit 22 and the control unit 30 is implemented by a general-purpose processor 102 and a memory 103, but each of the above structural elements can also be implemented by a dedicated hardware circuit. The processing circuit of the dedicated hardware circuit is a single circuit, a composite circuit, a programmed processor, a parallel-programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a circuit combining them. Each of the above structural elements can be implemented by a combination of the control circuit 100 and a dedicated hardware circuit.

[0165] The structures shown in the above embodiments represent an example, and can also be combined with other known technologies, can also combine the embodiments with each other, and a part of the structure can also be omitted or changed without departing from the gist.

[0166] Explanation of reference numerals

[0167] Wire discharge machining machines 1, 1a, 1b, 1c, machining electrode 10, workpiece 11, power supply unit 20, machining power supply 21, machining power supply control units 22, 22a, control unit 30, drive trajectory control units 31, 31b, average inter-electrode voltage detection unit 32, voltage calculation unit 33, machining speed control units 34, 34a, 34b, 34c, drive control unit 35, average inter-electrode voltage storage unit 36, discharge frequency storage unit 37, machining speed storage unit 38, drive trajectory storage unit 39, arithmetic units 40, 222, 312, 342, learning device 50, data acquisition units 51, 61, teacher data storage unit 52, model generation unit 53, trained model storage unit 54, inference device 60, inference unit 62, discharge frequency control unit 221, drive trajectory calculation unit 311, machining speed calculation unit 341.

Claims

1. A control device for a wire electrical discharge machining machine, which controls a driving locus of an electrode relative to a workpiece, a relative machining speed between the workpiece and the electrode, and a discharge frequency of a voltage periodically applied between the electrode and the workpiece in a wire electrical discharge machining machine that machines the workpiece by applying a voltage between the workpiece and the electrode to generate a discharge, The control device for the wire electrical discharge machining machine is characterized by comprising: a storage device that sets n as an integer greater than or equal to 2 and stores at least one data selected from an inter-electrode average voltage, which is the voltage applied between the workpiece and the electrode in the (n - 1)-th machining when machining a specified machining section of the workpiece n times, the discharge frequency, the machining speed, and the driving locus; an arithmetic device that calculates an inter-electrode distance in the n-th machining using a calculation model representing the relationship between the data in the (n - 1)-th machining and the machining shape of the workpiece, and calculates at least one command value selected from the machining speed, the discharge frequency, and the driving locus in the n-th machining according to the inter-electrode distance in the n-th machining; and a learning device, The learning device comprises: a data acquisition unit that acquires teacher data, which includes machining specification values of the workpiece in pre-machining, at least one data selected from the inter-electrode average voltage, the machining speed, the discharge frequency, and the driving locus in each machining, and calculation value information that is the object of the command value; and a model generation unit that generates a trained model for inferring a first calculation object command value that is the object of the command value according to the machining specification values acquired by the data acquisition unit and at least one data selected from the inter-electrode average voltage, the machining speed, the discharge frequency, and the driving locus.

2. The control device for the wire electrical discharge machining machine according to claim 1, wherein the storage device further stores at least one of the data selected from the inter-electrode average voltage, the discharge frequency, the machining speed, and the driving locus in a machining earlier than the (n - 1)-th machining, and the arithmetic device calculates at least one command value selected from the machining speed, the discharge frequency, and the driving locus in the n-th machining using the calculation model representing the relationship between the data in the machining before the (n - 1)-th machining and the machining shape of the workpiece.

3. The control device for the wire electrical discharge machining machine according to claim 1 or 2, wherein It further includes an instruction value calculation unit that uses at least one of the average voltage between electrodes, the discharge frequency, and the driving locus in the n-th machining to calculate at least one instruction value of the machining speed, the discharge frequency, and the driving locus, and corrects the calculated instruction value with at least one of the instruction values selected from the machining speed, the discharge frequency, and the driving locus in the n-th machining calculated by the arithmetic device.

4. The control device of a wire electrical discharge machining machine according to claim 1, wherein: When the machining section is in a straight line shape, the calculation model includes a function representing the relationship between at least one of the data selected from the average voltage between electrodes, the discharge frequency, the machining speed, and the driving locus in the (n-1)-th machining and the machining shape of the workpiece.

5. The control device of a wire electrical discharge machining machine according to claim 2, wherein: When the machining section is in a straight line shape, the calculation model includes a function representing the relationship between at least one of the data selected from the average voltage between electrodes, the discharge frequency, the machining speed, and the driving locus in the machining before the (n-1)-th machining and the machining shape of the workpiece.

6. The control device of a wire electrical discharge machining machine according to claim 1, wherein: When the machining section is at a corner of a curved shape, the calculation model includes a function representing the relationship between the corner diameter and the opening angle of the corner, at least one of the data selected from the average voltage between electrodes, the discharge frequency, the machining speed, and the driving locus in the (n-1)-th machining, and the shape of the workpiece.

7. The control device of a wire electrical discharge machining machine according to claim 2, wherein: When the machining section is at a corner of a curved shape, the calculation model includes a function representing the relationship between the corner diameter and the opening angle of the corner, at least one of the data selected from the average voltage between electrodes, the discharge frequency, the machining speed, and the driving locus in the machining before the (n-1)-th machining, and the shape of the workpiece.

8. The control device of a wire electrical discharge machining machine according to claim 1, wherein: It further includes an inference device having: A teacher data storage unit that stores teacher data representing the relationship between the machining specification values of the workpiece in pre-machining, the average voltage between electrodes, the machining speed, the discharge frequency, the driving locus in each machining, the calculated value information of the object of the instruction value, and the measurement result of the deviation from the desired shape with respect to the machining shape in each machining; A data acquisition unit that acquires the machining specification values of the workpiece and at least one data selected from the average voltage between electrodes, the machining speed, the discharge frequency, and the driving locus in the (n-1)-th machining; And An inference unit that uses a trained model for inferring a first calculation target command value, which is a command value of a calculation target in the nth machining, based on the machining specification value and at least one piece of data selected from the inter-electrode average voltage, the machining speed, the discharge frequency, and the drive locus in the (n - 1)th machining. The inference unit infers the first calculation target command value based on the machining specification value obtained by the data acquisition unit and at least one piece of data selected from the inter-electrode average voltage, the machining speed, the discharge frequency, and the drive locus in the (n - 1)th machining. Using the teacher data, when the command value of the calculation target is set as the first calculation target command value, the inference unit calculates a second calculation target command value, which is a command value of another calculation target that minimizes the error with respect to the desired shape.

9. A control method for a wire electrical discharge machining machine, which controls a drive locus of an electrode relative to a workpiece, a relative machining speed between the workpiece and the electrode, and a discharge frequency of a voltage periodically applied between the electrode and the workpiece by applying a voltage between the workpiece and the electrode to generate a discharge and machining the workpiece. The control method for the wire electrical discharge machining machine is characterized in that a storage device sets n as an integer greater than or equal to 2, and stores at least one piece of data selected from the inter-electrode average voltage, which is the voltage applied between the workpiece and the electrode in the (n - 1)th machining, the discharge frequency, the machining speed, and the drive locus in the case of machining a specified machining section of the workpiece n times. An arithmetic unit calculates an inter-electrode distance in the nth machining using a calculation model representing the relationship between the data in the (n - 1)th machining and the machining shape of the workpiece, and calculates at least one command value selected from the machining speed, the discharge frequency, and the drive locus in the nth machining corresponding to the machining amount required for the desired shape based on the machining program for machining the machining section from the inter-electrode distance in the nth machining. A learning process The learning process includes: A data acquisition process that acquires teacher data, which includes a machining specification value of the workpiece in pre-machining, at least one piece of data selected from the inter-electrode average voltage, the machining speed, the discharge frequency, and the drive locus in each machining, and calculation value information of an object that becomes the command value. And A model generation process that generates a trained model for inferring a first calculation target command value, which is an object of the command value, based on the machining specification value obtained by the data acquisition process and at least one piece of data selected from the inter-electrode average voltage, the machining speed, the discharge frequency, and the drive locus.

Images