JPWO2021048959A1 - Numerical control device, numerical control method and machine learning device - Google Patents

Abstract

A numerical control device (1X) that controls a machine tool that controls multiple drive shafts to move a tool, and generates a vibration waveform that indicates the vibration path of the tool when machining a workpiece while vibrating the tool. A plurality of drives based on the waveform generator (385), the vibration direction of the tool when vibrating the tool, the angle in the direction of the reference axis which is one of the drive shafts of the plurality of drive shafts, and the vibration waveform. It is provided with a vibration movement amount calculation unit (386) that generates a command vibration waveform for each drive shaft that indicates the movement path of the tool on each shaft.

Description

The present invention relates to a numerical control device, a numerical control method, and a machine learning device for controlling a machine tool that vibrates and cuts a workpiece.

As a conventional numerical control device, there is a numerical control device that can realize vibration cutting processing in which a workpiece is machined while vibrating a cutting tool at a low frequency.

Patent Document 1 describes a numerical control device that realizes machining in which vibration cutting is applied to thread cutting. The numerical control device described in Patent Document 1 controls a machine tool to machine a machine tool while vibrating a cutting tool along one of a drive shafts orthogonal to the main shaft, thereby performing vibration cutting for thread cutting. We have realized processing that applies processing.

In addition, some numerical control devices are capable of virtual axis control. Virtual axis control is the control of cutting tools, etc. along a virtual axis (hereinafter referred to as a virtual axis) that is different from any of the drive axes by performing synchronous control that synchronizes the control of a plurality of drive axes. It is a control to move an object.

Japanese Patent No. 5851670

The numerical control device described in Patent Document 1 has a configuration in which a cutting tool is vibrated along one drive shaft to perform machining, and when the cutting tool is moved by virtual axis control, that is, when thread cutting is performed. When the vibration direction of the cutting tool is the direction of the virtual axis realized by synchronously controlling a plurality of drive axes, there is a problem that the vibration cutting process cannot be performed.

The present invention has been made in view of the above, and obtains a numerical control device to which vibration cutting can be applied even when the vibration direction of the cutting tool is different from any of the drive shafts. With the goal.

In order to solve the above-mentioned problems and achieve the object, the present invention is a numerical control device that controls a machine tool that controls a plurality of drive shafts to move a tool, and processes a workpiece while vibrating the tool. The angle between the vibration waveform generator that generates the vibration waveform that indicates the vibration path of the tool, the vibration direction of the tool when the tool is vibrated, and the direction of the reference axis that is one of the multiple drive shafts. And a vibration movement amount calculation unit that generates a command vibration waveform for each drive shaft indicating the movement path of the tool in each of the plurality of drive shafts based on the vibration waveform.

The numerical control device according to the present invention has an effect that the vibration cutting process can be applied even when the vibration direction of the cutting tool is different from that of each drive shaft.

The figure for demonstrating the virtual axis control used in the numerical control apparatus concerning each embodiment. The figure for demonstrating the configuration example of the target machine tool controlled by the numerical control device which concerns on each embodiment. The figure which shows the structural example of the numerical control apparatus which concerns on Embodiment 1. The figure which shows the vibration direction of a tool at the time of vibrating cutting in thread cutting. The figure which shows the image of the vibration cutting in the thread cutting process A flowchart showing an example of an operation in which the numerical control device according to the first embodiment performs thread cutting while vibrating the tool by virtual axis control. The figure which shows an example of the command vibration waveform generated by the vibration waveform generation part of the numerical control apparatus which concerns on Embodiment 1. The figure which shows an example of the relationship between the command vibration waveform and the feedback vibration waveform. The figure which shows an example of the synthetic vibration waveform obtained by synthesizing the command vibration waveform of each axis before adjustment. The figure which shows an example of the synthetic vibration waveform obtained by synthesizing the command vibration waveform of each axis after adjustment. Diagram for explaining tilt axis control The figure which shows the configuration example of the machine tool which uses the tilt axis control. Diagram for explaining the configuration of a machine tool that performs vibration cutting The figure which shows the hardware configuration example of the control calculation part included in the numerical control device which concerns on Embodiment 1. The figure which shows the structural example of the numerical control apparatus which concerns on Embodiment 2.

Hereinafter, the numerical control device, the numerical control method, and the machine learning device according to the embodiment of the present invention will be described in detail with reference to the drawings. The present invention is not limited to this embodiment.

In each embodiment, a configuration in which the numerical control device performs thread cutting while vibrating the cutting tool by virtual axis control will be described. Therefore, first, the virtual axis control will be described, and the machine tool to be controlled by the numerical control device according to each embodiment will be described.

FIG. 1 is a diagram for explaining virtual axis control used in the numerical control device according to each embodiment. In virtual axis control, a controlled object is controlled using a virtual machine coordinate system obtained by rotating a real machine coordinate system, which is a coordinate system possessed by a machine tool controlled by a numerical control device. Each axis of the virtual machine coordinate system is called a virtual axis. In the example shown in FIG. 1, the actual machine coordinate system including the X-axis and the Y-axis is rotated by 45 ° clockwise, and the origin (described as the actual machine coordinate origin in FIG. 1) is set to −120 mm on the X-axis. The virtual machine coordinate system is the one that is shifted and shifted by -60 mm on the Y axis.

In the virtual axis control using the virtual machine coordinate system shown in FIG. 1, for example, the vibration direction of the cutting tool used in thread cutting is tilted by 45 ° from the X axis of the actual machine coordinate system (described as the actual X axis in FIG. 1). Used when In this virtual axis control, the vibration direction of the cutting tool and the direction of the virtual X axis coincide with each other.

FIG. 2 is a diagram for explaining a configuration example of a target machine tool controlled by the numerical control device according to each embodiment. As shown in FIG. 2, the target machine tool controlled by the numerical control device according to each embodiment has a tool post to which five types of cutting tools T1 to T5 are attached, and any one of the cutting tools. It is a configuration in which the work is machined using. FIG. 2 shows a cross section of the turret when viewed from the direction of the rotation axis of the work. Each cutting tool is installed so that the cutting tool faces the direction of the work. Each cutting tool is controlled to move in a straight line toward the workpiece. In each cutting tool, the numerical control device controls the X-axis servomotor that moves the tool post in the actual X-axis direction and the Y-axis servomotor that moves the tool post in the actual Y-axis direction, thereby centering the work. Move on a straight line passing through. The number of cutting tools is an example and is not limited to five types.

When vibration cutting is performed using a machine tool having the configuration shown in FIG. 2, the cutting tool is vibrated on a straight line passing through the center of the work by writing vibration commands for each of the X-axis and the Y-axis in the machining program. It is possible to make it. However, when the vibration cutting process is performed by such a method, the load for creating the processing program becomes large. Therefore, the numerical control device according to each embodiment enables the virtual axis control to be used in the control of the cutting tool. By using the virtual axis control, the tool can be vibrated along one virtual axis. That is, it is possible to vibrate the cutting tool on the virtual axis by a vibration command for one axis. The creator of the machining program does not need to be aware of the virtual axis when creating a machining program for vibration cutting, and vibrates the cutting tool in the direction of the reference axis with the actual X-axis or the actual Y-axis as the reference axis. Vibration conditions may be specified using commands. As a result, a work of creating a machining program for performing vibration cutting using a machine tool having the configuration shown in FIG. 2, that is, a machine tool having a configuration in which the vibration direction of the cutting tool is different from the direction of any drive shaft. Load is reduced.

When the numerical control device controls the machine tool having the configuration shown in FIG. 2, the setting of the virtual machine coordinates is changed every time the cutting tool used for machining the work is changed. FIG. 2 shows an example in which a work is machined using the cutting tool T2. In this case, the numerical control device virtualizes the direction in which the cutting tool T2 moves and the direction of the virtual X-axis coincide with each other. After setting the machine coordinates, the operation of the tool post is controlled using the virtual machine coordinates. The case where the cutting tool T2 is used has been described, but the same applies to the case where another cutting tool is used. For example, when the cutting tool T4 is used, the numerical control device changes the setting of the virtual machine coordinates so that the direction in which the cutting tool T4 moves and the direction of the virtual X-axis match. When the cutting tool T3 is used, the numerical control device changes the setting of the disguised machine coordinates so that the direction in which the cutting tool T3 moves and the direction of the virtual X-axis coincide with each other. In this case, the virtual X-axis is Matches the real X-axis. In the example shown in FIG. 2, it was decided to set so that the vibration direction of the cutting tool and the direction of the virtual X-axis match, but the setting is made so that the vibration direction of the cutting tool and the direction of the virtual Y-axis match. You may.

As described above, in the numerical control device according to each embodiment, the vibration direction of the cutting tool used and the direction of the virtual X-axis, which is the X-axis of the virtual machine coordinates, coincide with each other, or the vibration of the cutting tool used. The virtual machine coordinates are set so that the direction and the direction of the virtual Y-axis, which is the Y-axis of the virtual machine coordinates, match, and the virtual machine coordinates are used to control each axis of the machine tool. The numerical control device according to each embodiment controls the X-axis servomotor and the Y-axis servomotor shown in FIG. 2 to move the cutting tools T1 to T5 in the actual X-axis direction and the actual Y-axis direction. In each embodiment, the cutting tool is referred to as a "tool". Further, in the description of each embodiment, the actual X-axis may be referred to as an X-axis, and the actual Y-axis may be referred to as a Y-axis.

Embodiment 1.
FIG. 3 is a diagram showing a configuration example of the numerical control device according to the first embodiment. The numerical control device 1X according to the first embodiment includes an input operation unit 3, a display unit 4, and a control calculation unit 2X. In FIG. 3, the drive unit 90 provided in the machine tool controlled by the numerical control device 1X is also shown. The description of the components other than the drive unit 90 of the machine tool is omitted.

The drive unit 90 provided in the machine tool is a mechanism for driving one or both of the work and the tool, which are the objects to be machined, in at least two axial directions. Here, a plurality of servomotors 91 for moving one or both of the workpiece and the tool in the direction of each axis defined on the numerical control device 1X, and a plurality for detecting the position and rotation speed of the rotor of each servomotor 91. The detector 92 and the above are provided. Further, the drive unit 90 includes an X-axis servo control unit 93X, a Y-axis servo control unit 93Y, ... That controls the servomotor 91 based on the position and rotation speed detected by the detector 92. In the following, when it is not necessary to distinguish the directions of the drive axes, the servo control units (X-axis servo control unit 93X, Y-axis servo control unit 93Y, ...) Corresponding to each axis are simply referred to as servo control units 93. do. Further, the drive unit 90 includes a spindle motor 94 for rotating the spindle for rotating the work, a detector 95 for detecting the position and rotation speed of the rotor of the spindle motor 94, and a position and rotation detected by the detector 95. A spindle control unit 96 that controls the spindle motor 94 based on the number is provided.

Returning to the description of the numerical control device 1X, the input operation unit 3 is a means for inputting information to the numerical control device 1X. The input operation unit 3 is composed of a keyboard, operation buttons, a mouse, and the like, and receives inputs such as commands, processing programs, and parameters for the numerical control device 1X by the user, and delivers them to the control calculation unit 2X.

The display unit 4 is composed of a liquid crystal display device or the like, and displays information processed by the control calculation unit 2X and the like.

The control calculation unit 2X includes an input control unit 32, a data setting unit 33, a storage unit 34, a screen processing unit 31, a control signal processing unit 35, a PLC (Programmable Logic Controller) 36, and an analysis processing unit 37. , An interpolation processing unit 38X, an acceleration / deceleration processing unit 39, and an axis data input / output unit 40. The PLC 36 may be arranged outside the control calculation unit 2X.

The input control unit 32 receives the information input from the input operation unit 3. The data setting unit 33 stores the information received by the input control unit 32 in the storage unit 34. For example, when the input content is an edit of the machining program 343 held by the storage unit 34, the data setting unit 33 reflects the edited content in the machining program 343 held by the storage unit 34. When a parameter is input, the data setting unit 33 updates the parameter 341 held by the storage unit 34.

The storage unit 34 stores parameters 341 used in the processing of the control calculation unit 2X, display data 342 to be displayed on the display unit 4, processing program 343 to be executed, and the like.

Further, the storage unit 34 is provided with a shared area 344 for storing data other than the parameter 341, the display data 342, and the processing program 343. In the shared area 344, data generated by a process in which the control calculation unit 2X controls the drive unit 90 is temporarily stored. The screen processing unit 31 controls the display unit 4 to display the display data 342 held by the storage unit 34.

The analysis processing unit 37 includes a movement command analysis unit 371 and a vibration command analysis unit 372. The analysis processing unit 37 reads a machining program 343 including one or more blocks from the storage unit 34, and analyzes the read machining program 343 with the movement command analysis unit 371 or the vibration command analysis unit 372. The movement command analysis unit 371 analyzes the movement command included in the machining program 343 and writes the analysis result in the shared area 344 of the storage unit 34. The vibration command analysis unit 372 analyzes the vibration command included in the machining program 343 and writes the analysis result in the shared area 344 of the storage unit 34. The vibration command includes an argument indicating the content of the vibration of the tool in the vibration cutting process. The vibration command includes an argument indicating the vibration direction, that is, in which axis direction the tool is vibrated, an argument indicating the amplitude of vibration, an argument indicating the frequency of vibration, and the like. The frequency of vibration is indicated by, for example, the number of vibrations during one rotation of the spindle. The vibration command is indicated by, for example, a G165 code.

The control signal processing unit 35 indicates that the auxiliary command is commanded when the analysis processing unit 37 reads an auxiliary command as a command for operating a machine other than a command for operating the drive shaft, which is a numerical control axis. Notify PLC36. An example of an auxiliary command is an M code or a T code.

Upon receiving the notification from the control signal processing unit 35 that the auxiliary command has been commanded, the PLC 36 executes the process corresponding to the auxiliary command. The PLC 36 holds a ladder program in which the machine operation is described. When the PLC 36 receives the T code or M code which is the auxiliary command, the PLC 36 executes the process corresponding to the auxiliary command according to the ladder program. After executing the process corresponding to the auxiliary command, the PLC 36 sends a completion signal indicating that the process corresponding to the auxiliary command is completed to the control signal processing unit 35 in order to execute the next block of the machining program 343.

In the control calculation unit 2X, the control signal processing unit 35, the analysis processing unit 37, and the interpolation processing unit 38X are connected via the storage unit 34. The analysis processing unit 37, the control signal processing unit 35, and the interpolation processing unit 38X transfer various information via the shared area 344 of the storage unit 34. In the following, when explaining the transfer of information between the control signal processing unit 35, the analysis processing unit 37, and the interpolation processing unit 38X, the description that the storage unit 34 is interposed may be omitted.

When the analysis processing unit 37 analyzes a command including an argument related to the tool movement path, the interpolation processing unit 38X calculates the tool movement path by interpolation processing using the argument included in the analyzed command. The command including an argument related to the movement path of the tool is a command including one or more arguments indicating the position of the tool, the movement speed of the tool, the interpolation method used in the interpolation process, and the like. The vibration command also corresponds to a command including an argument related to the movement path of the tool.

The interpolation processing unit 38X includes a distribution ratio determination unit 381, a waveform information acquisition unit 382, a comparison unit 383, a command vibration waveform adjustment unit 384X, a vibration waveform generation unit 385, and a vibration movement amount calculation unit 386.

The vibration waveform generation unit 385 controls the movement of the tool based on the result of the vibration command analysis unit 372 analyzing the vibration command, that is, based on the argument included in the analyzed vibration command. Generates a vibration waveform indicating. The vibration waveform generated by the vibration waveform generation unit 385 indicates the vibration path of the tool when the tool is vibrated in the direction of the virtual axis.

The distribution ratio determination unit 381 determines the ratio of the vibration waveform generated by the vibration waveform generation unit 385 to be distributed to each drive shaft controlled when the tool is moved by virtual axis control. As described above, the virtual axis control moves the tool in the direction of the virtual axis by synchronously controlling a plurality of drive axes. Therefore, the amount of movement in the direction of the virtual axis of the tool is a combination of the amount of movement in each direction of each drive shaft that is synchronously controlled by the virtual axis control. On the other hand, the vibration waveform generated by the vibration waveform generation unit 385 is a waveform in the virtual machine coordinate system expressed by using the virtual axis, that is, a waveform showing vibration in the direction of the virtual axis. Therefore, the distribution ratio determination unit 381 determines at what ratio the components of the vibration waveform generated by the vibration waveform generation unit 385 should be distributed to each drive shaft to obtain a waveform showing vibration in each direction of the drive shaft. decide.

The waveform information acquisition unit 382 acquires information on the waveform indicating the actual movement path of the tool when the tool is controlled to vibrate from the drive unit 90.

The comparison unit 383 compares the vibration waveform generated by the vibration waveform generation unit 385 with the feedback waveform which is the waveform indicated by the information acquired by the waveform information acquisition unit 382. The comparison unit 383 separates the vibration waveform generated by the vibration waveform generation unit 385 into an X-axis component and a Y-axis component for comparison. Specifically, the comparison unit 383 uses information indicating the vibration waveform on the X-axis indicated by the component of the X-axis from the detector 92 that detects the position and speed of the rotor of the X-axis servomotor 91. Compare with the feedback waveform of the axis. Further, the comparison unit 383 feeds back the vibration waveform on the Y-axis indicated by the Y-axis component by the information obtained from the detector 92 that detects the position and speed of the rotor of the Y-axis servomotor 91. Compare with waveform.

The command vibration waveform adjusting unit 384X adjusts the command vibration waveform described later based on the comparison result by the comparison unit 383.

The vibration movement amount calculation unit 386 is an X-axis command vibration indicating the movement path of the tool on the X-axis based on the vibration waveform generated by the vibration waveform generation unit 385 and the ratio determined by the distribution ratio determination unit 381. A unit of a tool when a Y-axis command vibration waveform indicating a waveform and a tool movement path on the Y-axis is generated, and the tool is vibrated to perform vibration cutting based on the generated command vibration waveform of each axis. Calculate the amount of movement per hour. The vibration movement amount calculation unit 386 calculates the vibration movement amount, which is the movement amount of the tool per unit time, for each drive shaft. That is, the vibration movement amount calculation unit 386 calculates the vibration movement amount of the X-axis of the tool and the vibration movement amount of the Y-axis of the tool based on the command vibration waveform for each drive shaft.

The acceleration / deceleration processing unit 39 calculates the movement amount of each drive shaft per unit time received from the vibration movement amount calculation unit 386 of the interpolation processing unit 38X, based on a predetermined acceleration / deceleration pattern, in consideration of acceleration / deceleration. Convert to a movement command per hour.

The axis data input / output unit 40 outputs a movement command per unit time output from the acceleration / deceleration processing unit 39 to the servo control unit 93 that controls each drive axis. Further, the axis data input / output unit 40 acquires data indicating the position and rotation speed of each servomotor 91 from the drive unit 90.

The operation when the numerical control device 1X shown in FIG. 3 performs vibration cutting in thread cutting will be briefly described.

The analysis processing unit 37 of the numerical control device 1X reads one block from the machining program 343, analyzes the read block by the movement command analysis unit 371 when the read block is a thread cutting command, and when the read block is a vibration command, a vibration command. Analysis is performed by the analysis unit 372. For example, if the read block is a G33 code, the analysis processing unit 37 determines that it is a thread cutting command, and if the read block is a G165 code, it determines that it is a vibration command.

The thread cutting command includes an argument that specifies the movement path of the tool in one thread cutting process. The thread cutting command includes, for example, an argument indicating a position where thread cutting is started, an argument indicating a position where thread cutting is finished, and an argument (pitch) indicating the amount of movement of the tool per rotation of the work. The tool used in thread cutting is assumed to be specified in advance by a command for specifying the tool to be used, but an argument for specifying the tool to be used may be included in the thread cutting command.

When performing vibration cutting in thread cutting, unlike the case of performing vibration cutting in normal workpiece processing, the tool is placed in the direction perpendicular to the thread cutting direction, that is, in the direction perpendicular to the rotation axis of the work. Thread cutting is performed while vibrating (see FIG. 4). FIG. 4 is a diagram showing a vibration direction of a tool when vibration cutting is performed in thread cutting. It is assumed that the X-axis and Z are orthogonal to each other. As shown in FIG. 4, when the machining direction is the Z-axis direction, the direction in which the tool is vibrated by vibrating cutting is the X-axis direction.

When the vibration command analysis unit 372 analyzes the vibration command, the vibration waveform generation unit 385 acquires the analysis result of the vibration command via the common area 344, and the vibration waveform which is the basic waveform of the vibration based on the acquired analysis result. To generate. The vibration movement amount calculation unit 386 obtains, for example, the vibration movement amount of the X-axis by using the vibration waveform generated by the vibration waveform generation unit 385 and the movement path of the tool. Specifically, the vibration movement amount calculation unit 386 obtains a vibration forward position in which the amplitude of the vibration waveform is added to the movement path of the tool and a vibration reverse position in which the amplitude of the vibration waveform is subtracted from the movement path of the tool. Generates the amount of vibration movement of the shaft. The movement path of the tool is obtained from the analysis result of the thread cutting command by the movement command analysis unit 371.

The vibration movement amount calculated by the vibration movement amount calculation unit 386 is sent to the drive unit 90 via the acceleration / deceleration processing unit 39 and the shaft data input / output unit 40. In the drive unit 90, the X-axis servo control unit 93X controls the X-axis servomotor 91 based on the vibration movement amount received from the vibration movement amount calculation unit 386. That is, vibration cutting in thread cutting is performed. FIG. 5 is a diagram showing an image of vibration cutting in thread cutting. In FIG. 5, the dark portion indicates a portion where the depth of cut is increased due to the vibration of the tool and the work is deeply cut, and the light portion indicates a portion where the work is shallowly cut. When the phase of the vibration in the X-axis direction shown in FIG. 4 of the first thread cutting process and the vibration in the X-axis direction of the second thread cutting process are shifted by, for example, 180 °, the portion deeply cut in the first time is cut shallowly in the second time. It will be squeezed, and the chips will be divided at that part. That is, shredded chips will be discharged.

FIG. 6 is a flowchart showing an example of an operation in which the numerical control device 1X according to the first embodiment performs thread cutting while vibrating the tool by virtual axis control. Here, when performing virtual axis control, the machining program 343 includes a virtual axis control command instructing the start of virtual axis control. That is, when the tool is vibrated on the virtual axis by the virtual axis control, the creator of the machining program 343 has a machining program 343 having a configuration in which the numerical control device 1X first executes the virtual axis control command and then executes the vibration command. To create. The virtual axis control command includes various arguments indicating the relationship between the virtual machine coordinate system used in the virtual axis control and the real machine coordinate system. An example of the argument included in the virtual axis control command is the rotation angle when the real machine coordinate system is rotated and the virtual machine coordinate system is set, that is, each real of the real machine coordinate system is set when the virtual machine coordinate system is set. An argument indicating the rotation angle indicating how much the axis is rotated, an argument indicating how much the real machine coordinate system is shifted in the direction of each real axis when setting the virtual machine coordinate system, and the like. The vibration command includes an argument indicating the vibration direction, but the argument of the vibration command executed after starting the virtual axis control indicates the vibration direction in the virtual machine coordinate system. For example, when a vibration command including an argument indicating vibration in the X-axis direction is executed after the start of virtual axis control, the vibration direction indicated by the argument of the vibration command is the virtual X-axis direction.

In the operation shown in FIG. 6, first, the analysis processing unit 37 acquires the rotation angle from the virtual axis control command among the commands included in the machining program 343 (step S1). Although the rotation angle is acquired from the virtual axis control command here, the analysis processing unit 37 may acquire the rotation angle based on the command for designating the tool to be used. In the present embodiment, as shown in FIG. 2, the direction in which each tool used in thread cutting is moved during machining is predetermined. For example, the angle between the virtual X-axis and the real X-axis that coincide with the vibration direction of the cutting tool T2 is 45 °. Therefore, when the cutting tool T2 is designated as the tool to be used, the rotation angle is 45 °. Further, when the cutting tool T1 is designated as a tool to be used, the rotation angle is 90 °. Therefore, if the tool to be used is known, the angle between the virtual axis and the real axis, that is, the rotation angle can be known.

Next, the interpolation processing unit 38X distributes the vibration waveform to each of the actual axes based on the rotation angle (step S2). In this step S2, first, the distribution ratio determination unit 381 distributes the vibration waveform generated by the vibration waveform generation unit 385 to the actual X-axis and the actual Y-axis based on the rotation angle acquired in step S1 (hereinafter,). , Called the distribution ratio). The distribution ratio determination unit 381 notifies the vibration movement amount calculation unit 386 of the determined distribution ratio. The vibration waveform generation unit 385 generates a vibration waveform based on the analysis result of the vibration command and passes it to the vibration movement amount calculation unit 386. The vibration movement amount calculation unit 386 generates a command vibration waveform for each axis based on the vibration waveform generated by the vibration waveform generation unit 385 and the distribution ratio determined by the distribution ratio determination unit 381. That is, the vibration movement amount calculation unit 386 distributes the components of the vibration waveform generated by the vibration waveform generation unit 385 to each of the actual axes according to the distribution ratio, and generates a command vibration waveform for each axis. In the present embodiment, the vibration movement amount calculation unit 386 generates a commanded vibration waveform on the actual X-axis and a commanded vibration waveform on the actual Y-axis. Then, the vibration movement amount calculation unit 386 superimposes the command vibration waveform of the actual X axis on the movement amount of the tool on the actual X axis, and superimposes the command vibration waveform of the actual Y axis on the movement amount of the tool on the actual Y axis. By doing so, the amount of vibration movement of each axis is generated. The vibration movement amount of each axis generated by the vibration movement amount calculation unit 386 is determined by the servo control unit 93 (X-axis servo control unit) of each axis of the drive unit 90 via the acceleration / deceleration processing unit 39 and the axis data input / output unit 40. It is sent to 93X, Y-axis servo control unit 93Y, ...). The servo control unit 93 of each axis controls the servomotor 91 to be controlled based on the vibration movement amount received from the vibration movement amount calculation unit 386.

In the example shown in FIG. 2, the cutting tool T2 is selected, and the rotation angle of the virtual X-axis corresponding to the vibration direction of the cutting tool T2 is 45 °. In this case, the distribution ratio of the vibration waveform is 1/2 on the actual X-axis and 1/2 on the actual Y-axis. FIG. 7 shows an example of the commanded vibration waveform of the actual X-axis and the commanded vibration waveform of the actual Y-axis when the cutting tool T2 shown in FIG. 2 is selected and thread cutting is performed with vibration cutting. FIG. 7 is a diagram showing an example of a command vibration waveform generated by the vibration waveform generation unit 385 of the numerical control device 1X according to the first embodiment. FIG. 7A shows a commanded vibration waveform on the actual X-axis, and FIG. 7B shows a commanded vibration waveform on the actual Y-axis. For the sake of simplicity, FIG. 7 shows a command vibration waveform that does not consider the movement amount of each axis, that is, a vibration waveform when the movement amount of each axis is 0. Since the distribution ratio of the vibration waveform is 1/2 on the actual X-axis and 1/2 on the actual Y-axis, the commanded vibration waveform on the actual X-axis and the commanded vibration waveform on the actual Y-axis have the same amplitude and frequency. There is. If the rotation angle is other than 45 °, the amplitude of the commanded vibration waveform on the actual X-axis and the amplitude of the commanded vibration waveform on the actual Y-axis are different. The frequencies of the command vibration waveform of the actual X-axis and the command vibration waveform of the actual Y-axis have the same values even when the rotation angle is other than 45 °.

Returning to the description of FIG. 6, next, the waveform information acquisition unit 382 uses the feedback vibration waveform of each axis (hereinafter referred to as the FB vibration waveform) based on the data output from the detector 92 attached to each servomotor 91. To be acquired (step S3). The FB vibration waveform is the actual vibration waveform of the tool used in machining.

Next, the comparison unit 383 compares the command vibration waveform of each axis generated by the vibration movement amount calculation unit 386 with the FB vibration waveform of each axis acquired in step S3 (step S4). The command vibration waveform and the FB vibration waveform are compared for each axis. For example, when the command vibration waveform of each axis generated by the vibration movement amount calculation unit 386 is the waveform shown by the solid line in FIG. 8 and the FB vibration waveform of each axis is the waveform shown by the dotted line in FIG. 8, the comparison unit 383 , It is judged that the amplitude is different between the command vibration waveform of the actual X axis and the FB vibration waveform, and the amplitude and phase are different between the command vibration waveform of the actual Y axis and the FB vibration waveform. The comparison unit 383 determines, for example, whether the amplitudes match by comparing the maximum value of the amplitude of the command vibration waveform with the maximum value of the amplitude of the FB vibration waveform. Further, the comparison unit 383 determines whether the phases match by comparing, for example, the position where the amplitude of the command vibration waveform is maximum and the position where the amplitude of the FB vibration waveform is maximum. The comparison unit 383 determines that the amplitudes match when the difference between the amplitudes of the command vibration waveform and the FB vibration waveform is equal to or less than the specified value, and the phase difference between the command vibration waveform and the FB vibration waveform is determined. If it is less than or equal to the value, it may be determined that the phases match.

When the amplitudes of the command vibration waveform of each axis and the FB vibration waveform match (step S5: Yes), the interpolation processing unit 38X proceeds to the process of step S7. When the amplitudes of the commanded vibration waveform and the FB vibration waveform of each axis do not match, that is, when the amplitudes of the commanded vibration waveform and the FB vibration waveform of at least one of the actual X-axis and the actual Y-axis do not match (step S5). : No), the command vibration waveform adjusting unit 384X adjusts the command vibration waveform so that the difference between the amplitudes of the command vibration waveform and the FB vibration waveform of the axes whose amplitudes do not match approaches 0 (step S6). In the case of the example shown in FIG. 8, since the amplitudes do not match in both the actual X-axis and the actual Y-axis, the command vibration waveform adjusting unit 384X uses the command vibration waveform of the actual X-axis and the command vibration waveform of the actual Y-axis. Adjust the amplitude. The command vibration waveform adjustment unit 384X calculates, for example, the difference between the peak value of the command vibration waveform and the peak value of the FB vibration waveform at a plurality of points, and adds the average value to the peak value of the command vibration waveform to be created next time. Adjust the amplitude by subtracting. Hereinafter, the adjustment amount when adjusting the amplitude of the command vibration waveform is referred to as a vibration amplitude adjustment amount.

When the phases of the command vibration waveform of each axis and the FB vibration waveform match (step S7: Yes), the interpolation processing unit 38X returns to the processing of step S3. When the phases of the commanded vibration waveform and the FB vibration waveform of each axis do not match, that is, when the phases of the commanded vibration waveform and the FB vibration waveform of at least one of the actual X-axis and the actual Y-axis do not match (step S7). : No), the command vibration waveform adjusting unit 384X adjusts the command vibration waveform so that the phase shift between the command vibration waveform and the FB vibration waveform of the axes whose phases do not match approaches 0 (step S8). In the case of the example shown in FIG. 8, the command vibration waveform of the actual X-axis and the FB vibration waveform have the same phase, and the command vibration waveform of the actual Y-axis and the FB vibration waveform do not have the same phase. The adjusting unit 384X adjusts the phase of the command vibration waveform of the actual Y axis. When the phase of the commanded vibration waveform and the phase of the FB vibration waveform are out of phase, the phase of the FB vibration waveform is usually delayed from the phase of the commanded vibration waveform. The command vibration waveform adjusting unit 384X adjusts servo parameters such as current loop gain and speed loop gain in order to improve the responsiveness of the servo motor 91 and bring the phase shift closer to zero. These servo parameters are included in the parameter 341 held by the storage unit 34. The command vibration waveform adjusting unit 384X adjusts by increasing the value of each servo parameter required for phase adjustment among the servo parameters included in the parameter 341 by, for example, "5". The adjusted parameters are transmitted to the X-axis servo control unit 93X, the Y-axis servo control unit 93Y, ... Via the acceleration / deceleration processing unit 39 and the axis data input / output unit 40, and as a result, the current loop of the servomotor 91. Gain, speed loop gain, etc. are adjusted. Here, the adjustment amount of the servo parameter is set to "5", but other adjustment amount may be used. After executing step S8, the interpolation processing unit 38X returns to the processing of step S3.

A composite vibration waveform obtained by synthesizing the command vibration waveform of the actual X-axis and the command vibration waveform of the actual Y-axis by executing the operation shown in FIG. 6 and adjusting the command vibration waveform is shown in FIG. From the one shown in FIG. The synthetic vibration waveform shown in FIG. 9 is the synthetic vibration waveform before adjusting the amplitude and phase of the commanded vibration waveform of each axis, and the synthetic vibration waveform shown in FIG. 10 is adjusted in the amplitude and phase of the commanded vibration waveform of each axis. It is a later synthetic vibration waveform.

By performing the operation shown in FIG. 6, that is, the operation of adjusting the amplitude and phase of the command vibration waveform of each axis based on the FB vibration waveform of each axis for moving the tool, the responsiveness of the servomotor 91 of each axis is improved. Even if they are different, the synthetic vibration waveform can be brought close to the normal shape. As a result, chips can be reliably divided in the vibration cutting process performed by using the virtual axis control.

In the present embodiment, the case where the virtual axis control is applied to the vibration cutting process has been described, but the tilt axis control shown in FIG. 11 can also be applied. The tilt axis control shown in FIG. 11 controls a machine tool having the configuration shown in FIG. 12, that is, a machine tool having an angle other than 90 ° on the axes (X-axis and Y-axis in FIGS. 11 and 12) to which the servomotor is attached. Used when doing. In the case of the example shown in FIG. 11, the inclination angle between the actual Y-axis and the program Y-axis (the y-axis in the figure) is θ. When vibrating the tool in the y-axis direction, it is necessary to synchronously control the real X-axis and real Y-axis servomotors, but the vibration waveform of the y-axis is generated as in the case of applying the virtual axis control described above. Then, by distributing this to the actual X-axis and the actual Y-axis based on the inclination angle θ, the commanded vibration waveforms of the actual X-axis and the actual Y-axis may be generated.

It is also possible to perform vibration cutting with the machine tool having the configuration shown in FIG. The machine tool shown in FIG. 13 has a configuration in which a tool is rotated, positioned, and used. Further, the moving direction of the tool does not match the control axes (X-axis and Z-axis in the figure) of the machine tool. Therefore, when the tool is vibrated, it is necessary to synchronously control the servomotors of a plurality of axes. However, the command vibration waveforms of the plurality of axes may be generated by the same method as the virtual axis control and the tilt axis control described above. ..

When the cutting tools T1, T3, and T5 shown in FIG. 2 are used, the vibration command is executed without executing the virtual axis control command. In this case, the distribution ratio determination unit 381 of the interpolation processing unit 38X has a rotation angle of 90 ° when the cutting tool T1 is used, a rotation angle of 0 ° when the cutting tool T3 is used, and when the cutting tool T5 is used. Determines the distribution ratio with the rotation angle as −90 °.

Next, the hardware configuration of the control calculation unit 2X included in the numerical control device 1X will be described. FIG. 14 is a diagram showing a hardware configuration example of the control calculation unit 2X included in the numerical control device 1X according to the first embodiment.

The control calculation unit 2X can be realized by the processor 101 and the memory 102 shown in FIG. An example of the processor 101 is a CPU (Central Processing Unit, a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, a DSP (Digital Signal Processor)) or a system LSI (Large Scale Integration). An example of the memory 102 is a RAM (Random Access Memory) or a ROM (Read Only Memory).

The control calculation unit 2X is realized by the processor 101 reading and executing a program stored in the memory 102 for executing the operation of the control calculation unit 2X. It can also be said that this program causes the computer to execute the procedure or method of the control calculation unit 2X. The memory 102 is also used as a temporary memory when the processor 101 executes various processes.

The program executed by the processor 101 may be a computer program product having a computer-readable and non-transitory recording medium containing a plurality of instructions for performing data processing, which can be executed by a computer. .. The program executed by the processor 101 causes the computer to execute data processing by a plurality of instructions.

Further, the control calculation unit 2X may be realized by dedicated hardware. Further, the functions of the control calculation unit 2X may be partially realized by dedicated hardware and partly realized by software or firmware.

As described above, the numerical control device 1X according to the present embodiment is virtual by rotating the vibration waveform generation unit 385 that generates a vibration waveform indicating the movement path of the tool when the tool is vibrated and the actual machine coordinate system. A distribution ratio determination unit 381 that determines the distribution ratio when distributing the vibration waveform to each of the controlled target axes when the tool is vibrated, based on the rotation angle of the actual machine coordinate system when setting the machine coordinate system. A vibration movement amount calculation unit 386 is provided, which generates a vibration waveform of a tool on each of the control target axes based on the vibration waveform and the distribution ratio, and calculates the vibration movement amount of each control target axis based on the generated vibration waveform. .. As a result, vibration cutting can be performed even when the vibration direction of the tool is the direction of the virtual axis realized by synchronously controlling a plurality of drive axes. Further, the numerical control device 1X has a comparison unit 383 that compares the actual vibration waveform of the tool on each of the control target axes with the vibration waveform of the tool on each of the control target axes generated by the vibration movement amount calculation unit 386. A command vibration waveform adjusting unit 384X for adjusting the vibration waveform of the tool in each of the controlled target axes generated by the vibration moving amount calculation unit 386 based on the comparison result in the comparison unit 383 is provided. As a result, even if the responsiveness of the servomotor 91 of each control target axis is different, the chips can be reliably divided.

Embodiment 2.
FIG. 15 is a diagram showing a configuration example of the numerical control device according to the second embodiment. In FIG. 15, the same components as those of the numerical control device 1X according to the first embodiment are designated by the same reference numerals.

The numerical control device 1Y according to the second embodiment has a configuration in which the control calculation unit 2X of the numerical control device 1X according to the first embodiment is replaced with the control calculation unit 2Y. The control calculation unit 2Y has a configuration in which the interpolation processing unit 38X of the control calculation unit 2X according to the first embodiment is replaced with the interpolation processing unit 38Y, and the machine learning device 50 is added. Since the configurations other than the interpolation processing unit 38Y and the machine learning device 50 of the control calculation unit 2Y are the same as those of the control calculation unit 2X according to the first embodiment, the description thereof will be omitted.

The interpolation processing unit 38Y has a configuration in which the command vibration waveform adjustment unit 384X of the interpolation processing unit 38X is replaced with the command vibration waveform adjustment unit 384Y. The command vibration waveform adjusting unit 384Y corrects the command vibration waveform based on the vibration waveform correction information output by the machine learning device 50. The machine learning device 50 includes a learning unit 51 and a state observing unit 52.

The machine learning device 50 comprises the inertia information of each axis stored in the common area 344, the vibration waveform deviation amount of each axis obtained by the comparison unit 383 comparing the vibration waveform of each axis and the FB vibration waveform, and the vibration waveform deviation amount of each axis. Machine learning is performed using the change information of the servo parameters of each axis included in the parameter 341, and the vibration waveform correction information used when the command vibration waveform adjustment unit 384Y corrects the command vibration waveform is generated.

The vibration waveform deviation amount used by the machine learning device 50 for learning to estimate the vibration waveform correction information is a difference in amplitude and a difference in phase between the vibration waveform and the FB vibration waveform. The difference in amplitude between the vibration waveform and the FB vibration waveform is the vibration amplitude adjustment amount described above. Further, regarding the change information of the servo parameters of each axis, the servo parameters of each axis are a current loop gain, a speed loop gain, and the like which are changed in step S8 of FIG. 6 described above in order to bring the phase shift closer to zero. The servo parameter change information is information indicating the change contents of the servo parameter.

Further, regarding the inertia information of each axis, the servomotor 91 of each axis is connected to the ball screw of the machine via a link mechanism such as a coupling. Further, a mechanical structure is attached to the ball screw, and the attached structure is usually different for each axis. Therefore, even if each servomotor 91 has the same capacity, the inertia applied to each of the servomotors 91 will be different because the structures to be attached are different. In fact, when setting up a new machine, the servo motor is adjusted so that the servo motor can produce an appropriate torque. When making this adjustment, inertia information related to the servo motor is required. The numerical control device 1Y according to the present embodiment has a function of measuring the inertia applied to the servomotor 91 of each axis, and the inertia information measured in advance using this function is stored in the shared area 344. And. It should be noted that it is not essential for the numerical control device 1Y to have an inertia measurement function in order to realize the present invention. When the machine is set up, the operator measures the inertia of each axis, creates inertia information, and stores the created inertia information in the shared area 344 of the storage unit 34 or another area in advance, so that the machine learning device can be used. 50 can perform learning.

The state observation unit 52 outputs a data set, which is the result of data observation, to the learning unit 51. The learning unit 51 learns the vibration waveform correction information based on the data set input from the state observation unit 52. That is, the state observation unit 52 is generated by the comparison unit 383, the inertia information of each axis stored in the common area 344, the change information indicating the change contents of the servo parameters of each axis by the command vibration waveform adjustment unit 384Y. The amount of vibration waveform deviation of each axis is observed as a state variable, and the data set created based on the state variable is output to the learning unit 51. The learning unit 51 learns the vibration waveform correction information based on the data set output from the state observation unit 52. Here, the data set is data in which the vibration amplitude adjustment amount, the inertia information of each axis, the servo parameter change information of each axis, and the vibration waveform deviation amount of each axis are associated with each other. The vibration waveform correction information includes the amplitude correction amount, that is, the vibration amplitude adjustment amount when the command vibration waveform is corrected and brought closer to the FB vibration waveform, and the FB vibration waveform of each axis described in the first embodiment is brought closer to the command vibration waveform. The adjustment contents of the servo parameters to be performed for this purpose are shown.

The machine learning device 50 may be connected to the numerical control device 1X according to the first embodiment via a network, and may be a device separate from the numerical control device 1X. In this case, the machine learning device 50 may exist on the cloud server. Further, the machine learning device 50 may be built in the numerical control device 1Y as shown in FIG.

The learning unit 51 learns the inertia information of each axis, the servo parameter change information of each axis, and the vibration waveform deviation amount of each axis from the data set associated with each other by, for example, supervised learning according to the neural network model. Here, supervised learning refers to a model in which a large number of sets of data of a certain input and a result are given to a learning device, features in those data sets are learned, and the result is estimated from the input. In the numerical control device 1Y according to the present embodiment, the amount of vibration waveform deviation of each axis is the teacher data.

A neural network is composed of an input layer composed of a plurality of neurons, an intermediate layer composed of a plurality of neurons, and an output layer composed of a plurality of neurons. The middle layer is also called the hidden layer. The intermediate layer may be one layer or two or more layers.

For example, in the case of a three-layer neural network, when multiple inputs are input to the input layer, the values are weighted and input to the intermediate layer, and the result is further weighted and output from the output layer. .. This output result depends on the value of each weight.

In the machine learning device 50 according to the present embodiment, the neural network learns the vibration waveform correction information by so-called supervised learning according to the data set generated by the state observation unit 52.

That is, the input layer of the neural network is input with the inertia information of each axis, the servo parameter change information of each axis, and the data set in which the vibration waveform deviation amount of each axis is associated with each other. In the neural network, each time a data set is input, the amount of vibration waveform deviation of each axis approaches 0 when the commanded vibration waveform adjusting unit 384Y adjusts the commanded vibration waveform according to the vibration waveform correction information output from the output layer. By adjusting each of the above weights individually as described above, learning is performed.

The neural network can also learn the vibration waveform correction information by so-called unsupervised learning. In unsupervised learning, a large amount of input data is given to the machine learning device 50 to learn how the input data is distributed, and even if the corresponding teacher output data is not given, the input data is subjected to learning. It is a method of learning a device that performs compression, classification, shaping, and so on. Unsupervised learning can cluster features in the input dataset among similar people. Using this result, it is possible to predict the output by setting some criteria and allocating the output to optimize it.

Further, when the machine learning device 50 is not built in the numerical control device 1Y, the learning unit 51 outputs vibration waveform correction information as a learning result according to the data sets created for the plurality of numerical control devices 1Y. You may. The learning unit 51 may acquire data sets from a plurality of numerical control devices 1Y used at the same site, or may obtain numerical control devices 1Y of a plurality of machine tools that operate independently at different sites. You may use the dataset collected from. In this configuration, it is also possible to add the numerical control device 1Y that collects the data set to the target on the way, or conversely, to separate it from the target. Further, the machine learning device 50 that has been trained using the data set acquired from a certain numerical control device 1Y is attached to another numerical control device 1Y, and the data set is acquired from the other numerical control device 1Y. The learning result may be updated by re-learning.

Further, as the learning algorithm used in the learning unit 51, deep learning that learns the extraction of the feature amount itself can also be used, and other known methods such as genetic programming, functional logic programming, and support can be used. Machine learning may be performed according to a vector machine or the like.

The vibration waveform correction information output by the machine learning device 50 is passed to the command vibration waveform adjusting unit 384Y. The commanded vibration waveform adjusting unit 384Y adjusts the commanded vibration waveform of each axis according to the vibration waveform correction information.

As described above, the numerical control device 1Y according to the present embodiment uses the inertia information of each axis, the vibration waveform deviation amount of each axis, the vibration amplitude adjustment amount, and the change information of the servo parameters of each axis. A machine learning device 50 for performing machine learning is provided. The commanded vibration waveform adjusting unit 384Y adjusts the commanded vibration waveform of each axis according to the vibration waveform correction information generated by machine learning by the machine learning device 50. As a result, even if the responsiveness of the servomotor 91 of each control target axis is different, the chips can be reliably divided.

The control calculation unit 2Y and the machine learning device 50 included in the numerical control device 1Y according to the second embodiment are the processors shown in FIG. 14, similar to the control calculation unit 2X included in the numerical control device 1X according to the first embodiment. This can be achieved with 101 and memory 102.

The configuration shown in the above-described embodiment shows an example of the content of the present invention, can be combined with another known technique, and is one of the configurations without departing from the gist of the present invention. It is also possible to omit or change the part.

1X, 1Y numerical control device, 2X, 2Y control calculation unit, 3 input operation unit, 4 display unit, 31 screen processing unit, 32 input control unit, 33 data setting unit, 34 storage unit, 35 control signal processing unit, 36 PLC , 37 Analysis processing unit, 38X, 38Y interpolation processing unit, 39 acceleration / deceleration processing unit, 40-axis data input / output unit, 50 machine learning device, 51 learning unit, 52 state observation unit, 90 drive unit, 91 servo motor, 92, 95 Detector, 93X X-axis servo control unit, 93Y Y-axis servo control unit, 94 spindle motor, 96 spindle control unit, 341 parameters, 342 display data, 343 machining program, 344 shared area, 371 movement command analysis unit, 372 vibration Command analysis unit, 381 distribution ratio determination unit, 382 waveform information acquisition unit, 383 comparison unit, 384X, 384Y command vibration waveform adjustment unit, 385 vibration waveform generation unit, 386 vibration movement amount calculation unit.

In order to solve the above-mentioned problems and achieve the object, the present invention is a numerical control device that controls a machine tool that controls a plurality of drive shafts to move a tool according to a machining program, and vibrates the tool. While processing the work, the vibration waveform generator that generates the vibration waveform showing the vibration path of the tool and the vibration waveform are at a constant angle with respect to the first axis, which is one of the plurality of drive axes. As a vibration waveform in the direction of the second axis, a vibration movement amount calculation unit for generating a command vibration waveform for each drive shaft indicating a tool movement path in each of the plurality of drive shafts is provided.

Claims (8)

A numerical control device that controls a machine tool that controls multiple drive shafts to move tools.
A vibration waveform generator that generates a vibration waveform that indicates the vibration path of the tool when the work is machined while vibrating the tool.
In each of the plurality of drive shafts, based on the vibration direction of the tool when the tool is vibrated, the angle in the direction of the reference axis which is one of the plurality of drive shafts, and the vibration waveform. A vibration movement amount calculation unit that generates a command vibration waveform for each drive shaft indicating the movement path of the tool, and a vibration movement amount calculation unit.
A numerical control device characterized by being provided with. A distribution ratio determining unit that determines a distribution ratio for distributing a component of the vibration waveform to each of the plurality of drive shafts based on an angle between the vibration direction of the tool and the direction of the reference axis when the tool is vibrated.
With
The vibration movement amount calculation unit distributes the components of the vibration waveform to each of the plurality of drive shafts according to the distribution ratio to generate the command vibration waveform of each drive shaft.
The numerical control device according to claim 1. The vibration movement amount calculation unit generates the movement amount per unit time of the tool when machining the work for each drive shaft by superimposing the command vibration waveform on the movement amount of the tool.
The numerical control device according to claim 1 or 2. A comparison unit that compares the feedback vibration waveform for each drive shaft, which indicates the actual movement path of the tool when the tool is vibrated, with the command vibration waveform for each drive shaft.
A command vibration waveform adjusting unit that adjusts the command vibration waveform for each drive shaft based on the comparison result by the comparison unit, and a command vibration waveform adjusting unit.
The numerical control device according to any one of claims 1 to 3, wherein the numerical control device is provided. The command vibration waveform adjusting unit adjusts the command vibration waveform by changing the servo parameters of each drive shaft.
The numerical control device according to claim 4. The comparison result by the comparison unit, the change information indicating the vibration amplitude adjustment amount of each drive shaft when the command vibration waveform adjustment unit adjusts the command vibration waveform, and the change contents of the servo parameters of each drive shaft, and the change information. A state observing unit that observes the inertia information of each of the plurality of drive axes as a state variable, and
A learning unit that learns vibration waveform correction information indicating the adjustment content of the vibration amplitude of each drive shaft and the adjustment content of the servo parameter of each drive shaft according to the data set created based on the state variable.
The numerical control device according to claim 5, wherein the numerical control device is provided. It is a numerical control method executed by a numerical control device that controls a machine tool that controls multiple drive axes to move tools.
A vibration waveform generation step that generates a vibration waveform indicating a vibration path of the tool when machining a work while vibrating the tool, and a vibration waveform generation step.
In each of the plurality of drive shafts, based on the vibration direction of the tool when the tool is vibrated, the angle in the direction of the reference axis which is one of the plurality of drive shafts, and the vibration waveform. A command vibration waveform generation step for generating a command vibration waveform for each drive shaft indicating the movement path of the tool, and a command vibration waveform generation step.
A numerical control method characterized by including. The direction of the reference axis, which is one of the multiple drive shafts controlled by the numerical control device that controls the machine tool when moving the tool, is different from the vibration direction of the tool when the tool is vibrated. In this case, it is a machine learning device that learns adjustment contents for adjusting a command vibration waveform for each drive shaft indicating a vibration path of the tool in each of the plurality of drive shafts when machining a work while vibrating the tool. hand,
The comparison result of the feedback vibration waveform for each drive shaft and the command vibration waveform for each drive shaft showing the actual movement path of the tool when the tool is vibrated, and the command vibration waveform for each drive shaft are shown. A state observation unit that observes the amount of vibration amplitude adjustment of each drive shaft at the time of adjustment, change information indicating the change contents of the servo parameters of each drive shaft, and inertia information of each of the plurality of drive shafts as state variables. When,
A learning unit that learns vibration waveform correction information indicating the adjustment content of the vibration amplitude of each drive shaft and the adjustment content of the servo parameter of each drive shaft according to the data set created based on the state variable.
A machine learning device characterized by being equipped with.

Images