Wire electrical discharge machining apparatus and wire electrical discharge machining method

Abstract

The wire electrical discharge machining apparatus (100) comprises a wire electrode (1) having cutting wire portions (1a) that are spaced apart in parallel from each other and facing the workpiece (W), a nozzle (7) having injection holes for supplying machining fluid to the gap between the cutting wire portions (1a) and the workpiece (W), an outer peripheral component arranged on the outer circumference of the workpiece (W), and a measuring unit connected to the interior of the space formed by the outer peripheral component and acquiring data for estimating the flow velocity of the machining fluid flowing through the machining groove formed in the workpiece (W) by electrical discharge machining.

Description

【Technical Field】 【0001】 The present disclosure relates to a wire electrical discharge machining apparatus and a wire electrical discharge machining method for discharging an electrical discharge between a wire electrode and a workpiece to perform electrical discharge machining on the workpiece. 【Background Art】 【0002】 In a multi-wire electrical discharge machining apparatus, an electrical discharge is generated between a plurality of wire electrodes and a workpiece, and a plurality of plate-like members are cut out from the workpiece at once. The multi-wire electrical discharge machining apparatus is used, for example, in a semiconductor manufacturing process for slice machining to cut out a plurality of wafers from an ingot. Machining debris always occurs in the gap between the wire electrode and the workpiece during electrical discharge machining. The machining debris caused by the electrical discharge becomes a factor that makes the electrical discharge machining unstable. Therefore, in order to perform stable electrical discharge machining, it is necessary to inject the machining fluid from the nozzle toward the machining groove for the purpose of discharging the machining debris from the gap. 【0003】 When the flow velocity of the machining fluid inside the machining groove decreases, the machining debris tends to stay in the machining groove, and the frequency of electrical discharge from the wire electrode to the machining debris increases. When the frequency of electrical discharge to the machining debris increases, the frequency of electrical discharge in the machining progress direction decreases, so that the machining efficiency by the multi-wire electrical discharge machining apparatus decreases accordingly. In addition, the more the staying machining debris increases, the easier it is for a short circuit to occur between the wire electrode and the machining debris. When a short-circuit current having energy larger than the discharge current flows between the wire electrode and the machining debris, the wire electrode is likely to be broken. Therefore, in the multi-wire electrical discharge machining apparatus, in order to enable stable electrical discharge machining, it is desirable to be able to discharge the machining debris from the machining groove before the machining debris increases to such an extent that a short circuit occurs. And in this case, there is a desire to grasp the flow velocity of the machining fluid flowing in the machining groove in order to appropriately control the injection amount of the machining fluid. 【0004】 Patent Document 1 describes that by providing a machining fluid flow path restricting portion surrounding the periphery of a columnar workpiece, the machining fluid easily enters the machining groove of the workpiece, and the machining debris is easily discharged. [Prior art documents] [Patent Documents] 【0005】 [Patent Document 1] International Publication No. 2022 / 234659 [Overview of the project] [Problems that the invention aims to solve] 【0006】 However, in semiconductor wafer manufacturing, the required product thickness is extremely thin, requiring a spacing of approximately 0.6 mm between adjacent machining grooves. This makes it difficult to directly attach measuring instruments such as flow meters to measure the flow rate of the machining fluid. Furthermore, the wire electrical discharge machining apparatus described in Patent Document 1 could not determine the flow velocity of the machining fluid within the machining grooves. In addition, although the wire electrical discharge machining apparatus described in Patent Document 1 is said to have close contact between the nozzle and the workpiece, in reality, due to manufacturing errors in each component of the machining fluid flow path limiting section, it is difficult to assemble the components without any gaps, and small gaps exist in various parts of the machining fluid flow path limiting section. While it is conceivable to determine the flow rate of the machining fluid flowing through the machining grooves from the amount of machining fluid injected and the number of machining grooves, it is difficult to grasp the flow rate of the machining fluid that has passed through the machining grooves because the machining fluid may leak out through the gaps in the aforementioned machining fluid flow path limiting section. 【0007】 This disclosure has been made in view of the above, and aims to provide a wire electrical discharge machining apparatus that can grasp the flow velocity of the machining fluid flowing in the machining groove. [Means for solving the problem] 【0008】 To solve the above-mentioned problems and achieve the objective, the wire electrical discharge machining apparatus according to the present disclosure comprises: wire electrodes having cutting wire portions that are spaced apart in parallel from each other and facing the workpiece; nozzles having injection holes for supplying machining fluid to the gap between the cutting wire portions and the workpiece; outer peripheral components arranged on the outer circumference of the workpiece; and a measuring unit connected to the interior of the space formed by the outer peripheral components and acquiring data for estimating the flow velocity of the machining fluid flowing through the machining grooves formed in the workpiece by electrical discharge machining. [Effects of the Invention] 【0009】 According to this disclosure, the effect is that it becomes possible to determine the flow velocity of the machining fluid flowing within the machining groove. [Brief explanation of the drawing] 【0010】 [Figure 1] A perspective view showing the main components of an example configuration of a wire electrical discharge machining apparatus according to Embodiment 1. [Figure 2] A perspective view showing an example of the configuration of the machining fluid guide section of a wire electrical discharge machining apparatus according to Embodiment 1. [Figure 3] Exploded perspective view showing an example of the configuration of the machining fluid guide section of the wire electrical discharge machining apparatus according to Embodiment 1. [Figure 4] Cross-sectional view showing an example of the configuration of the machining fluid guide section of the wire electrical discharge machining apparatus according to Embodiment 1. [Figure 5] Cross-sectional view showing an example of the configuration of the machining fluid guide section of the wire electrical discharge machining apparatus according to Embodiment 1. [Figure 6] Enlarged view showing a specific area in Figure 5. [Figure 7] This figure shows an example of the functional configuration of the control device for a wire electrical discharge machining apparatus according to Embodiment 1. [Figure 8] Flowchart showing the procedure for wire electrical discharge machining using the wire electrical discharge machining apparatus according to Embodiment 1 [Figure 9] This figure shows an example of the functional configuration of the control device for a wire electrical discharge machining apparatus according to Embodiment 2. [Figure 10] Diagram showing the configuration of the learning device according to Embodiment 2. [Figure 11] Figure showing the method for estimating the accumulation state of machining chips in the machining groove according to Embodiment 2 [Figure 12] Flowchart showing the processing procedure of the learning process by the learning device according to Embodiment 2 [Figure 13] Figure showing the configuration of the neural network used by the learning device according to Embodiment 2 [Figure 14] Figure showing the configuration of the inference device according to Embodiment 2 [Figure 15] Flowchart showing the processing procedure of the inference process by the inference device according to Embodiment 2 [Figure 16] Figure showing the configuration in which the respective functions of the control units according to Embodiments 1 and 2 are realized by hardware [Figure 17] Figure showing the configuration in which the respective functions of the control units according to Embodiments 1 and 2 are realized by software 【Embodiments for Carrying Out the Invention】 【0011】 Hereinafter, the wire electrical discharge machining apparatus and the wire electrical discharge machining method according to the embodiments will be described in detail based on the drawings. 【0012】 Embodiment 1. FIG. 1 is a perspective view showing a main part of a configuration example of a wire electrical discharge machining apparatus 100 according to Embodiment 1. The wire electrical discharge machining apparatus 100 is an apparatus that generates a discharge between a wire electrode 1 and a workpiece W to perform electrical discharge machining on the workpiece W. Specifically, the wire electrical discharge machining apparatus 100 generates a discharge between a plurality of cutting wire portions 1a that run in parallel and spaced apart from each other and the workpiece W, and performs electrical discharge machining on the workpiece W by the energy of the discharge, and is a multi-wire electrical discharge machining apparatus that simultaneously cuts a plurality of plate-like members from the workpiece W. 【0013】 In FIG. 1, the X-axis, Y-axis, and Z-axis of a three-axis orthogonal coordinate system are shown. The X-axis direction coincides with the traveling direction of the wire electrode 1 on the workpiece W, that is, the traveling direction of the wire electrode 1 with respect to the workpiece W arranged in the wire electrical discharge machining apparatus 100. Also, the X-axis direction coincides with the left-right direction when the wire electrical discharge machining apparatus 100 is viewed from the front. The arrow 101 in FIG. 1 indicates the traveling direction of the wire electrode 1 on the workpiece W. The Y-axis direction coincides with the direction in which the wire electrodes 1 are arranged in parallel on the workpiece W, that is, the direction in which the wire electrodes 1 are arranged in parallel with respect to the workpiece W arranged in the wire electrical discharge machining apparatus 100. Also, the Y-axis direction coincides with the depth direction when the wire electrical discharge machining apparatus 100 is viewed from the front. The Z-axis direction coincides with the height direction of the wire electrical discharge machining apparatus 100. The height direction of the wire electrical discharge machining apparatus 100 coincides with the up-down direction. 【0014】 The wire electrical discharge machining apparatus 100 includes a wire electrode 1, a plurality of guide rollers 2, a power supply unit 3, a moving device 4, a control device 5, a machining fluid guide portion 6, a nozzle 7, and a workpiece pressing and holding device 8. 【0015】 Examples of the material of the workpiece W include materials such as tungsten, molybdenum, silicon carbide, single-crystal silicon, single-crystal silicon carbide, gallium nitride, and polycrystalline silicon. Silicon carbide is also referred to as silicon carbide. 【0016】 The plurality of guide rollers 2 serve to guide the traveling of the wire electrode 1. The shape of the guide roller 2 is a columnar shape extending in the Y-axis direction. The guide roller 2 is rotated by a motor (not shown). The number of guide rollers 2 is four in the first embodiment, but may be a plurality other than four. Hereinafter, when distinguishing the four guide rollers 2, they are referred to as guide rollers 2a, 2b, 2c, and 2d. Each of the guide rollers 2a, 2b, 2c, and 2d is rotatably arranged with the Y-axis as the rotation axis. The rotation axes of each of the guide rollers 2a, 2b, 2c, and 2d are parallel to each other. By the rotation axes of each of the guide rollers 2a, 2b, 2c, and 2d being parallel to each other, the wire electrode 1 can be made to travel with high precision. 【0017】 Guide rollers 2a, 2b, 2c, and 2d are positioned in a plane perpendicular to the axis of rotation, i.e., in the XZ plane in Figure 1, spaced apart from each other in the X-axis and Z-axis directions. Specifically, guide rollers 2a and 2b are positioned at the same height and spaced apart from each other in the X-axis direction. Guide rollers 2c and 2d are also positioned at the same height and spaced apart from each other in the X-axis direction. Guide roller 2c is positioned below guide roller 2b and spaced apart from it. Guide roller 2d is positioned below guide roller 2a and spaced apart from it. The axis of rotation of each guide roller 2a, 2b, 2c, and 2d is positioned to coincide with each vertex of the rectangle. 【0018】 The wire electrode 1 plays the role of cutting the workpiece W. A single wire electrode 1, fed from a feed bobbin (not shown), is repeatedly wound around guide rollers 2a, 2b, 2c, and 2d in that order. Specifically, the wire electrode 1 is wound multiple times around the outer circumferential surface of each of the guide rollers 2a, 2b, 2c, and 2d, at intervals in the direction along the respective rotation axes of the guide rollers 2a, 2b, 2c, and 2d. The wire electrode 1 travels along with the rotation of the guide rollers 2a, 2b, 2c, and 2d, and is finally wound from guide roller 2b onto a winding bobbin (not shown). 【0019】 The wire electrode 1 has a plurality of cutting wire sections 1a that cut the workpiece W. The plurality of cutting wire sections 1a are parts of the wire electrode 1 that are spaced apart from each other and facing the workpiece W. The plurality of cutting wire sections 1a are parts of the wire electrode 1 that are stretched between the guide roller 2c and the guide roller 2d. The plurality of cutting wire sections 1a are spaced apart from each other and arranged in the direction along the respective rotation axes of the guide rollers 2c and 2d. It is preferable that the plurality of cutting wire sections 1a are arranged parallel to each other. 【0020】 The electron supply unit 3 supplies power to the wire electrode 1, generating a discharge between the wire electrode 1 and the workpiece W. The electron supply unit 3 is cylindrical, extending in the Y-axis direction. In this embodiment 1, there are two electron supply units 3, but this can be changed as appropriate. Hereafter, the two electron supply units 3 will be referred to as electron supply units 3a and 3b. Each of the electron supply units 3a and 3b is in contact with the wire electrode 1. Specifically, each of the electron supply units 3a and 3b is positioned below each cutting wire section 1a and is in contact with each cutting wire section 1a. The electron supply units 3a and 3b are positioned spaced apart from each other in the X-axis direction, with the workpiece W in between. 【0021】 One electron supply terminal 3a is positioned in the X-axis direction between the guide roller 2c and the nozzle 7a, which will be described later. The other electron supply terminal 3b is positioned in the X-axis direction between the guide roller 2d and the nozzle 7b, which will be described later. The wire electrical discharge machining apparatus 100 is equipped with a power supply, such as a power panel (not shown). The power supply terminals of the power supply are electrically connected to the electron supply terminals 3a and 3b, respectively. The ground terminal of the power supply is electrically connected to the workpiece W. The voltage (voltage pulse) output from the power supply is applied between each cutting wire portion 1a of the wire electrode 1 and the workpiece W. This makes it possible to generate a discharge between each cutting wire portion 1a and the workpiece W. 【0022】 The moving device 4 plays the role of relatively moving the workpiece W between the multiple cutting wire sections 1a, on which the wire electrodes 1 are arranged in parallel, and the workpiece W. Specifically, the moving device 4 changes the relative position between each of the parallel cutting wire sections 1a and the workpiece fixing plate 62 on which the workpiece W is placed (details to be described later). In this embodiment 1, the position of each cutting wire section 1a in the vertical direction, i.e., the Z-axis direction, and the position of the processing fluid escape prevention plate 63 of the processing fluid guide section 6 (details to be described later) are fixed, and the workpiece fixing plate 62 and the workpiece W can be moved vertically by the moving device 4. The moving device 4 is positioned below the workpiece fixing plate 62 and the workpiece W. The upper end of the moving device 4 is fixed to the workpiece fixing plate 62. The workpiece W is fixed to the moving device 4 via the workpiece fixing plate 62. 【0023】 The moving device 4 moves the components inside the processing fluid guide section 6 together with the workpiece W in the vertical direction relative to the pair of processing fluid escape prevention plates 63, i.e., in the Z-axis direction. As a result, the wire electrical discharge machining apparatus 100 performs electrical discharge machining on the workpiece W by moving the workpiece fixing plate 62 on which the workpiece W is placed closer to or further away from the cutting wire section 1a. In addition, the electrical discharge machining on the workpiece W forms a processing groove Wz along the cutting wire section 1a, which will be described later, in the workpiece W. Finally, the workpiece W is cut into a plurality of plate-like members. The moving device 4 may be movable in the X-axis direction, Y-axis direction, and Z-axis direction. 【0024】 The nozzle 7 plays the role of removing machining debris generated by electrical discharge machining by spraying machining fluid 9 between the wire electrode 1 and the workpiece W. Specifically, the nozzle 7 sprays machining fluid 9 towards the machining grooves formed in the workpiece W by electrical discharge machining, so as to discharge the machining debris generated within the grooves from within the grooves. The machining fluid 9 is supplied to the nozzle 7 from a machining fluid tank (not shown) through a pump (not shown) and a machining fluid supply pipe (not shown). One nozzle 7 is located on each side of the workpiece W in the X-axis direction. Specifically, the nozzle 7 is located between the electron supply 3 and the machining fluid guide section 6. In this embodiment 1, there are two nozzles 7, but this number may be changed as appropriate. Hereinafter, the two nozzles 7 will be referred to as nozzles 7a and 7b. 【0025】 One nozzle 7a is positioned in the X-axis direction between the processing fluid guide section 6 and the electron supply section 3a, that is, between the workpiece W and the electron supply section 3a. The other nozzle 7b is positioned in the X-axis direction between the processing fluid guide section 6 and the electron supply section 3b, that is, between the workpiece W and the electron supply section 3b. The insides of nozzles 7a and 7b are filled with processing fluid 9. Nozzle 7 has injection holes 71 that spray the processing fluid 9 filled inside toward the workpiece W in the processing fluid guide section 6. The parallel cutting wire sections 1a are inserted through the injection holes 71 of nozzles 7a and 7b and travel along them. 【0026】 Figure 2 is a perspective view showing an example configuration of the processing fluid guide section 6 of the wire electrical discharge machining apparatus 100 according to Embodiment 1. Figure 3 is an exploded perspective view showing an example configuration of the processing fluid guide section 6 of the wire electrical discharge machining apparatus 100 according to Embodiment 1. Figure 4 is a cross-sectional view showing an example configuration of the processing fluid guide section 6 of the wire electrical discharge machining apparatus 100 according to Embodiment 1. Figure 4 shows a cross-section in the XZ plane showing the state in which cutting is being performed on a cylindrical workpiece W by the cutting wire section 1a. Figure 5 is a cross-sectional view showing an example configuration of the processing fluid guide section 6 of the wire electrical discharge machining apparatus 100 according to Embodiment 1. Figure 5 shows a cross-section in the YZ plane at a position corresponding to the line segment VV in Figure 4. Figure 6 is an enlarged view showing a specific region A in Figure 5. 【0027】 The processing fluid guide section 6 is an outer peripheral component that is arranged to surround the outer circumference of the workpiece W and forms a sealed space around the workpiece W. The processing fluid guide section 6 comprises a pair of processing fluid rectifier plates 61, namely processing fluid rectifier plate 61a and processing fluid rectifier plate 61b, arranged to surround the workpiece W, a workpiece fixing plate 62, a pair of processing fluid escape prevention plates 63, namely processing fluid escape prevention plate 63a and processing fluid escape prevention plate 63b, and a workpiece holding section 64. The pair of processing fluid escape prevention plates 63 constitute the first member of the processing fluid guide section 6, over which the cutting wire section 1a is stretched. The pair of processing fluid rectifier plates 61 and the workpiece fixing plate 62 constitute the second member of the processing fluid guide section 6, which together with the first member form a sealed space in which the workpiece W is fixed and processing fluid 9 flows into the workpiece W from the first member. The workpiece holding portion 64 is held at a position spaced above the cutting wire portion 1a and constitutes a third member that holds the workpiece W, which is inserted into the space and divided during cutting, from above. 【0028】 The pair of processing fluid rectifier plates 61 are arranged parallel to the direction of travel of the cutting wire section 1a and rectify the flow of processing fluid 9 sprayed from the nozzle 7 into the processing fluid guide section 6. During processing of the workpiece W, the pair of processing fluid rectifier plates 61, namely processing fluid rectifier plate 61a and processing fluid rectifier plate 61b, are positioned in close contact with the workpiece W, sandwiching it in the depth direction, i.e., the Y-axis direction. 【0029】 The workpiece W is placed and fixed on the workpiece fixing plate 62. The workpiece W is fixed on the workpiece fixing plate 62 by a jig (not shown) for fixing the workpiece W. The workpiece W is fixed on the workpiece fixing plate 62 such that each end face of the workpiece W is sandwiched between a pair of processing fluid rectifier plates 61 in the depth direction, i.e., the Y-axis direction, and is in close contact with the pair of processing fluid rectifier plates 61. The workpiece fixing plate 62 is fixed to a moving device 4 that can move linearly in the height direction, i.e., the Z-axis direction. 【0030】 In the processing fluid guide section 6, the workpiece fixing plate 62 and a pair of processing fluid rectifier plates 61 move up and down relative to a pair of processing fluid escape prevention plates 63 as the moving device 4 moves up and down. When processing the workpiece W, the moving device 4 raises the workpiece fixing plate 62 and the pair of processing fluid rectifier plates 61 at an appropriate speed, thereby performing electrical discharge machining. As shown in Figures 5 and 6, a processing groove Wz is formed in the workpiece W along the cutting wire section 1a, and finally a thin plate is cut off from the workpiece W. 【0031】 The pair of processing fluid escape prevention plates 63 are in close contact with the end faces of the workpiece fixing plate 62 and the pair of processing fluid rectifier plates 61, and are positioned on both sides of the workpiece W sandwiched between the processing fluid rectifier plates 61. The pair of processing fluid escape prevention plates 63 are fixed in place and do not move up or down. 【0032】 A pair of processing fluid escape prevention plates 63 are connected to the nozzle 7. Processing fluid escape prevention plate 63a is connected to nozzle 7a. Processing fluid escape prevention plate 63b is connected to nozzle 7b. In the pair of processing fluid escape prevention plates 63, through holes 631 are formed in the parts that come into contact with the injection holes 71 of nozzles 7a and 7b (described later), allowing the processing fluid 9 to be ejected and passed through the parallel-running cutting wire section 1a. The injection holes 71 of nozzles 7a and 7b and the through holes 631 of the processing fluid escape prevention plates 63 are the same size. In Figures 2 and 3, for convenience, the through holes 631 of the processing fluid escape prevention plates 63 and the injection holes 71 of nozzle 7 are shown as rectangular parallelepiped openings. The processing fluid escape prevention plates 63 are also in close contact with the workpiece holding section 64. 【0033】 The workpiece holder 64 is moved and held in the height direction by a workpiece holder and holding device 8 equipped with a linear actuator. At the start of the cutting process, the workpiece holder 64 is held in an initial cutting position, spaced upward from the workpiece W and the cutting wire 1a. After the start of the cutting process, the workpiece holder 64 moves vertically to fix the thin plate being processed from the workpiece W. 【0034】 The workpiece holder 8 supports the workpiece holder 64. The workpiece holder 8 also has a linear actuator that moves the workpiece holder 64 vertically. The workpiece holder 8 is installed in a location where its relative position to the cutting wire 1a does not change, such as on the surface plate of the wire electrical discharge machining apparatus 100. 【0035】 A cutting fluid 9 is supplied from a pair of nozzles 7 to the workpiece W via a pair of cutting fluid escape prevention plates 63. Specifically, the cutting fluid 9 is supplied from nozzle 7a to the workpiece W via the cutting fluid escape prevention plate 63a. Also, the cutting fluid 9 is supplied from nozzle 7b to the workpiece W via the cutting fluid escape prevention plate 63b. 【0036】 When a certain voltage is applied between the electrodes of the cutting wire section 1a and the workpiece W, and the distance between the electrodes reaches a certain range, a discharge occurs between the electrodes. The high heat generated by this discharge melts the workpiece W, and as a result, multiple plate-shaped members are cut out at once. During processing, if the processing fluid 9 is supplied to the gap between the workpiece W and the cutting wire section 1a, the processing debris generated between the workpiece W and the cutting wire section 1a can be discharged to the outside of the gap. This processing debris can cause a short circuit between the workpiece W and the cutting wire section 1a. Therefore, by supplying the processing fluid 9 to the gap between the workpiece W and the cutting wire section 1a during processing, the frequency of short circuits can be reduced. 【0037】 Furthermore, a machining fluid tank (not shown) and a pump (not shown) are connected to the nozzles 7a and 7b, and machining fluid 9 is supplied from the machining fluid tank. Alternatively, a machining fluid guide unit 6 to which the workpiece W is fixed may be installed inside a machining tank (not shown) containing the machining fluid 9, and electrical discharge machining may be performed with the workpiece W immersed in the machining fluid 9. 【0038】 Furthermore, the wire electrical discharge machining apparatus 100 is equipped with a plurality of measuring units 10 attached to the processing fluid rectifier plate 61. The measuring units 10 are connected to the inside of a sealed space formed by the processing fluid guide portion 6, which is an outer peripheral component, and acquire data at a predetermined period to estimate the flow velocity of the processing fluid 9 flowing through the processing groove Wz formed in the workpiece W by electrical discharge machining during the machining of the workpiece W. The measuring units 10 transmit the acquired data to the control device 5. In Figures 1 to 3 and 5, eight pressure sensors 11, which are an example of measuring units 10, are attached to the processing fluid rectifier plate 61a. The pressure sensors 11 are attached in a position that does not overlap with the workpiece W in the in-plane direction of the XZ plane, which is the in-plane direction of the processing fluid rectifier plate 61a. 【0039】 The pressure sensor 11 measures the pressure in the sealed space inside the processing fluid guide section 6, which consists of a pair of processing fluid rectifier plates 61, a pair of processing fluid escape prevention plates 63, a workpiece fixing plate 62, and a workpiece holding section 64 that surround the workpiece W. The pressure sensor 11 also measures the pressure in the space around the pressure sensor 11 within the sealed space inside the processing fluid guide section 6. In the wire electrical discharge machining apparatus 100, eight pressure sensors 11 are attached to the processing fluid rectifier plate 61a, so the pressure at eight locations within the space inside the processing fluid guide section 6 is measured by these eight pressure sensors 11. The pressure values ​​measured by the pressure sensors 11 indirectly include information about the size of the gaps between the nozzle 7 and the components constituting the processing fluid guide section 6. 【0040】 The processing fluid guide section 6 is configured to form a sealed space inside, but in reality, due to manufacturing tolerances of each component, it is difficult for the components to be assembled without any gaps at all. As a result, small gaps exist between the nozzle 7 and the processing fluid escape prevention plate 63, and between the processing fluid rectifier plate 61 and the workpiece holding section 64, etc. The size of these gaps causes changes in the pressure at various points within the internal space of the processing fluid guide section 6. The information about the size of the gaps between the nozzle 7 and the components constituting the processing fluid guide section 6 is indirectly included in the pressure values ​​at various points within the internal space of the processing fluid guide section 6. 【0041】 In Embodiment 1, an example was shown in which the pressure sensor 11 is placed on the processing fluid rectifier plate 61. However, the mounting location of the pressure sensor 11 is not limited as long as the pressure in the space within the processing fluid guide section 6 can be measured. Furthermore, the measurement section 10 is not limited to the pressure sensor 11, as long as it can obtain information for estimating the flow velocity of the processing fluid 9 in the processing groove Wz. 【0042】 The control device 5 plays a role in controlling the operation of the entire wire electrical discharge machining apparatus 100. The control device 5 controls, for example, the injection of machining fluid 9 between the wire electrode 1 and the workpiece W, the moving device 4, the workpiece holding device 8, and the machining power supply (not shown). The control device 5 drives the moving device 4 to control the relative position in the height direction between the workpiece fixing plate 62 on which the workpiece W is placed and each cutting wire section 1a. The control device 5 outputs a voltage application command to the machining power supply to control the generation of a discharge between each cutting wire section 1a and the workpiece W. 【0043】 Figure 7 shows an example of the functional configuration of the control device 5 of the wire electrical discharge machining apparatus 100 according to Embodiment 1. The control device 5 comprises a calculation unit 51, a control unit 52, and a storage unit 53. 【0044】 The calculation unit 51 acquires pressure data from the pressure sensor 11 at multiple locations within the sealed space inside the machining fluid guide unit 6 when the workpiece W is being machined. Based on this pressure data, the calculation unit 51 estimates the flow velocity of the machining fluid 9 flowing through the machining groove Wz formed in the workpiece W by electrical discharge machining. 【0045】 One method by which the calculation unit 51 estimates the flow velocity of the processing fluid 9 is to use a regression equation to determine the flow velocity of the processing fluid 9 flowing in the processing groove Wz, with the pressure at each point in the sealed space inside the processing fluid guide unit 6, the flow rate of the processing fluid 9 sprayed from the nozzle 7, the depth of the processing groove Wz, and the number of parallel processing grooves Wz as variables. When processing the workpiece W, the calculation unit 51 can calculate an estimated value of the flow velocity of the processing fluid 9 flowing in the actual processing groove Wz by substituting the measured pressure at each point in the sealed space inside the processing fluid guide unit 6 obtained by the pressure sensor 11, the flow rate of the processing fluid 9 sprayed from the nozzle 7, the depth of the processing groove Wz, and the number of parallel processing grooves Wz into the regression equation. The calculation unit 51 transmits the calculation result, which is the data of the flow velocity of the processing fluid 9 flowing in the actual processing groove Wz, to the control unit 52. 【0046】 The regression equation is created by performing a fluid analysis in advance using a 3D model that simulates the workpiece W and the workpiece W surrounding the workpiece W, with the pressure at each point in the sealed space inside the workpiece fluid guide section 6, the flow rate of the workpiece fluid 9 sprayed from the nozzle 7, the depth of the workpiece groove Wz, and the number of parallel workpiece grooves Wz as variables, and is stored in the memory section 53. 【0047】 The control unit 52 controls the operation of the entire wire electrical discharge machining apparatus 100. For example, the control unit 52 controls the injection of machining fluid 9 between the wire electrode 1 and the workpiece W, the moving device 4, the workpiece holding device 8, and the machining power supply (not shown). 【0048】 Furthermore, the control unit 52 includes a flow rate control unit 521 that controls the flow rate of the processing fluid 9 sprayed from the nozzle 7. 【0049】 The flow rate control unit 521 automatically adjusts the flow rate setting of the machining fluid 9 sprayed from the nozzle 7 based on the data of the flow velocity value of the machining fluid 9 flowing through the machining groove Wz formed in the workpiece W by electrical discharge machining, which is estimated by the calculation unit 51. The flow rate of the machining fluid 9 sprayed from the nozzle 7 is automatically controlled to an appropriate flow rate for cooling between the electrodes between the cutting wire section 1a and the workpiece W, and for the discharge of machining debris generated in the machining groove Wz by electrical discharge machining. In this case, it is sufficient that information showing the relationship between the flow velocity value of the machining fluid 9 flowing through the machining groove Wz and the flow rate value of the machining fluid 9 sprayed from the nozzle 7 is determined in advance by simulation or the like and stored in the flow rate control unit 521. 【0050】 Furthermore, the operator, who is the user of the wire electrical discharge machining apparatus 100, may adjust the set value of the flow rate of the processing fluid 9 sprayed from the nozzle 7 based on the data of the flow velocity value of the processing fluid 9 flowing through the processing groove Wz formed in the workpiece W by electrical discharge machining, which is estimated by the calculation unit 51. In other words, the flow rate control unit 521 can control the flow rate of the processing fluid 9 sprayed from the nozzle 7 based on the set value of the flow rate of the processing fluid 9 sprayed from the nozzle 7, which is set by the user to the flow rate control unit 521. 【0051】 Furthermore, the nozzle 7 is supplied with processing fluid 9 from the processing fluid tank 72 through a pump 73 and a processing fluid supply pipe (not shown). The flow rate of the processing fluid 9 supplied to the nozzle 7 is switched by the operation of a flow control valve 74 installed between the pump 73 and the nozzle 7, and the operation of the flow control valve 74 is controlled by a flow control unit 521. 【0052】 Here, a target flow rate value for the processing fluid 9 flowing through the processing groove Wz may be set in advance in the flow rate control unit 521. The flow rate control unit 521 compares the flow rate value of the processing fluid 9 estimated by the calculation unit 51 with the target flow rate value and opens and closes the flow rate adjustment valve 74 so that the flow rate value of the processing fluid 9 estimated by the calculation unit 51 approaches the target flow rate value, thereby automatically controlling the flow rate of the processing fluid 9 sprayed from the nozzle 7. The target flow rate value should be determined in advance by simulation or the like and stored in the flow rate control unit 521 so as to efficiently perform inter-electrode cooling between the cutting wire section 1a and the workpiece W, and discharge of processing chips generated in the processing groove Wz by electrical discharge machining from the processing groove Wz. 【0053】 The memory unit 53 stores various types of information used to control the wire electrical discharge machining apparatus 100. 【0054】 Next, a wire electrical discharge machining method for machining a workpiece W using the wire electrical discharge machining apparatus 100 according to Embodiment 1 will be described. Figure 8 is a flowchart showing the procedure for the wire electrical discharge machining method using the wire electrical discharge machining apparatus 100 according to Embodiment 1. 【0055】 In step S10, electrical discharge machining of the workpiece W is started. Specifically, the control device 5 controls each component of the wire electrical discharge machining apparatus 100 to start the electrical discharge machining of the workpiece W. In the electrical discharge machining of the workpiece W started in step S10, the following steps are performed: running multiple cutting wire sections 1a over the workpiece W, whose outer circumference is surrounded by a machining fluid guide section 6 which is an outer peripheral component, to form a machining groove Wz in the workpiece W by electrical discharge machining; and supplying machining fluid 9 from the nozzle 7 into the gap between the cutting wire section 1a and the workpiece W. After that, the process proceeds to step S20. 【0056】 In step S20, the pressure in the sealed space inside the processing fluid guide section 6 is measured. Specifically, multiple pressure sensors 11 acquire pressure data at multiple locations within the sealed space inside the processing fluid guide section 6. The multiple pressure sensors 11 transmit the measured pressure data to the calculation unit 51 of the control device 5. Step S20 can be described as a step to acquire data for estimating the flow velocity of the processing fluid 9 flowing through the processing groove Wz in the space formed by the processing fluid guide section 6, which is an outer peripheral component. The process then proceeds to step S30. 【0057】 In step S30, the flow velocity of the machining fluid 9 flowing through the machining groove Wz formed in the workpiece W by electrical discharge machining is estimated. Specifically, the calculation unit 51 estimates the flow velocity of the machining fluid 9 flowing through the machining groove Wz formed in the workpiece W by electrical discharge machining. The calculation unit 51 calculates an estimated value of the flow velocity of the machining fluid 9 flowing in the actual machining groove Wz by substituting the current measured pressure at each point in the sealed space inside the machining fluid guide unit 6 obtained from the pressure sensor 11, the current flow rate of the machining fluid 9 sprayed from the nozzle 7, the current depth of the machining groove Wz, and the number of parallel machining grooves Wz into a regression equation. The calculation unit 51 transmits the data of the estimated flow velocity of the machining fluid 9 flowing in the actual machining groove Wz, which is the calculation result, to the flow rate control unit 521 of the control unit 52. Then, the process proceeds to step S40. 【0058】 In step S40, the flow rate of the processing fluid 9 sprayed from the nozzle 7 is controlled based on the flow velocity value of the processing fluid 9 estimated by the calculation unit 51. Specifically, the flow rate control unit 521 controls the flow rate of the processing fluid 9 sprayed from the nozzle 7 to an appropriate flow rate, i.e., an appropriate spray amount, based on the flow velocity value of the processing fluid 9 estimated by the calculation unit 51. Then, the process returns to step S20. 【0059】 According to the wire electrical discharge machining apparatus 100 of the above-described embodiment 1, a wire electrical discharge machining apparatus is realized that comprises a wire electrode having cutting wire portions that are spaced apart from each other in parallel and facing the workpiece, a nozzle having an injection hole for supplying machining fluid to the gap between the cutting wire portions and the workpiece, an outer peripheral component arranged on the outer circumference of the workpiece, and a measuring unit connected to the inside of the space formed by the outer peripheral component and acquiring data for estimating the flow velocity of the machining fluid flowing through the machining groove formed in the workpiece by electrical discharge machining. 【0060】 The wire electrical discharge machining apparatus 100 according to the above-described embodiment 1 is equipped with a pressure sensor 11 as a measuring unit that is connected to the inside of the space formed by the processing fluid guide section 6, which is an outer peripheral component, and acquires data for estimating the flow velocity of the processing fluid 9 flowing through the processing groove Wz formed in the workpiece W by electrical discharge machining. As a result, the wire electrical discharge machining apparatus 100 can estimate the flow velocity of the processing fluid 9 flowing through the processing groove Wz formed in the workpiece W by electrical discharge machining based on the pressure measurement data from multiple locations in the sealed space inside the processing fluid guide section 6 obtained by the pressure sensor 11. Furthermore, the wire electrical discharge machining apparatus 100 improves the accuracy of estimating the flow velocity of the processing fluid 9 by using the pressure from multiple locations inside the processing fluid guide section 6 as measurement data in the sealed space inside the processing fluid guide section 6. 【0061】 Furthermore, the wire electrical discharge machining apparatus 100 can efficiently perform both electrode cooling between the cutting wire section 1a and the workpiece W, and discharge of machining debris generated in the machining groove Wz by electrical discharge machining, by controlling the flow rate of the machining fluid 9 sprayed from the nozzle 7 based on the estimated flow velocity value of the machining fluid 9. 【0062】 Therefore, the wire electrical discharge machining apparatus 100 according to Embodiment 1 has the effect of making it possible to determine the flow velocity of the machining fluid 9 flowing in the machining groove Wz. 【0063】 Embodiment 2. In Embodiment 1 described above, the flow velocity of the machining fluid 9 flowing through the machining groove Wz was estimated by fluid analysis using a 3D model. However, in reality, it is necessary to create a 3D model for each EDM condition, and the fluid analysis is time-consuming. In Embodiment 2, in view of the above, we will describe a case in which the appropriate flow rate value of the machining fluid 9 sprayed from the nozzle 7 can be estimated using machine learning without performing fluid analysis using a 3D model or estimating the flow velocity of the machining fluid 9 flowing through the machining groove Wz. 【0064】 Figure 9 shows an example of the functional configuration of the control device 5a of the wire electrical discharge machining apparatus 100a according to Embodiment 2. The wire electrical discharge machining apparatus 100a according to Embodiment 2 differs from the wire electrical discharge machining apparatus 100 according to Embodiment 1 described above in that it is equipped with a control device 5a instead of a control device 5. In addition to the configuration of the control device 5 according to Embodiment 1, the control device 5a according to Embodiment 2 further includes a machine learning device 200 which comprises a learning device 210 and an inference device 220. The control device 5a calculates the flow rate value of the machining fluid 9 sprayed from the nozzle 7 using the machine learning device 200. 【0065】 First, we will explain the learning phase in the learning device 210 of the machine learning device 200. 【0066】 Based on the learning data, the learning device 210 learns the appropriate flow rate of the machining fluid 9 injected from the nozzle 7, which corresponds to the pressure value measured by the pressure sensor 11 acquired by the data acquisition unit 211, the depth of the machining groove Wz calculated from the amount of rise of the moving device 4 from the start of machining, the number of parallel machining grooves Wz, and the state of machining debris accumulation in the machining groove Wz. As a result, the learning device 210 can learn the appropriate flow rate of the machining fluid 9 injected from the nozzle 7 without performing fluid analysis using a 3D model. 【0067】 Figure 10 shows the configuration of the learning device 210 according to Embodiment 2. The learning device 210 comprises a data acquisition unit 211, a model generation unit 212, and a trained model storage unit 213. The data acquisition unit 211 is the first data acquisition unit in the machine learning device 200. The trained model storage unit 213 may be provided outside the learning device 210, or it may be provided in the storage unit 53 of the control device 5a. 【0068】 The data acquisition unit 211 acquires the following information as training data: the pressure value measured by the pressure sensor 11, the depth of the machining groove Wz formed in the workpiece W, which is calculated from the amount the moving device 4 rises from the start of machining, the number of parallel machining grooves Wz, the state of accumulation of machining debris in the machining groove Wz, and the flow rate value of the machining fluid 9 sprayed from the nozzle 7. 【0069】 The pressure value measured by the pressure sensor 11 is data used to estimate the flow velocity of the machining fluid 9 flowing through the machining groove Wz formed in the workpiece W by electrical discharge machining. The number of parallel machining grooves Wz can be expressed as the number of parallel cutting wire sections 1a. In the following, the pressure value measured by the pressure sensor 11 may be referred to as "pressure value," the depth of the machining groove Wz calculated from the rise of the moving device 4 from the start of machining may be referred to as "depth of machining groove Wz," the number of parallel machining grooves Wz may be referred to as "number of parallel machining grooves Wz," the state of accumulation of machining debris in the machining groove Wz may be referred to as "state of accumulation of machining debris," and the flow rate value of the machining fluid 9 sprayed from the nozzle 7 may be referred to as "flow rate value of machining fluid 9." In Figure 10, the pressure value measured by the pressure sensor 11 is labeled "Pressure Value," the depth of the machining groove Wz calculated from the amount of rise of the moving device 4 from the start of machining is labeled "Depth of Machining Groove," the number of parallel machining grooves Wz is labeled "Number of Parallel Machining Grooves," the state of accumulation of machining debris in the machining groove Wz is labeled "Debris Accumulation State," and the flow rate value of the machining fluid 9 sprayed from the nozzle 7 is labeled "Flow Rate Value of Machining Fluid." 【0070】 The accumulation of machining debris in the machining groove Wz is estimated by the current value flowing between the cutting wire section 1a and the workpiece W, which is obtained by current sensors (not shown) connected to each of the cutting wire sections 1a arranged in parallel in the wire electrical discharge machining apparatus 100a. Figure 11 is a diagram showing the method for estimating the accumulation of machining debris in the machining groove Wz according to Embodiment 2. Figure 11 shows an example of a graph representing the change in the measured current value of the discharge current flowing between the cutting wire section 1a and the workpiece W each time an applied voltage (voltage pulse) is applied between each cutting wire section 1a and the workpiece W. 【0071】 In Figure 11, the vertical axis represents the current value of the discharge current, and the horizontal axis represents time. In Figure 11, the threshold Th is the threshold value used to determine whether or not machining debris is accumulating in the machining groove Wz. The determination of whether or not machining debris is accumulating in the machining groove Wz is performed, for example, by the control unit 52. The control unit 52 acquires data on the current value flowing between the cutting wire section 1a and the workpiece W, obtained by the current sensor, and uses the threshold Th to determine whether or not machining debris is accumulating in the machining groove Wz. The threshold Th is predetermined and stored in the control unit 52. The threshold Th may also be stored in the storage unit 53 of the control device 5a. 【0072】 If the discharge current value exceeds the threshold Th, it is presumed that a relatively large amount of machining debris accumulated in the machining groove Wz, causing a short circuit between the cutting wire section 1a and the workpiece W, resulting in a large current flowing between the cutting wire section 1a and the workpiece W. If the discharge current value is below the threshold Th, it is presumed that a relatively small amount of machining debris accumulated in the machining groove Wz, and no short circuit occurred between the cutting wire section 1a and the workpiece W. 【0073】 Therefore, when the discharge current value exceeds the threshold Th, the data acquisition unit 211 generates information about the accumulation status of machining debris in the machining groove Wz, for example, "there is a lot of machining debris accumulation in the machining groove Wz." Also, when the discharge current value is less than or equal to the threshold Th, the data acquisition unit 211 generates information about the accumulation status of machining debris in the machining groove Wz, for example, "there is little machining debris accumulation in the machining groove Wz." In this case, the data acquisition unit 211 acquires either the information "there is a lot of machining debris accumulation in the machining groove Wz" or the information "there is little machining debris accumulation in the machining groove Wz" as information about the accumulation status of machining debris. Note that the information about the accumulation status of machining debris may be one of three or more classifications of the state of machining debris accumulation in the machining groove Wz. 【0074】 The flow rate value of the processing fluid 9 acquired as training data is the appropriate flow rate value of the processing fluid 9 to be sprayed from the nozzle 7 in a state corresponding to the pressure value, depth of the processing groove Wz, number of parallel processing grooves Wz, and the state of processing chip accumulation, which are acquired as training data. 【0075】 The appropriate flow rate value for the processing fluid 9 is the flow rate of the processing fluid 9 that, in a state corresponding to the information contained in the training data excluding the flow rate value of the processing fluid 9, can efficiently perform inter-electrode cooling between the cutting wire portion 1a and the workpiece W, and discharge the processing chips generated in the processing groove Wz by electrical discharge machining from the processing groove Wz, and is the ideal flow rate of the processing fluid 9 sprayed from the nozzle 7. 【0076】 Here, the learning data is data that associates the following information: pressure value, depth of machining groove Wz, number of parallel machining grooves Wz, chip retention status, and flow rate value of machining fluid 9. The data acquisition unit 211 acquires the learning data from the control unit 52 of the control device 5a. Alternatively, the learning data may be input to the data acquisition unit 211 from an external device of the control device 5a. The data acquisition unit 211 generates the learning data by associating the pressure value, depth of machining groove Wz, number of parallel machining grooves Wz, chip retention status, and flow rate value of machining fluid 9. 【0077】 The model generation unit 212 learns the appropriate flow rate of the processing fluid 9 to be sprayed from the nozzle 7, corresponding to the pressure value, depth of the processing groove Wz, number of parallel processing grooves Wz, and processing debris retention status, based on training data created based on combinations of the pressure value, depth of the processing groove Wz, number of parallel processing grooves Wz, and processing debris retention status, acquired by the data acquisition unit 211. In other words, the model generation unit 212 generates a trained model 233 for inferring the appropriate flow rate of the processing fluid 9 to be sprayed from the nozzle 7, corresponding to the pressure value, depth of the processing groove Wz, number of parallel processing grooves Wz, and processing debris retention status, acquired by the data acquisition unit 211. 【0078】 Next, the processing procedure of the learning device 210 will be explained using Figure 12. Figure 12 is a flowchart showing the processing procedure of the learning device 210 according to Embodiment 2. 【0079】 In step S110, the data acquisition unit 211 acquires training data. Specifically, the data acquisition unit 211 acquires information such as the pressure value, the depth of the machining groove Wz, the number of parallel machining grooves Wz, the state of machining debris accumulation, and the flow rate of the machining fluid 9 as training data. The data acquisition unit 211 may acquire the information on the pressure value, the depth of the machining groove Wz, the number of parallel machining grooves Wz, the state of machining debris accumulation, and the flow rate of the machining fluid 9 at the same time, or at different times. In other words, the data acquisition unit 211 may acquire each of the above pieces of information at any time, as long as the information on the pressure value, the depth of the machining groove Wz, the number of parallel machining grooves Wz, the state of machining debris accumulation, and the flow rate of the machining fluid 9 can be associated with each other. 【0080】 In step S120, the model generation unit 212 performs a learning process by supervised learning, according to the training data obtained by the data acquisition unit 211, which is a combination of pressure value, depth of machining groove Wz, number of parallel machining grooves Wz, state of machining debris accumulation, and flow rate value of machining fluid 9. The model generation unit 212 generates a trained model 233 by supervised learning according to the training data, for example. 【0081】 In step S130, the trained model storage unit 213 stores the trained model 233. That is, the model generation unit 212 causes the generated trained model 233 to be stored in the trained model storage unit 213. 【0082】 The model generation unit 212 can use any known learning algorithm, such as supervised learning, unsupervised learning, or reinforcement learning. As an example, we will describe the case where the model generation unit 212 performs supervised learning using a neural network. 【0083】 The model generation unit 212 learns the appropriate flow rate value of the processing fluid 9 to be sprayed from the nozzle 7 by so-called supervised learning, for example, according to a neural network model. Here, supervised learning is a method in which a learning device is given pairs of input and result (label) data, learns features in the training data, and infers the result from the input. 【0084】 A neural network consists of an input layer made up of multiple neurons, a hidden layer (intermediate layer) also made up of multiple neurons, and an output layer also made up of multiple neurons. The hidden layer can be one or more layers. 【0085】 Figure 13 shows the configuration of the neural network used by the learning device 210 according to Embodiment 2. For example, in a three-layer neural network as shown in Figure 13, when multiple data are input to input layers X1 to X3, these values ​​are multiplied by weights w11 to w16 and input to hidden layers Y1 to Y2, and the result is further multiplied by weights w21 to w26 and output from output layers Z1 to Z3. This output result varies depending on the values ​​of weights w11 to w16 and weights w21 to w26. 【0086】 In Embodiment 2, the neural network learns the flow rate of the processing fluid 9 ejected from the nozzle 7 by so-called supervised learning, according to training data (dataset) created based on a combination of pressure value, depth of processing groove Wz, number of parallel processing grooves Wz, processing chip retention status, and flow rate value of processing fluid 9, all acquired by the data acquisition unit 211. 【0087】 In other words, the neural network learns by inputting information acquired by the data acquisition unit 211, such as pressure values, depth of machining grooves Wz, number of parallel machining grooves Wz, and state of machining debris retention, into input layers X1 to X3, and adjusting weights w11 to w16 and w21 to w26 so that the output from output layers Z1 to Z3 approaches the flow rate of the machining fluid 9 that corresponds to the state indicated by the information acquired by the data acquisition unit 211, such as pressure values, depth of machining grooves Wz, number of parallel machining grooves Wz, and state of machining debris retention. 【0088】 The model generation unit 212 generates and outputs a trained model 233 by performing the training described above. 【0089】 The trained model storage unit 213 stores the trained model 233 output from the model generation unit 212. 【0090】 Next, we will explain the application phase of the machine learning device 200 in the inference device 220. 【0091】 Figure 14 shows the configuration of the inference device 220 according to Embodiment 2. The inference device 220 comprises a data acquisition unit 221 and an inference unit 222. The data acquisition unit 221 is the second data acquisition unit in the machine learning device 200. The inference unit 222 is connected to the trained model storage unit 213. 【0092】 The data acquisition unit 221 acquires the following information as inference data: the pressure value measured by the pressure sensor 11, the depth of the machining groove Wz calculated from the amount the moving device 4 rises from the start of machining, the number of parallel machining grooves Wz, and the status of machining debris accumulation in the machining groove Wz. The inference data is input from the control unit 52 to the data acquisition unit 221. 【0093】 The inference unit 222 uses the trained model 233 to output a flow rate value of the processing fluid 9 sprayed from the nozzle 7 that corresponds to the state indicated by the inference data. The inference unit 222 reads the trained model 233 from the trained model storage unit 213. The inference unit 222 inputs the pressure value, the depth of the processing groove Wz, the number of parallel processing grooves Wz, and the state of processing chip accumulation into the trained model 233. As a result, the inference unit 222 infers the flow rate value of the processing fluid 9 sprayed from the nozzle 7 from the information of the pressure value, the depth of the processing groove Wz, the number of parallel processing grooves Wz, and the state of processing chip accumulation. In other words, by inputting inference data into the trained model 233 for inferring the flow rate value of the processing fluid 9, the inference unit 222 can output a flow rate value of the processing fluid 9 sprayed from the nozzle 7 that corresponds to the state indicated by the inference data. 【0094】 Next, using Figure 15, the processing procedure of the inference device 220 for obtaining the flow rate value of the processing fluid 9 sprayed from the nozzle 7 will be explained. Figure 15 is a flowchart showing the processing procedure of the inference device 220 according to Embodiment 2. 【0095】 In step S210, the data acquisition unit 221 acquires inference data. Specifically, the data acquisition unit 221 acquires the following information as inference data: the pressure value measured by the pressure sensor 11, the depth of the machining groove Wz calculated from the amount the moving device 4 rises from the start of machining, the number of parallel machining grooves Wz, and the state of machining debris accumulation in the machining groove Wz. 【0096】 In step S220, the inference unit 222 inputs the following information, which is inference data acquired by the data acquisition unit 221: pressure value, depth of machining groove Wz, number of parallel machining grooves Wz, and state of machining debris accumulation, into the trained model 233 stored in the trained model storage unit 213. As an inference result obtained by the trained model 233, the inference unit 222 obtains the flow rate value of the machining fluid 9 sprayed from the nozzle 7 corresponding to the input information. 【0097】 In step S230, the inference result obtained by the trained model 233 is output to the flow rate control unit 521 of the control unit 52. Specifically, the inference unit 222 outputs the flow rate value information of the processing fluid 9 sprayed from the nozzle 7, which is the inference result obtained by the trained model 233, to the flow rate control unit 521 of the control unit 52. 【0098】 In step S240, the flow rate control unit 521 controls the flow rate of the processing fluid 9 sprayed from the nozzle 7 based on the inference result output by the inference unit 222 in step S230 described above. That is, the flow rate control unit 521 adjusts the flow rate of the processing fluid 9 sprayed from the nozzle 7 by controlling the operation of the flow rate adjustment valve 74 using the flow rate value of the processing fluid 9, which is the inference result. As a result, the wire electrical discharge machining apparatus 100a can automatically control the flow rate of the processing fluid 9 sprayed from the nozzle 7, and efficiently perform inter-electrode cooling between the cutting wire section 1a and the workpiece W, and discharge of processing chips generated in the processing groove Wz by electrical discharge machining from the processing groove Wz. 【0099】 The learning device 210 and the inference device 220 are used to learn the flow rate value of the processing fluid 9 sprayed from the nozzle 7. However, the learning device 210 and the inference device 220 may be separate devices from the control device 5a, connected to the control device 5a via a network such as the Internet. At least one of the learning device 210 and the inference device 220 may be connected to the control device 5a via a network, for example. At least one of the learning device 210 and the inference device 220 may be a separate device from the control device 5a. Furthermore, at least one of the learning device 210 and the inference device 220 may reside on a cloud server. At least one of the learning device 210 and the inference device 220 may be built into the control device 5a. 【0100】 Furthermore, in Embodiment 2, the flow rate value of the processing fluid 9 sprayed from the nozzle 7 was described as being output using a trained model 233 learned by the model generation unit 212 of the control device 5a. However, a trained model may also be obtained from an external device such as another control device 5a, and the flow rate value of the processing fluid 9 sprayed from the nozzle 7 may be output based on this trained model. 【0101】 Furthermore, while Embodiment 2 described a case where supervised learning is applied to the learning algorithm used by the model generation unit 212, it is not limited to this. In addition to supervised learning, reinforcement learning, unsupervised learning, or semi-supervised learning can also be applied to the learning algorithm. 【0102】 Furthermore, the model generation unit 212 may learn the flow rate value of the machining fluid 9 sprayed from the nozzle 7 according to the training data created for multiple wire EDM machines 100a. The model generation unit 212 may acquire training data from multiple wire EDM machines 100a used in the same area, or it may learn the flow rate value of the machining fluid 9 sprayed from the nozzle 7 using training data collected from multiple wire EDM machines 100a operating independently in different areas. It is also possible to add or remove wire EDM machines 100a from the target midway through the process of collecting training data. Moreover, a training device that has learned the flow rate value of the machining fluid 9 sprayed from the nozzle 7 for one wire EDM machine 100a may be applied to another wire EDM machine 100a, and the flow rate value of the machining fluid 9 sprayed from the nozzle 7 for that other wire EDM machine 100a may be retrained and updated. 【0103】 Furthermore, the learning algorithm used in the model generation unit 212 can be deep learning, which learns to extract the features themselves, or machine learning can be performed according to other known methods, such as genetic programming, functional logic programming, or support vector machines. 【0104】 According to the learning device 210 described above, a learning device is realized that includes a first data acquisition unit that acquires learning data including data for estimating the flow velocity of the machining fluid flowing through the machining grooves formed in the workpiece by electrical discharge machining, the depth of the machining grooves formed in the workpiece by electrical discharge machining, the number of parallel machining grooves which is the number of machining grooves arranged in parallel in the workpiece, the state of accumulation of machining debris in the machining grooves, and the flow rate of the machining fluid sprayed from the nozzle, and a model generation unit that uses the learning data to generate a trained model for inferring the flow rate of the machining fluid sprayed from the nozzle from data for estimating the flow velocity of the machining fluid flowing through the machining grooves formed in the workpiece by electrical discharge machining, the depth of the machining grooves, the number of parallel machining grooves, and the state of accumulation of machining debris in the machining grooves. 【0105】 According to the inference device 220 described above, an inference device is realized that includes a second data acquisition unit that acquires inference data including data for estimating the flow velocity of the machining fluid flowing through the machining grooves formed in the workpiece by electrical discharge machining, the depth of the machining grooves formed in the workpiece by electrical discharge machining, the number of parallel machining grooves which is the number of machining grooves arranged in parallel in the workpiece, and the state of accumulation of machining debris in the machining grooves, and an inference unit that uses a trained model to infer the flow rate of the machining fluid sprayed from the nozzle from the data for estimating the flow velocity of the machining fluid flowing through the machining grooves formed in the workpiece by electrical discharge machining, the depth of the machining grooves, the number of parallel machining grooves, and the state of accumulation of machining debris, and outputs a flow rate value of the machining fluid corresponding to the data, the depth of the machining grooves, the number of parallel machining grooves, and the state of accumulation of machining debris. 【0106】 The wire electrical discharge machining apparatus 100a according to Embodiment 2 can efficiently perform both electrode cooling between the cutting wire section 1a and the workpiece W, and discharge of machining debris generated in the machining groove Wz by electrical discharge machining, by controlling the flow rate of the machining fluid 9 sprayed from the nozzle 7 based on the estimated flow rate value of the machining fluid 9. 【0107】 Furthermore, in the wire electrical discharge machining apparatus 100a according to Embodiment 2, the effort required in the wire electrical discharge machining apparatus 100 according to Embodiment 1 to create a 3D model by fluid analysis for each electrical discharge machining condition is eliminated. 【0108】 Next, the hardware configurations of the control units 80 according to Embodiments 1 and 2 will be described. The control units 80 according to Embodiments 1 and 2 correspond to the calculation unit 51 and control unit 52 of the control devices 5 and 5a of the wire electrical discharge machining apparatuses 100 and 100a, and to the machine learning apparatus 200 of the control device 5a of the wire electrical discharge machining apparatus 100a, respectively. The functions of the control units 80 according to Embodiments 1 and 2 are realized by processing circuits. The processing circuits may be dedicated hardware or processing devices that execute programs stored in memory. 【0109】 When the processing circuit is dedicated hardware, the processing circuit may be a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an application-specific integrated circuit, a field-programmable gate array, or a combination thereof. Figure 16 shows a configuration in which each function of the control unit 80 according to Embodiments 1 and 2 is realized in hardware. The processing circuit 81 incorporates a logic circuit 81a that realizes the functions of the control unit 80. 【0110】 If the processing circuit 81 is a processing unit, the functions of the control unit 80 are realized by software, firmware, or a combination of software and firmware. 【0111】 Figure 17 shows a configuration in which each function of the control unit 80 according to Embodiments 1 and 2 is implemented by software. The processing circuit 81 includes a processor 811 that executes program 81b, a random access memory 812 used by the processor 811 as a work area, and a storage device 813 that stores program 81b. The processor 811 loads program 81b stored in the storage device 813 onto the random access memory 812 and executes it, thereby realizing the functions of the control unit 80. The software or firmware is written in a programming language and stored in the storage device 813. The processor 811 can be a central processing unit, but is not limited to that. The storage device 813 can be a semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), or EEPROM (Electrically Erasable Programmable Read Only Memory). The semiconductor memory may be non-volatile memory or volatile memory. Furthermore, the storage device 813 can be a magnetic disk, flexible disk, optical disk, compact disk, minidisc, or DVD (Digital Versatile Disc) in addition to semiconductor memory. The processor 811 may output data such as calculation results to the storage device 813 for storage, or it may store such data in an auxiliary storage device (not shown) via the random access memory 812. By integrating the processor 811, random access memory 812, and storage device 813 onto a single chip, the functions of the control unit 80 can be realized by a microcomputer. 【0112】 The processing circuit 81 realizes the functions of the control unit 80 by reading and executing the program 81b stored in the memory device 813. The program 81b can also be described as instructing the computer to execute the procedures and methods for realizing the functions of the control unit 80. 【0113】 Furthermore, the processing circuit 81 may implement some of the functions of the control unit 80 using dedicated hardware, and some of the functions of the control unit 80 using software or firmware. 【0114】 Thus, the processing circuit 81 can realize each of the above-mentioned functions through hardware, software, firmware, or a combination thereof. 【0115】 The configurations shown in the above embodiments are merely examples, and it is possible to combine them with other known technologies, combine different embodiments, and omit or modify parts of the configuration without departing from the gist of the invention. [Explanation of symbols] 【0116】 1 Wire electrode, 1a Cutting wire section, 2,2a,2b,2c,2d Guide rollers, 3,3a,3b Electron supply, 4 Moving device, 5,5a Control device, 6 Processing fluid guide section, 7,7a,7b Nozzle, 8 Workpiece holding device, 9 Processing fluid, 10 Measurement section, 11 Pressure sensor, 51 Calculation section, 52 Control section, 53 Storage section, 61,61a,61b Processing fluid rectifier plate, 62 Workpiece fixing plate, 63,63a,63b Processing fluid escape prevention plate, 64 Workpiece holding section, 71 Spray hole, 72 Processing fluid tank, 73 Pump, 74 Flow rate adjustment valve, 100,100a Wire EDM machine, 101 Arrow, 200 Machine learning device, 210 Learning device, 211,221 Data acquisition unit, 212 Model generation unit, 213 Trained model storage unit, 220 inference device, 222 inference unit, 233 trained model, 521 flow control unit, 631 through hole, A specific region, Th threshold, W workpiece, Wz machining groove.

Claims

[Claim 1] A wire electrode having cutting wire portions that are spaced apart from each other in parallel and facing the workpiece, A nozzle having an injection hole for supplying processing fluid to the gap between the cutting wire portion and the workpiece, An outer peripheral component member arranged on the outer circumference of the workpiece, A measuring unit connected to the interior of the space formed by the outer peripheral component, which acquires data for estimating the flow velocity of the machining fluid flowing through the machining groove formed in the workpiece by electrical discharge machining, A wire electrical discharge machining apparatus characterized by being equipped with the following features. [Claim 2] A calculation unit that estimates the flow velocity of the processing fluid flowing inside the processing groove from the data acquired by the measurement unit, A flow control unit that controls the flow rate of the processing fluid sprayed from the nozzle, Equipped with, The flow rate control unit controls the flow rate of the processing fluid sprayed from the nozzle based on the flow velocity value of the processing fluid estimated by the calculation unit. A wire electrical discharge machining apparatus according to claim 1, characterized by the following: [Claim 3] The flow rate control unit compares the flow velocity value of the processing fluid estimated by the calculation unit with a target flow velocity value, which is a target value for the flow velocity of the processing fluid, and controls the flow rate of the processing fluid sprayed from the nozzle so that the estimated flow velocity value of the processing fluid approaches the target flow velocity value. A wire electrical discharge machining apparatus according to claim 2, characterized by the following: [Claim 4] A calculation unit that estimates the flow velocity of the processing fluid flowing inside the processing groove from the data acquired by the measurement unit, A flow control unit that controls the flow rate of the processing fluid sprayed from the nozzle, Equipped with, The flow rate control unit controls the flow rate of the processing fluid sprayed from the nozzle based on a setting value set by the user to the flow rate control unit. A wire electrical discharge machining apparatus according to claim 1, characterized by the following: [Claim 5] A first data acquisition unit acquires learning data including the aforementioned data, the depth of the machining groove formed in the workpiece by electrical discharge machining, the number of parallel machining grooves in the workpiece, the state of debris accumulation in the machining grooves, and the flow rate value of the machining fluid sprayed from the nozzle. A model generation unit generates a trained model for inferring the flow rate value of the machining fluid ejected from the nozzle, using the aforementioned training data, the depth of the machining groove, the number of parallel machining grooves, and the state of debris accumulation in the machining groove. The learning device has the following features: A wire electrical discharge machining apparatus according to claim 1, characterized by the following: [Claim 6] A second data acquisition unit acquires inference data including the aforementioned data, the depth of the machining groove formed in the workpiece by electrical discharge machining, the number of parallel machining grooves in the workpiece, and the state of debris accumulation in the machining grooves. An inference unit that uses a trained model to infer the flow rate of the processing fluid sprayed from the nozzle based on the data, the depth of the processing groove, the number of parallel processing grooves, and the state of processing debris accumulation, and outputs the flow rate of the processing fluid corresponding to the data, the depth of the processing groove, the number of parallel processing grooves, and the state of processing debris accumulation, The inference device has the following features: A wire electrical discharge machining apparatus according to claim 1, characterized by the following: [Claim 7] The system includes a flow control unit that controls the flow rate of the processing fluid sprayed from the nozzle, The flow rate control unit controls the flow rate of the processing fluid sprayed from the nozzle using the flow rate value of the processing fluid output from the inference unit. A wire electrical discharge machining apparatus according to claim 6, characterized by the following: [Claim 8] The aforementioned data is the pressure in the space formed by the outer peripheral component. A wire electrical discharge machining apparatus according to any one of claims 1 to 5, characterized by the above. [Claim 9] A wire electrical discharge machining method in a wire electrical discharge machining apparatus, wherein a discharge is generated between a workpiece and a plurality of cutting wire sections that travel parallel to each other and spaced apart, and the workpiece is subjected to electrical discharge machining by the energy from the discharge, thereby simultaneously cutting a plurality of plate-shaped members from the workpiece, The steps include: running a plurality of cutting wires over the workpiece, whose outer circumference is surrounded by the outer peripheral component, to form a machining groove in the workpiece by electrical discharge machining; The steps include supplying processing fluid from a nozzle into the gap between the cutting wire portion and the workpiece, The steps include: acquiring data for estimating the flow velocity of the processing fluid flowing through the processing groove in the space formed by the outer peripheral component by a measuring unit connected to the inside of the space formed by the outer peripheral component; A wire electrical discharge machining method characterized by including the following. [Claim 10] The calculation unit estimates the flow velocity of the processing fluid flowing inside the processing groove from the data, The flow control unit controls the flow rate of the processing fluid sprayed from the nozzle based on the estimated flow velocity value of the processing fluid. The wire electrical discharge machining method according to claim 9, characterized by including the following: