Multi-wire electrical discharge machining (EDM) machine, multi-wire EDM system

Abstract

A multi-wire electrical discharge machine (1) comprises: a machining power supply (7) for applying a machining pulse voltage between a workpiece (2) and a wire cutting section (6a); and a wire breakage detection section (20) for detecting a wire breakage of a wire electrode (6). The wire breakage detection section (20) is electrically connected in parallel to a wire electrode section extending from a portion of the wire electrode (6) on the extraction side to a portion of the wire electrode (6) on the discharge side, and comprises: a wire breakage detection circuit (206) which is an electrical circuit in which a wire breakage detection power supply (204) for supplying a direct current to the wire electrode section at least during a cutting process of the workpiece (2) and a current limiting resistor (205) are connected in series; and a wire breakage determination section (207) for determining a wire breakage based on a current flowing in the current limiting resistor (205).

Description

Technical Field

[0001] The present invention relates to a multi-wire electrical discharge machine, and more particularly to a multi-wire electrical discharge machine having a cutting wire section composed of a plurality of wire electrode sections (hereinafter referred to as cutting wires) which are stretched in parallel by winding a wire electrode between a plurality of guide rollers and arranged in parallel at positions opposing a workpiece, and wherein a pulse voltage is applied between the cutting wire section and the workpiece to cause discharge, thereby cutting the workpiece into a plurality of pieces. Background Art

[0002] As a method for cutting chip-shaped thin plates from a columnar workpiece such as a semiconductor ingot, the following multi-wire electrical discharge machining technology has been disclosed, namely, by winding a wire electrode between a plurality of guide rollers, a cutting line portion consisting of a plurality of cutting lines is formed relative to the columnar workpiece, and a pulse voltage is independently applied between the cutting line portion and the columnar workpiece to cause discharge, thereby cutting a plurality of thin plates from the workpiece at the same time (for example, refer to patent document 1).

[0003] In electrical machining such as multi-wire electrical discharge machining, when a wire electrode is broken (hereinafter referred to as wire break), it is important to immediately stop the machining process in order to prevent damage to the workpiece or the machine. Therefore, when a wire break occurs, it is required to be able to detect the wire break at high speed and high precision. Patent document 1 proposes a method in which a pulse voltage for wire break detection is applied to the wire electrode only during a period in which a pulse voltage is not applied by a machining power supply, and the conductive state of the wire electrode is confirmed, thereby detecting whether there is a wire break.

[0004] Patent Document 1: Japanese Patent Application Publication No. 2009-184071 Summary of the invention

[0005] In the conventional multi-wire electrical discharge machine, a single wire electrode is wound between a plurality of guide rollers. Therefore, if the number of windings increases, the inductance of the wire electrode increases, and it becomes difficult to flow a high-frequency current through the wire electrode. Therefore, in the method of applying a wire break detection pulse voltage to the wire electrode and checking the conductive state of the wire electrode as in Patent Document 1, there is a problem that if the number of windings of the wire electrode increases, wire breakage cannot be detected.

[0006] The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to obtain a multi-wire electric discharge machine capable of detecting wire breakage at high speed and with high accuracy.

[0007] In order to solve the above-mentioned problems and achieve the purpose, a multi-wire electrical discharge machine of the present invention has a cutting wire section composed of a plurality of cutting wires which are stretched in parallel by winding a wire electrode between a plurality of guide rollers and are arranged in parallel at a position opposing a workpiece. The multi-wire electrical discharge machine generates discharge between the cutting wire section and the workpiece, thereby cutting the workpiece, and is characterized in that it has: a machining power supply which applies a machining pulse voltage between the workpiece and the cutting wire section; and a wire breakage detection section which detects wire breakage of the wire electrode. The wire breakage detection section is electrically connected in parallel to a wire electrode section stretched in a section from a position on the extraction side of the wire electrode to a position on the discharge side of the wire electrode, and has: a wire breakage detection circuit which is an electrical circuit formed by connecting in series a wire breakage detection power supply and a current limiting resistor which always supply a direct current to the wire electrode section at least during the cutting process of the workpiece; and a wire breakage determination section which determines wire breakage of the wire electrode based on the current flowing in the current limiting resistor.

[0008] Effects of the Invention

[0009] The multi-wire electrical discharge machine according to the present invention has an effect of being able to detect wire breakage at high speed and high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 It is a perspective view showing the main structure of the multi-wire electrical discharge machine according to the first embodiment.

[0011] Figure 2 This is a diagram showing the relationship between the wire electrode portion between the connection target 20a and the connection target 20b of the wire breakage detection unit in the electric discharge machining of the multi-wire electric discharge machine according to the first embodiment, using an equivalent circuit.

[0012] Figure 3 This is a circuit diagram showing a configuration of a wire breakage detection unit of the multi-wire electrical discharge machine according to the first embodiment.

[0013] Figure 4 This is a block diagram showing the hardware configuration of the processing control device according to the first embodiment.

[0014] Figure 5 This is a circuit diagram showing a configuration of a wire breakage detection unit of a multi-wire electrical discharge machine according to the second embodiment.

[0015] Figure 6 This is a circuit diagram showing a configuration of a wire breakage detection unit of a multi-wire electrical discharge machine according to a third embodiment.

[0016] Figure 7This is a diagram for explaining the response of the binarization circuit before and after the line breakage according to the third embodiment.

[0017] Figure 8 This is a diagram for explaining the response of the low-pass characteristic portion before and after the line breakage according to the third embodiment.

[0018] Fig. 9 This is a circuit diagram showing a configuration of a wire breakage detection unit of a multi-wire electrical discharge machine according to a fourth embodiment.

[0019] Fig.10 This is a diagram for explaining the response of the frequency detection unit before and after the line breakage according to the fourth embodiment.

[0020] Fig.11 It is a perspective view showing the main structure of the multi-wire electrical discharge machine involved in the fifth embodiment.

[0021] Fig.12 This is a diagram showing the structure of a machine learning device according to the fifth embodiment.

[0022] Fig.13 This is a diagram showing an example of a three-layer neural network involved in implementation mode 5.

[0023] Fig.14 This is a flowchart related to the learning process of the machine learning device involved in Implementation Example 5.

[0024] Fig.15 It is a configuration diagram of a wire breakage prediction unit related to a multi-wire electrical discharge machine according to a fifth embodiment.

[0025] Fig.16 It is a diagram showing the configuration of a multi-wire electrical discharge machining system according to a fifth embodiment.

[0026] Fig.17 It is a diagram showing another configuration of the multi-wire electrical discharge machining system according to the fifth embodiment.

[0027] Fig.18 This is a flowchart related to processing for obtaining prediction information using the line disconnection prediction unit according to the fifth embodiment. DETAILED DESCRIPTION

[0028] Hereinafter, a multi-wire electrical discharge machine according to an embodiment of the present invention will be described in detail with reference to the drawings. However, the present invention is not limited to the embodiment.

[0029] Implementation method 1.

[0030] Figure 1 It is a perspective view showing the main structure of the multi-wire electrical discharge machine in the first embodiment of the present invention.

[0031] The multi-wire electrical discharge machine 1 involved in the first embodiment includes: a wire reel 4 that supplies a wire electrode 6; a wire reel driving unit 17 that drives the wire reel 4; a wire discharge roller 5 that discharges the wire electrode 6 to the outside of the multi-wire electrical discharge machine 1; guide rollers 3a, 3b, 3c, and 3d, which are used to make the wire electrode 6 move appropriately; a guide roller driving unit 18 that drives the guide roller 3a; a power supply 12 that contacts the wire electrode 6 and supplies a pulse voltage for processing (hereinafter referred to as a processing pulse voltage); a driving unit 14 that drives a processing table 70 capable of placing a workpiece 2 in a vertical direction; a processing power supply 7 that applies a processing pulse voltage between the cutting wire portion 6a and the workpiece 2 via the power supply 12; a processing control device 9 that controls electrical discharge processing; a wire breakage detection unit 20 that detects the breakage of the wire electrode; and a processing state detection device 15 that detects the processing state in the cutting wire portion 6a. Furthermore, the multi-wire electrical discharge machine 1 has an orthogonal coordinate system in which the vertical direction is defined as the Z-axis direction, and two directions orthogonal to each other in the horizontal plane are defined as the X-axis direction and the Y-axis direction.

[0032] The four guide rollers 3a to 3d are arranged in parallel with the axial direction and separated from each other. In addition, in a plane orthogonal to the axial direction, the positions of the axes are arranged in a quadrilateral. Specifically, among the four guide rollers 3a to 3d, the guide rollers 3a and 3b are arranged at the highest position, the guide roller 3c is arranged below the guide roller 3b, and the guide roller 3d is arranged below the guide roller 3a and aligned with the guide roller 3c. In addition, on the outer periphery of the four guide rollers 3a to 3d, a plurality of guide grooves for guiding the travel of the wire electrode 6 are formed at intervals in the respective axial directions.

[0033] As described above, one wire electrode 6 drawn out from the bobbin 4 is guided by the guide grooves of the guide rollers 3a, 3b, 3c, and 3d that rotate clockwise in the example shown in the figure, and after being wound multiple times between the four guide rollers 3a, 3b, 3c, and 3d with the intervals between the guide grooves being spaced apart from each other, it is discharged to the outside of the multi-wire electrical discharge machine 1 through the wire discharge roller 5. Here, the portion of the wire electrode 6 that is stretched parallel to the guide roller 3c and the guide roller 3d is the cut wire portion 6a.

[0034] In the example shown in the figure, the guide roller 3a is driven by the guide roller driving unit 18, and the bobbin driving unit 17 drives the bobbin 4 so that the tension of the wire electrode 6 becomes constant. In addition, the guide roller driving unit 18 and the bobbin driving unit 17 drive the guide roller 3a and the bobbin 4, respectively, so that the desired wire travel speed is achieved.

[0035] Regarding the workpiece 2, a position command is output from the processing control device 9 to the drive unit 14, and the processing table 70 is driven in the Z-axis direction, thereby relatively changing the positional relationship between the workpiece 2 and the cutting line portion 6a. Therefore, by the position command output from the processing control device 9, the workpiece 2 is moved in a manner close to the cutting line portion 6a, and a pulse voltage is applied between the cutting line portion 6a and the workpiece 2 to generate discharge, thereby cutting a plurality of thin plates from the workpiece 2. In addition, the processing control device 9 performs feed control in the cutting direction in a manner that maintains an appropriate inter-electrode distance between the workpiece 2 and the cutting line portion 6a (hereinafter referred to as inter-electrode).

[0036] In addition, an ingot-shaped material to be cut into a plurality of thin plates can be used as the workpiece 2. The material of the workpiece 2 includes, for example, metals such as tungsten and molybdenum that are sputtering targets, ceramics such as polycrystalline silicon carbide that are used as various structural components, semiconductor materials such as single crystal silicon, single crystal silicon carbide, single crystal gallium nitride, single crystal gallium oxide, and single crystal diamond that are semiconductor device wafers, and solar cell materials such as single crystal and polycrystalline silicon that are solar cell wafers.

[0037] The metals mentioned above do not hinder the application of electrical discharge machining, but the materials that can be used for electrical discharge machining through semiconductor materials and solar cell materials are materials with a relative resistance of approximately less than or equal to 100Ωcm and sufficiently low, and more preferably less than or equal to 10Ωcm. Therefore, as the workpiece 2, metals or materials with a relative resistance equivalent to metal to a range of less than or equal to 100Ωcm, preferably less than or equal to 10Ωcm, especially semiconductor materials and solar cell materials having a relative resistance in the above range are suitable.

[0038] Then, machining fluid is supplied to the gap between the workpiece 2 and the cutting wire portion 6 a by blowing or dipping, similarly to a so-called single-type wire electrical discharge machine.

[0039] The machining power supply 7 generates a machining pulse voltage applied to the inter-electrode of the workpiece 2 and the cutting line portion 6a based on a machining voltage application instruction from the machining control device 9. The machining pulse voltage is generated, for example, by a switching power supply method. In addition, the machining power supply 7 is composed of a plurality of machining power supply units 8 that can apply voltages independently of each other. Moreover, a grounding electrode 10 is provided on the machining power supply 7 across the plurality of machining power supply units 8. The grounding wires of the plurality of machining power supply units 8 are connected to the grounding electrode 10, and are connected to the workpiece 2 through a cable 11. In addition, the polarity of the generated machining pulse voltage can be appropriately reversed as needed.

[0040] The power supply 12 is composed of a plurality of power supply units 13 insulated from each other. In the example shown in the figure, the wire electrode portion stretched in parallel between the guide rollers 3b and 3c becomes the power supply line portion 6b to which the machining pulse voltage is supplied, and each power supply unit 13 is configured in a one-to-one sliding contact relationship with each wire electrode of the power supply line portion 6b. Moreover, each power supply unit 13 independently supplies the machining pulse voltage generated by each machining power supply unit 8 to each wire electrode of the power supply line portion 6b. Therefore, the machining pulse voltage is independently applied to each cutting wire of the cutting wire portion 6a from each machining power supply unit 8.

[0041] The machining state detection device 15 is a device for detecting the machining state in the cutting line portion 6a. The machining state detection device 15 is provided on the wiring 19 connecting the power supply 12 and the machining power supply 7, and monitors the voltage state (hereinafter referred to as the inter-electrode voltage) applied to the workpiece 2 and the cutting line portion 6a through the power supply 12, the number of discharges per unit time, the number of short circuits per unit time, and the output voltage of the machining power supply 7, that is, the machining pulse voltage, thereby detecting the machining state in the cutting line portion 6a. In addition, the machining state detection device 15 is composed of a plurality of machining state detection units 16 in a one-to-one correspondence with each cutting line of the cutting line portion 6a, and detects the machining state of the cutting line portion 6a for each cutting line pair.

[0042] One connection target 20a of the wire break detection unit 20 is connected to the wire electrode 6 between the bobbin 4 and the power supply unit 13 where the wire electrode 6 drawn out from the bobbin 4 first comes into sliding contact via the wire break detection power supply 20d. In addition, another connection target 20b of the wire break detection unit 20 comes into sliding contact with the wire discharge roller 5. That is, the wire break detection unit 20 is electrically connected in parallel to the wire electrode portion stretched from a portion on the drawing side of the wire electrode 6 to a portion on the discharge side of the wire electrode 6. Here, the wire discharge roller 5 is made of a conductive material, and the connection target 20b of the wire break detection unit 20 and the wire electrode 6 on the discharge side are electrically connected via the wire discharge roller 5.

[0043] Next, the configuration of the wire breakage detection unit 20 and the principle of a method for detecting a wire breakage will be described. Figure 2This is a diagram showing the relationship between the wire electrode portion between the connection target 20a and the connection target 20b of the wire break detection unit 20 in the multi-wire electrical discharge machining and the wire break detection unit 20 by an equivalent circuit. In the multi-wire electrical discharge machining, discharge occurs at the inter-electrode portion between the workpiece 2 and the cutting wire portion 6a with probability according to the state of the inter-electrode portion (for example, the relative resistance of the machining fluid, the local temperature of the workpiece 2 and the cutting wire, the amplitude of the machining pulse voltage, etc.). That is, the voltage generated in the wire electrode portion between the connection target 20a and the connection target 20b of the wire break detection unit 20 (that is, the input signal to the wire break detection unit 20) can be regarded as an analog voltage source 60 that randomly changes according to the machining pulse voltage and the state of the discharge. That is, the wire electrode portion between the connection target 20a and the connection target 20b of the wire break detection unit 20 generates a high-frequency voltage that randomly changes according to the machining pulse voltage and the state of the discharge. Therefore, the relationship between the wire electrode portion between the connection target 20a and the connection target 20b of the wire breakage detection unit 20 in the multi-wire electrical discharge machining can be determined by Figure 2 The equivalent circuit representation of .

[0044] Figure 3 1 is a circuit diagram showing the structure of the wire disconnection detection unit 20 of the present embodiment. The wire disconnection detection unit 20 has a wire disconnection detection circuit 206 which is an electrical circuit formed by connecting in series a wire disconnection detection power supply 204 for supplying a DC current to a wire electrode portion between a connection target 20a and a connection target 20b of the wire disconnection detection unit 20 and a current limiting resistor 205.

[0045] The wire break detection power supply 204 applies a DC voltage of a value lower than the machining pulse voltage applied by the machining power supply 7 as a wire break detection voltage, and always applies it at least during the cutting process of the workpiece 2, and supplies a DC current to the wire electrode portion between the connection target 20a and the connection target 20b of the wire break detection unit 20. That is, at least during the cutting process of the workpiece 2, a DC current is always supplied to the wire electrode portion between the connection target 20a and the connection target 20b of the wire break detection unit 20, but the DC current for wire break detection is about several mA to several hundred mA, which is sufficiently smaller than the machining current of approximately several A to several hundred A, and therefore has no effect on machining. The current limiting resistor 205 limits the current flowing in the wire break detection circuit 206. That is, the current limiting resistor 205 limits the DC current supplied from the wire break detection power supply 204 to the wire electrode portion between the connection target 20a and the connection target 20b of the wire break detection unit 20, and limits the current output from the analog voltage source 60 to the wire break detection unit 20. In addition, the wire break detection unit 20 has a wire break determination unit 207 that determines whether the wire is broken based on the current flowing in the current limiting resistor 205. The wire break determination unit 207 outputs a wire break signal indicating the determination result of the wire break to the processing control device 9, and notifies the detection result of the wire break.

[0046] Next, a specific method of detecting a wire breakage by the wire breakage detection unit 20 will be described. Figure 3 In the circuit shown, the current flowing through the current limiting resistor 205 is a current obtained by superimposing the direct current output from the wire break detection power supply 204 and the high-frequency current output from the analog voltage source 60. When the wire electrode 6 is in a non-broken state, the wire break detection power supply 204 is a direct current voltage source, so a direct current flows through the current limiting resistor 205. In addition, since the analog voltage source 60 inputs a random voltage signal to the wire break detection unit 20, the current flowing through the current limiting resistor 205 by the analog voltage source 60 becomes a high-frequency current.

[0047] Therefore, the current detected by the wire break determination unit 207 becomes a current obtained by superimposing the direct current output by the wire break detection power supply 204 and the high-frequency current output by the analog voltage source 60. On the other hand, when the wire electrode 6 is in a broken state, the connection targets 20a and 20b of the wire break detection unit 20 are not electrically connected. Therefore, a current does not flow in the current limiting resistor 205. At this time, the wire break determination unit 207 determines that the wire is broken when it can detect that the current does not flow in the current limiting resistor 205. As described above, by successively monitoring the current flowing in the current limiting resistor 205, it is possible to detect a wire break.

[0048] In addition, the current limiting resistor 205 can also be used as the current limiting resistor 205 by sufficiently extending the length of the wire electrode 6 connected to the connection targets 20a and 20b of the wire break detection unit 20. If the wire break is detected by successively monitoring the current flowing in the current limiting resistor 205, its circuit structure is not limited.

[0049] Next, a description will be given of the hardware configuration of the machining control device 9 included in the multi-wire electrical discharge machine 1 according to Embodiment 1. Each function of the machining control device 9 is realized by using a computer system such as a personal computer or a general-purpose computer. Figure 4 This is a diagram showing an example of a hardware configuration when the function of the process control device 9 according to the first embodiment is realized using a computer system.

[0050] The processing control device 9 has: a processor 91 that performs various processes; a memory 92 as a built-in memory; an external storage device 93 that stores various information; an input unit 94 that obtains information input by an operator; and an output unit 95 that prompts the operator with various information.

[0051] The processor 91 is a CPU (Central Processing Unit). The processor 91 may be a processing device, a computing device, a microprocessor, a microcomputer, or a DSP (Digital Signal Processor). The functions of the processing control device 9 are implemented by the processor 91 and software, firmware, or a combination of software and firmware. The software or firmware is described as a program and stored in the external storage device 93. The processor 91 reads the software or firmware stored in the external storage device 93 to the memory 92 and executes it.

[0052] The memory 92 is a nonvolatile or volatile semiconductor memory, such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), or EEPROM (registered trademark) (Electrically Erasable Programmable Read Only Memory). The external storage device 93 is a HDD (Hard Disk Drive) or SSD (Solid State Drive).

[0053] The input unit 94 includes input devices represented by input keys and pointing devices. The input unit 94 obtains information input by the operator of the multi-wire electrical discharge machine 1 and notifies the processor 91 of the obtained information. The output unit 95 includes output devices represented by LCD (Liquid Crystal Display) and speakers. The output unit 95 and the pointing device constituting the input unit 94 are integrally formed into a touch screen. The output unit 95 presents various information to the operator according to the instructions of the processor 91.

[0054] As described above, according to the present embodiment, the multi-wire electrical discharge machine 1 applies the DC voltage output from the wire break detection power supply 204 to the wire electrode 6 at all times during at least the cutting process of the workpiece 2, and checks the conduction state of the wire electrode 6, thereby not affecting the inductance generated by winding one wire electrode 6 between a plurality of guide rollers, and can detect wire breakage with high accuracy. In addition, the DC voltage output from the wire break detection power supply 204 is applied to the wire electrode 6 not only during the period when the machining pulse voltage is not applied by the machining power supply 7, but also during the multi-wire electrical discharge machining, and the conduction state of the wire is checked by superimposing the high-frequency current input from the wire electrode 6 to the wire break detection unit 20 and the DC current supplied from the wire break detection power supply 204, thereby detecting wire breakage at high speed.

[0055] Implementation method 2.

[0056] In implementation mode 1, by Figure 3 The structure shown in the figure is used to explain the structure of the wire breakage detection unit 20, but the structure of the wire breakage detection unit 20 may also be Figure 5 The structure shown. Figure 5 The structure of Figure 3In the structure of, there are added rectifier circuits 201 and 202 for rectifying the high-frequency voltage generated in the wire electrode portion between the connection target 20a and the connection target 20b of the wire disconnection detection unit 20 according to the state of the machining pulse voltage and the discharge, and a smoothing circuit 203 for smoothing the current rectified by the rectifier circuits 201 and 202, and the other parts are the same structure. The rectifier circuits 201 and 202 can be any one as long as they can rectify the high-frequency voltage, and for example, they are composed of rectifier diodes 201a and 202a. The smoothing circuit 203 can be any one as long as it can smooth the current rectified by the rectifier circuits 201 and 202, and for example, it is composed of a circuit in which a high-frequency disconnection inductor 203a and a damping diode 203b are connected in parallel. With the above structure, a current obtained by superimposing the direct current output by the wire disconnection detection power supply 204 and the current obtained by smoothing the high-frequency current output by the analog voltage source 60 flows through the current limiting resistor 205. The wire breakage determination unit 207 detects the current and checks the conductive state of the wire in the same manner as in the first embodiment, thereby detecting wire breakage.

[0057] Implementation method 3.

[0058] When the conduction state of the wire electrode 6 is confirmed by the current flowing to the current limiting resistor 205, the current flowing to the current limiting resistor 205 vibrates immediately after the wire breakage occurs, so in order to accurately detect the wire breakage, it is necessary to wait until the current stabilizes. In the third embodiment, a method for detecting the wire breakage at a high speed is described. The description of the same parts as the first embodiment is omitted, and the parts different from the first embodiment are described.

[0059] Figure 6 : This is a circuit diagram showing the structure of the wire breakage detection unit 30 of the multi-wire electrical discharge machine involved in the third embodiment. The wire breakage determination unit 305 has: a binary circuit 304, which converts the current flowing through the current limiting resistor 205 into a binary signal of a low level and a high level; and a signal processing circuit 303, which outputs a wire breakage signal based on the binary signal. The binary circuit 304 can be realized by a structure combining a photocoupler 304a and a pull-up resistor 304b as shown in the figure as an example, and outputs a low-level analog signal if current flows through the current limiting resistor 205, and outputs a high-level analog signal if current does not flow through the current limiting resistor 205, thereby converting the current flowing through the current limiting resistor 205 into a binary signal.

[0060] The signal processing circuit 303 outputs a wire break signal based on the result of detecting whether the signal output from the binarization circuit 304 is a high level signal or a low level signal. As described above, by monitoring the wire break signal, it is possible to detect the wire break. In the multi-wire electrical discharge machine according to the third embodiment, in order to detect whether the analog signal output from the binarization circuit 304 is a high level signal or a low level signal, the signal processing circuit 303 includes: a low-pass characteristic section 301, which is a circuit having a low-pass characteristic; and a wire break signal output section 302, which outputs the wire break signal based on the signal output from the low-pass characteristic section 301. As the structure of the low-pass characteristic section 301, there are CR type low-pass filters, active low-pass filters using operational amplifiers, digital filters using D flip-flops, FIR (Finite Impulse Response) filters, etc., but as long as it is a circuit having a low-pass characteristic, its circuit structure is not limited.

[0061] Figure 7 This is a diagram for explaining the response of the binary circuit 304 before and after the line break. Before the line break (between time a and b), the connection targets 20a and 20b of the line break detection unit 30 are in a conductive state through the line electrode 6, and current flows to the current limiting resistor 205, so the output of the binary circuit 304 indicates a Low level. Immediately after the line break occurs (between time b and c), the output current of the line break detection power supply 204 becomes zero, but there is a current that flows into the line break detection unit 30 in an unsteady state until the output current of the analog voltage source 60 converges to zero, so the binary signal vibrates until it moves to a High level. Figure 8 This is a diagram for explaining the response of the low-pass characteristic section 301 before and after the line is disconnected. If a binary signal that vibrates is input to the low-pass characteristic section 301, the vibration component is removed by the low-pass filter, and the signal becomes Figure 8 When the signal after passing through the low-pass characteristic unit 301 is greater than or equal to a preset threshold, the line disconnection signal output unit 302 determines that the line is disconnected and outputs a line disconnection signal. That is, the signal processing circuit 303 outputs a line disconnection signal based on the comparison result of the signal after the binary signal is passed through the low-pass filter and the predetermined threshold.

[0062] As described above, according to this embodiment, the line break determination unit 305 is constructed to have a binary circuit 304 that converts a binary signal into a Low level and a High level based on the presence or absence of current flowing in the current limiting resistor 205, and a signal processing circuit 303 with a low-pass characteristic. Thus, even if the current flowing to the current limiting resistor 205 that is monitored for confirming the conduction state vibrates right after the line break occurs, the line break can be detected at a high speed.

[0063] Implementation method 4.

[0064] In the third embodiment, the method of detecting the line breakage at high speed is described in that the line breakage determination unit 305 is configured to include the binary circuit 304 and the signal processing circuit 303 having a low-pass characteristic, thereby enabling the line breakage to be detected at high speed even if the current flowing to the current limiting resistor 205 is vibrating immediately after the line breakage occurs. However, in the fourth embodiment, a method of detecting the line breakage at high speed is described even if the signal processing circuit has a structure that does not have a low-pass characteristic. The description of the same parts as those in the third embodiment is omitted, and the parts different from the third embodiment are described.

[0065] Fig. 9 1 is a circuit diagram showing the structure of the wire breakage detection unit 40 of the multi-wire electrical discharge machine according to the fourth embodiment. The signal processing circuit 403 is different from the third embodiment, but has the same structure as the other parts. The signal processing circuit 403 includes a frequency detection unit 401 for detecting the frequency of the signal output from the binarization circuit 304, and a wire breakage signal output unit 402 for outputting a wire breakage signal based on the detected frequency. As described above, the binarized signal immediately after the wire breakage occurs becomes Figure 7 A waveform that vibrates periodically and repeatedly as shown. Fig.10 1 is a diagram for explaining the response of the frequency detection unit 401 before and after the wire is disconnected. If a binary signal that periodically repeats vibration is input to the frequency detection unit 401, the frequency detection unit 401 detects the frequency of the binary signal and converts it into Fig.10 The detection result shown is output to the line disconnection signal output unit 402. If the frequency detected by the line disconnection signal output unit 402 is higher than a predetermined threshold value (e.g., several tens of Hz to several kHz), it is determined that the line is disconnected and a line disconnection signal is output. That is, the signal processing circuit 403 detects the frequency of the binary signal and outputs a line disconnection signal based on the comparison result with the predetermined threshold value.

[0066] As described above, according to this embodiment, the frequency of the binary signal is detected to determine the line break, thereby enabling high-speed detection of the line break even if the current flowing to the current limiting resistor 205 monitored for confirmation of the conduction state vibrates immediately after the line break occurs.

[0067] Implementation method 5.

[0068] In Embodiments 1 to 4, a method for detecting a wire break when a wire break occurs is described, but in Embodiment 5, a method for predicting a wire break in advance using a result learned by a machine learning device and suppressing the wire break is described. The description of the same parts as in Embodiments 1 to 4 is omitted, and the parts different from Embodiments 1 to 4 are described.

[0069] Fig.11 This is a perspective view showing the main structure of a multi-wire electrical discharge machine 1a according to Embodiment 5. The main structure of the multi-wire electrical discharge machine 1 in Embodiment 1 is supplemented with a wire breakage prediction unit 21 and a machine learning device 22.

[0070] Next, the operations of the line disconnection prediction unit 21 and the machine learning device 22 will be described by dividing into a learning phase and an effective use phase.

[0071] <Learning Stage>

[0072] Fig.12 1 is a diagram showing a configuration of a machine learning device 22 according to Embodiment 5. The machine learning device 22 includes a data acquisition unit 221 , a state observation unit 222 , a model generation unit 223 , and a trained model storage unit 224 .

[0073] The data acquisition unit 221 acquires the wire break detection result output by the wire break detection unit 20 via the processing control device 9, and the state observation unit 222 acquires information indicating the processing state during processing, that is, state data, via the processing control device 9. As an example of the state data, there are the relative distance between the workpiece 2 and the cutting wire portion 6a, the wire travel speed, the wire tension, and a plurality of output results output by the processing state detection device 15.

[0074] The model generation unit 223 learns information indicating the possibility of wire breakage of the wire electrode 6 (hereinafter referred to as predicted information) according to the learning data created based on the combination of the wire breakage detection result output from the data acquisition unit 221 and the state data output from the state observation unit 222. That is, a trained model is generated that infers the best predicted information based on the wire breakage detection result and the state data. Here, the learning data is data that associates the wire breakage detection result and the state data with each other.

[0075] In addition, the machine learning device 22 uses the pre-known information for learning, but it can also be a device connected to the multi-wire electrical discharge machine via a network and separated from the multi-wire electrical discharge machine. In addition, the machine learning device 22 can also be built into the multi-wire electrical discharge machine. In addition, the machine learning device 22 can also exist on a cloud server. Fig.16 1 is a diagram showing the configuration of a multi-wire electrical discharge machining system 500 including a multi-wire electrical discharge machining machine 1b and a machine learning device 22b. Fig.11 The difference in the structure is that a machine learning device 22b is provided outside the multi-wire electrical discharge machine 1b. The machine learning device 22b obtains the wire break detection result and state data from the multi-wire electrical discharge machine 1b as learning data, and generates a trained model through machine learning. The wire break prediction unit 21 of the multi-wire electrical discharge machine 1b obtains the trained model from the machine learning device 22b, and outputs prediction information using the trained model.

[0076] return Fig.12 The learning algorithm used by the model generation unit 223 can use a known algorithm of tutor learning. As an example, the case where a neural network is applied is described. The model generation unit 223 learns the predicted information through so-called tutor learning according to the neural network model, for example. Here, tutor learning is a method of learning the features existing in these learning data by giving a set of input and result (label) data to a learning device, and inferring the result based on the input.

[0077] The neural network is composed of an input layer composed of multiple neurons, an intermediate layer (hidden layer) composed of multiple neurons, and an output layer composed of multiple neurons. The intermediate layer can be 1 layer or more than or equal to 2 layers.

[0078] For example, if it is Fig.13 In the three-layer neural network shown in FIG. 1 , if multiple inputs are input to the input layer (X1-X3), the values ​​are multiplied by the weight W1 (w11-w16) and input to the intermediate layer (Y1-Y2), and the result is further multiplied by the weight W2 (w21-w26) and output from the output layer (Z1-Z3). The output result changes according to the values ​​of the weights W1 and W2.

[0079] In the present application, the neural network learns the predicted information by so-called supervised learning according to the learning data created based on the combination of the wire break detection result acquired by the data acquisition unit 221 and the state data acquired from the state observation unit 222 .

[0080] That is, the neural network learns by adjusting the weights W1 and W2 so that the state data is input to the input layer and the result output from the output layer is close to the detection result of the wire breakage.

[0081] The model generation unit 223 generates a trained model by executing the above learning and outputs it.

[0082] The trained model storage unit 224 stores the trained model output from the model generation unit 223 .

[0083] Next, use Fig.14 , the learning process performed by the machine learning device 22 is explained. Fig.14 This is a flowchart related to the learning process of the machine learning device 22.

[0084] In step b1, the data acquisition unit 221 acquires the detection result of the wire breakage, and the state observation unit 222 acquires the state data. In addition, the detection result of the wire breakage and the state data are acquired simultaneously, but as long as the detection result of the wire breakage and the state data can be associated and input, the detection result of the wire breakage and the state data can also be acquired at different timings.

[0085] In step b2, the model generation unit 223 generates a trained model by learning the predicted information through so-called supervised learning according to the learning data created based on the combination of the wire break detection result obtained by the data acquisition unit 221 and the state data obtained by the state observation unit 222.

[0086] In step b3 , the trained model storage unit 224 stores the trained model generated by the model generation unit 223 .

[0087] <Effective use stage>

[0088] Fig.15 1 is a block diagram of a wire breakage prediction unit 21 related to a multi-wire electrical discharge machine. The wire breakage prediction unit 21 includes a state observation unit 211 and an estimation unit 212 .

[0089] The state observation unit 211 acquires state data.

[0090] The inference unit 212 infers the prediction information obtained by using the trained model. That is, by inputting the state data obtained by the state observation unit 211 to the trained model, the prediction information inferred from the state data can be output.

[0091] In addition, in the present embodiment, the prediction information is output using the trained model learned by the model generation unit 223 of the multi-wire electrical discharge machine 1a, but the trained model may be obtained from outside such as another multi-wire electrical discharge machine, and the prediction information may be output based on the trained model.

[0092] Next, use Fig.18 , a process for obtaining prediction information using the line disconnection prediction unit 21 will be described.

[0093] In step c1, the state observation unit 211 acquires state data.

[0094] In step c2 , the line disconnection prediction unit 21 inputs the state data to the trained model stored in the trained model storage unit 224 , and acquires prediction information.

[0095] In step c3 , the inference unit 212 outputs the prediction information obtained by the trained model to the process control device 9 .

[0096] In step c4, the processing control device 9 uses the outputted predicted information to control the electrical discharge machining so as to suppress the wire electrode from breaking. When the predicted information indicates that there is a possibility of wire breaking, the processing control device 9 issues an instruction to the processing power supply 7 in a manner of reducing the processing energy. That is, the processing control device 9 controls the processing power supply 7 in a manner lower than the current processing energy. In addition, the processing control device 9 may also output an instruction to reduce the wire tension or an instruction to increase the wire travel speed to the guide roller drive unit 18 and the bobbin drive unit 17. That is, the processing control device 9 controls the guide roller drive unit and the bobbin drive unit in a manner of reducing the current wire tension or in a manner of speeding up the current wire travel speed. In this way, wire breaking in multi-wire electrical discharge machining is suppressed, and the workpiece 2 can be processed stably.

[0097] In addition, in this embodiment, although the case where the tutor learning is applied to the learning algorithm used by the model generation unit 223 is described, the present invention is not limited to this. As for the learning algorithm, in addition to the tutor learning, semi-tutor learning and the like can also be applied.

[0098] In addition, the model generation unit 223 can learn the predicted information according to the learning data created for multiple multi-wire EDMs. In addition, the model generation unit 223 can also obtain learning data from multiple multi-wire EDMs used in the same area, and can also learn the predicted information using learning data collected from multiple multi-wire EDMs operating independently in different areas. In addition, the multi-wire EDM that collects the learning data can be added to or removed from the object midway. Furthermore, the machine learning device that has learned the predicted information about a certain multi-wire EDM can be applied to different other multi-wire EDMs, and the predicted information about the other multi-wire EDMs can be re-learned and updated.

[0099] In addition, as a learning algorithm used in the model generation unit 223, deep learning that learns the extraction of feature quantities themselves can also be used, and machine learning can also be performed according to other well-known methods such as genetic programming, functional logic programming, support vector machines, etc.

[0100] In the above example, the prediction information is calculated inside the multi-wire electrical discharge machines 1a and 1b, but the prediction information may be generated in a wire break prediction device different from the multi-wire electrical discharge machines 1a and 1b. Fig.17 1 is a diagram showing the structure of a multi-wire electrical discharge machining system 600 having a multi-wire electrical discharge machine 1c and a wire break prediction device 23. The wire break prediction device 23 has the same wire break prediction unit 21 and machine learning device 22 as described above. In the learning phase, the wire break prediction device 23 obtains the wire break detection result and state data from the multi-wire electrical discharge machine 1c as learning data, and generates a trained model through the same processing as described in the above-mentioned <learning phase>. In the effective use phase, the state data is newly obtained from the multi-wire electrical discharge machine 1c, and the prediction information is calculated by inputting it into the trained model generated in the learning phase, and outputs it to the processing control device 9 of the multi-wire electrical discharge machine 1c. The processing control device 9 uses the obtained prediction information to control the electrical discharge machining so as to suppress the wire breakage by the same method as described above. With the structure of the multi-wire electrical discharge machining system 600 as described above, it is possible to perform machine learning and inference processing that bears the load outside the multi-wire electrical discharge machine.

[0101] According to the fifth embodiment, wire breakage can be predicted in multi-wire EDM and machining conditions can be controlled to suppress wire breakage. In addition, the following effect is achieved, that is, the accuracy of predicting wire breakage is continuously improved because the learning stage of the machine learning device can be performed for each wire breakage.

[0102] The configuration described in the above embodiment is an example of the content of the present invention, and can be combined with other known technologies. Part of the configuration can be omitted or changed without departing from the scope of the present invention.

[0103] Description of the label

[0104] 1, 1a, 1b, 1c multi-wire electrical discharge machine, 2 workpiece, 3a, 3b, 3c, 3d guide roller, 4 bobbin, 5 wire discharge roller, 6 wire electrode, 6a wire cutting part, 6b power supply part, 7 machining power supply, 8 machining power supply unit, 9 machining control device, 10 ground electrode, 11 cable, 12 power supply, 13 power supply unit, 14 drive unit, 15 machining state detection device, 16 machining state detection unit, 17 bobbin drive unit, 18 guide roller drive unit, 19 wiring, 20, 30, 40, 50 wire break detection part, 20a, 20b connection target of wire break detection part, 20d wire break detection power supply, 21 wire break prediction part, 22, 22b machine learning device, 23 wire break prediction device, 60 analog voltage source, 70 processing table, 91 processor, 92 memory, 93 external storage device, 94 input unit, 95 output unit, 201, 202 rectification circuit, 201a, 202a rectification diode, 203 smoothing circuit, 203a high frequency disconnection inductor, 203b damping diode, 204 wire break detection power supply, 205 current limiting resistor, 206 wire break detection circuit, 207, 305, 404 wire break determination unit, 211, 222 state observation unit, 212 inference unit, 221 data acquisition unit, 223 model generation unit, 224 trained model storage unit, 301 low pass characteristic unit, 302, 402 wire break signal output unit, 303, 403 signal processing circuit, 304 Binary circuit, 304a photocoupler, 304b pull-up resistor, 401 frequency detection unit, 500, 600 multi-line electrical discharge machining system.

Claims

1. A multi-wire electrical discharge machine, comprising a cutting wire section composed of a plurality of cutting wires which are stretched in parallel by winding a wire electrode between a plurality of guide rollers and arranged in parallel at positions opposed to a workpiece, wherein an electrical discharge is generated between the cutting wire section and the workpiece, thereby cutting the workpiece. The multi-wire electrical discharge machine is characterized by having: a machining power source for applying a machining pulse voltage between the workpiece and the cutting line portion; and a wire breakage detection unit for detecting a breakage of the wire electrode; The wire breakage detection unit is electrically connected in parallel to a portion of the wire electrode that is stretched in a section from a location on the extraction side of the wire electrode to a location on the discharge side of the wire electrode. have: a wire breakage detection circuit, wherein a wire breakage detection power supply and a current limiting resistor are connected in series for supplying a direct current to the wire electrode portion at least during cutting of the workpiece; and A wire breakage determination unit determines whether the wire electrode is broken based on the current flowing through the current limiting resistor.

2. The multi-wire electrical discharge machine according to claim 1, characterized in that: The wire breakage detection circuit is an electrical circuit formed by further connecting a rectifier circuit and a smoothing circuit in series, wherein the rectifier circuit rectifies the high-frequency voltage generated in the wire electrode portion corresponding to the machining pulse voltage and the state of the discharge, and the smoothing circuit smoothes the current rectified by the rectifier circuit.

3. The multi-wire electrical discharge machine according to claim 1, characterized in that: The wire breakage determination unit determines that the wire electrode is broken when it can be detected that no current flows through the current limiting resistor.

4. The multi-wire electrical discharge machine according to claim 3, characterized in that: The wire breakage determination unit comprises: a binary circuit that converts the current into a binary signal of a Low level and a High level according to the presence or absence of the current flowing through the current limiting resistor; as well as The signal processing circuit detects a state in which no current flows through the current limiting resistor based on the binary signal, and outputs a wire disconnection signal indicating a result of determination of disconnection of the wire electrode.

5. The multi-wire electrical discharge machine according to claim 4, characterized in that: The signal processing circuit outputs the line disconnection signal based on a comparison result between a signal obtained by passing the binary signal through a low-pass filter and a predetermined threshold value.

6. The multi-wire electrical discharge machine according to claim 4, characterized in that The signal processing circuit detects the frequency of the binary signal, and outputs the line disconnection signal based on a comparison result with a predetermined threshold value.

7. The multi-wire electrical discharge machine according to any one of claims 1 to 6, characterized in that: The apparatus further comprises a wire breakage prediction unit that outputs prediction information indicating a possibility of the wire electrode being broken. The line disconnection prediction unit comprises: a state observation unit that acquires state data, which is information indicating a processing state; and An inference unit infers the predicted information from the state data input from the state observation unit using a trained model generated by machine learning based on a combination of the state data and a detection result of a break in the wire electrode.

8. The multi-wire electrical discharge machine according to claim 7, characterized in that: The state data includes at least one of a relative distance between the workpiece and the cutting wire portion, a wire travel speed, a wire tension, an inter-electrode voltage, a number of discharges per unit time, a number of short circuits per unit time, and the machining pulse voltage.

9. The multi-wire electrical discharge machine according to claim 7, characterized in that: It also has a machine learning device, which learns the predicted information. The machine learning device has: a state observation unit that obtains the state data; a data acquisition unit that acquires a detection result of a break in the wire electrode; as well as A model generation unit generates the trained model through machine learning according to learning data created based on a combination of the state data and a detection result of disconnection of the wire electrode.

10. The multi-wire electrical discharge machine according to claim 7, characterized in that: The invention further includes a machining control device that controls the electric discharge machining based on the foreknowledge information so as to suppress the wire electrode from being broken.

11. The multi-wire electrical discharge machine according to claim 10, characterized in that The machining control device controls the machining power source so as to reduce machining energy compared to current machining energy, when the predicted information indicates that there is a possibility of disconnection of the wire electrode.

12. The multi-wire electrical discharge machine according to claim 10, characterized in that Also features: a guide roller driving unit that drives the guide roller; and A bobbin driving unit drives the bobbin. When the predicted information indicates that the wire electrode may break, the processing control device controls the guide roller driving unit and the bobbin driving unit to reduce the wire tension or increase the wire travel speed compared to a current wire tension or a current wire travel speed.

13. A multi-wire electrical discharge machining system, characterized in that: have: The multi-wire electrical discharge machine according to any one of claims 1 to 6; and a wire breakage prediction device for outputting prediction information indicating the possibility of the wire electrode being broken, The wire breakage prediction device comprises: a state observation unit that acquires state data, which is information indicating a processing state; a data acquisition unit that acquires a detection result of a break in the wire electrode; a model generating unit that generates a trained model by machine learning according to learning data created based on a combination of the state data and a detection result of a break in the wire electrode; and An inference unit uses the trained model to infer the predicted information based on the state data input from the state observation unit.

14. The multi-wire electrical discharge machining system according to claim 13, characterized in that: The state data includes at least one of a relative distance between the workpiece and the cutting wire portion, a wire travel speed, a wire tension, an inter-electrode voltage, a number of discharges per unit time, a number of short circuits per unit time, and the machining pulse voltage.

Images