JP6896196B1 - Numerical control device and learning device - Google Patents

Abstract

The command generator (11) that analyzes the machining program (12) and generates an operation command signal, which is a control condition for controlling the motor (4) of the machine tool (2), and the vibration of the machine tool (2). A drive control unit that controls the motor (4) based on the first fluctuation condition during the fluctuation section when the vibration determination unit (5) for determining the presence or absence and the vibration determination unit (5) determine that vibration is present. (6), the actual fluctuation time which is the time when the motor (4) is controlled based on the first fluctuation condition, and the maximum value of the temperature of the motor (4) in the actual fluctuation time are input, and the motor (4) It is provided with a learning unit (10) that generates a fluctuation condition model so as to estimate a second fluctuation condition in which the maximum value of the temperature of the machine tool (2) can be suppressed within the permissible temperature of the motor (4). It is a feature.

Description

This disclosure relates to numerical control devices and learning devices.

A machine tool is a device that processes a work by applying force or energy to the work using a tool. One of the processing defects in machine tools is a phenomenon called chatter vibration. This is a phenomenon in which the tool or the structure of the work or machine tool body vibrates due to the cutting resistance generated when the tool and the work come into contact with each other. As a method of suppressing chatter vibration, a method of varying the spindle rotation speed is known. By fluctuating the rotation speed of the spindle, the generation of vibration becomes non-uniform, and the excitation of chatter vibration can be hindered. When the spindle rotation speed is changed, a waveform in which a periodic waveform such as a triangular wave or a sine wave is superimposed on the reference rotation speed waveform of the spindle before the fluctuation is generally set. Therefore, in order to fluctuate the spindle speed, it is necessary to determine the combination of the amplitude and frequency of the overlapping periodic waveforms. In the following, the amplitude and frequency superimposed for fluctuation are referred to as fluctuation amplitude and fluctuation frequency, respectively, and the combination of both is referred to as fluctuation condition.

The method described in Patent Document 1 has been proposed as a method of changing the spindle speed of a machine tool after determining the fluctuation conditions. The method described in Patent Document 1 uses information related to motor drive including the rated output, and presets the ratio of the motor output used for fluctuation of the spindle rotation speed to the rated output of the motor. Therefore, the fluctuation condition with the ratio as the upper limit can be set.

JP 2012-130983

Here, the relationship between the temperature rise of the motor and the fluctuation condition will be described. Repeated fluctuations in the spindle speed of the spindle motor to suppress chatter vibration cause the motor temperature to rise. If the motor temperature continues to rise, it will lead to deterioration and failure of the motor, so it must be operated within a range that does not affect the motor. On the other hand, it is known that it is necessary to give a sufficiently large fluctuation amplitude and fluctuation frequency to the motor in order to suppress chatter vibration. Therefore, there is a trade-off relationship between the suppression of the motor temperature rise and the suppression of chatter vibration.

When the fluctuation conditions are the same, the temperature rise of the motor differs depending on the length of the fluctuation time. When the time for changing the spindle speed is short, there is a margin in the difference between the permissible value of the motor temperature and the actual motor temperature as compared with the case where the time is long. Therefore, the fluctuation amplitude and the fluctuation frequency can be set to higher values than when the fluctuation time is long. In other words, when the time for fluctuating the spindle speed is short, it is possible to set a fluctuation condition having a higher suppressing effect on chatter vibration than when the fluctuation time is long. However, since the method described in Patent Document 1 does not consider the time for fluctuating the spindle speed, the limitation of the fluctuation condition is uniquely provided based on the motor output. Therefore, there is a problem that it is not possible to select a fluctuation condition having a high suppression effect on chatter vibration even when the time for fluctuating the spindle speed is short.

Therefore, the present disclosure has been made in view of the above problems, and is a numerical value capable of controlling the motor under control conditions in which fluctuation conditions that are within the permissible range of the motor temperature and have a high suppression effect on chatter vibration are superimposed. The purpose is to provide a control device.

This disclosure includes a command generator that analyzes a machining program and generates an operation command signal, which is a control condition for controlling a machine tool motor, a vibration determination unit that determines the presence or absence of vibration of the machine tool, and a vibration determination. When the unit determines that there is vibration, the drive control unit that controls the motor based on the first fluctuation condition and the actual fluctuation time that is the time that the motor is controlled based on the first fluctuation condition during the fluctuation section. And the maximum value of the motor temperature during the actual fluctuation time are input, and the fluctuation condition model is used to estimate the second fluctuation condition that can suppress the vibration of the machine tool within the maximum value of the motor temperature within the allowable temperature of the motor. It is provided with a learning unit to be generated.

The numerical control device according to the present disclosure can operate the motor under the optimum fluctuation conditions that can suppress chatter vibration within the allowable range of the motor temperature.

A block diagram showing a configuration of a numerical control device according to the first embodiment. The figure which shows the relationship between the fluctuation condition model of the numerical control apparatus which concerns on Embodiment 1 and the input / output signal of a learning stage. The figure which shows the relationship between the fluctuation condition model of the numerical control apparatus which concerns on Embodiment 1 and the input / output signal of an inference stage. Flow chart of the learning stage of the numerical control device according to the first embodiment Flow chart of the estimation stage of the numerical control device according to the first embodiment The figure which shows an example of the hardware configuration of the numerical control device which concerns on Embodiment 1. Diagram showing the difference in the range in which fluctuation conditions can be set due to the difference in the length of the fluctuation section. Diagram showing the difference between the spindle speed and the motor temperature due to the difference in the fluctuation section length A block diagram showing the configuration of the numerical control device according to the second embodiment. A block diagram showing a configuration of a numerical control device according to a third embodiment.

Embodiment 1.
FIG. 1 is a block diagram showing a configuration of a numerical control device 1 according to a first embodiment of the present disclosure. The numerical control device 1 of the first embodiment transmits a control signal to the spindle motor 4 provided in the machine tool 2 based on the machining program 12 given from the outside, and relatively moves a tool or a workpiece (not shown). The work is processed by letting it work. The machine tool 2 is provided with a sensor 3 for detecting vibration generated in the machine tool 2 and a spindle motor 4 for relatively moving a tool or a workpiece. The sensor 3 is, for example, an encoder for positioning provided in advance in the machine tool 2. Further, as another example, an acceleration sensor, a force sensor, or a microphone additionally provided in the machine tool 2 may be used.

The numerical control device 1 has a vibration determination unit 5 that receives a sensor signal from a sensor 3 provided in the machine tool 2, a command generation unit 11 that analyzes a machining program 12 and generates an operation command signal, and a command generation unit 11. The drive control unit 6 that receives the operation command signal and the fluctuation condition from the main shaft motor 4 and transmits the control signal to the spindle motor 4, the temperature prediction unit 7 that receives the fluctuation section signal from the command generation unit 11, and the temperature prediction unit 7 continuously determine the continuation. It includes a fluctuation condition setting unit 8 that receives a signal and a fluctuation section length and transmits a fluctuation condition signal to the drive control unit 6, and a learning unit 10 that learns the fluctuation condition.

The machining program 12 describes machining conditions and a machining path for machining the work into a predetermined shape. Here, the machining conditions are the spindle speed, the feed rate relative to the work of the tool, and the depth of cut of the tool with respect to the work. The machining path is a path followed by the tool and is time-series data. Further, in the machining program 12, when chatter vibration occurs, fluctuation conditions for suppressing the chatter vibration are described. Further, the machining program 12 describes a variable section command which is a command indicating a section in which the spindle rotation speed is changed. The variable section command is described so as to be a section in which the spindle speed fluctuates in a series of machining blocks corresponding to an arbitrary machining section. Therefore, if chatter vibration occurs during the fluctuation section, the fluctuation of the motor continues until the fluctuation section ends. That is, when chatter vibration occurs during the fluctuation section, the motor is operated in a state where the initial processing conditions are changed to the fluctuation conditions until the fluctuation section ends.

Hereinafter, the operation of each part of the numerical control device 1 of the first embodiment will be described. The vibration determination unit 5 receives a sensor signal from the sensor 3 of the machine tool 2, determines the presence or absence of chatter vibration in the machine tool 2, and transmits the determination result as a vibration detection signal to the drive control unit 6.

The command generation unit 11 analyzes the command described in the machining program 12, generates an operation command signal, and transmits the operation command signal to the drive control unit 6. The operation command signal is a signal for controlling the spindle motor 4 according to the machining program 12. In addition, when the command generation unit 11 reads the fluctuation section command in the machining program 12, it transmits a fluctuation section signal to the temperature prediction unit 7 and sends the fluctuation conditions described in the machining program 12 to the drive control unit 6. Send. The fluctuation condition transmitted from the command generation unit 11 is the first fluctuation condition. The first fluctuation condition is a condition relating to motor control superimposed on the operation command signal during the fluctuation section. The variable section command is information on the variable section length and is described in the machining program 12. The variable interval length can be derived from the variable interval command.

The temperature prediction unit 7 receives the motor temperature signal and the motor current from the drive control unit 6. Further, the temperature prediction unit 7 receives the fluctuation section signal from the command generation unit 11 and determines whether or not the motor fluctuation can be continued during the fluctuation section based on the motor temperature signal. The temperature prediction unit 7 is a temperature determination unit. The temperature prediction performed by the temperature prediction unit 7 is performed based on the mathematical formula 1.

Here, n is the number of time series data indicating the time, Δt is the sampling period, T is the temperature time constant, τ is the temperature, R is the heat coefficient, and I is the motor current required to control the motor. .. Equation 1 represents the transition of the motor temperature when the motor current is generated. Since the speed of the fluctuating motor changes periodically, the motor current for generating acceleration / deceleration also changes periodically. Therefore, if the motor current during fluctuation for one cycle and the initial motor temperature (motor temperature at the start of fluctuation) are given to Equation 1, the temperature at each time can be derived. The motor temperature signal received by the temperature prediction unit 7 from the drive control unit 6 is, for example, the temperature of the motor at time 0 (the first of the time series data). By designating the temperature of the motor at a certain time (for example, time 0) by the motor temperature signal, the temperature prediction unit 7 can calculate the temperature at each time from Equation 1. The motor temperature at a certain time is actually acquired by the temperature sensor.

The temperature prediction unit 7 calculates the fluctuation section length from the fluctuation section signal, and calculates the temperature at each time up to the fluctuation section length. Then, the temperature prediction unit 7 compares the calculated temperature with the preset allowable motor temperature. The temperature prediction unit 7 predicts the motor temperature during the fluctuation section (from a certain time until the length of the fluctuation section elapses) by calculating based on Equation 1, and the maximum motor temperature during the fluctuation section is the motor tolerance. If the temperature is below the temperature, it is determined that the motor fluctuation in the fluctuation section can be continued, and if not, it is determined that the motor fluctuation in the fluctuation section cannot be continued. The temperature prediction unit 7 transmits the continuation determination signal, which is the result of the determination, and the fluctuation section length to the fluctuation condition setting unit 8.

In this case, the motor temperature is calculated by the temperature prediction unit 7 according to the mathematical formula 1, but the method is not limited to this. For example, it may be calculated according to another mathematical formula.

The allowable motor temperature is a value determined according to the mechanical and electrical design specifications of the spindle motor 4, and is an upper limit capable of continuing the operation of the spindle motor 4 without causing an abnormality in the spindle motor 4. Represents temperature. The allowable motor temperature is registered in the storage unit of the numerical control device 1 in advance before the start of the numerical control device 1, and the temperature prediction unit 7 reads the allowable motor temperature in the storage unit when calculating the temperature. Alternatively, the allowable motor temperature may be described in the machining program 12.

The drive control unit 6 controls the spindle motor 4 based on the machining program 12 by the operation command signal from the command generation unit 11. The drive control unit 6 receives a vibration detection signal from the vibration determination unit 5, and when the vibration determination unit 5 determines that chatter vibration has occurred, the fluctuation condition transmitted from the command generation unit 11 or the fluctuation condition , The spindle rotation speed is changed according to the fluctuation condition (hereinafter referred to as the second fluctuation condition) transmitted from the fluctuation condition setting unit 8, and this is continued until the fluctuation section ends. When the fluctuation condition signal is input from the fluctuation condition setting unit 8, the drive control unit 6 changes the fluctuation condition from the condition described in the machining program 12 to the second fluctuation condition set by the fluctuation condition setting unit 8. To do. The second variable condition is a variable condition different from the condition described in the machining program 12.

In addition, the drive control unit 6 outputs the learning data signal to the learning unit 10 and outputs the inference data signal to the fluctuation condition setting unit 8. The learning data signal is a data signal for constructing the fluctuation condition model 9 described later, and the inference data signal is a data signal for calculating the fluctuation condition from the fluctuation condition model 9 described later. The learning data is data composed of machining condition information, motor specification information, motor state information, fluctuation time information, and fluctuation condition information. The inference data is data composed of machining condition information, motor specification information, and motor state information.

The fluctuation condition setting unit 8 receives the continuation determination signal and the fluctuation section length from the temperature prediction unit 7, and receives the inference data signal from the drive control unit 6. When the temperature prediction unit 7 determines that the motor fluctuation during the fluctuation section is impossible, the fluctuation condition setting unit 8 inputs the fluctuation section length and the inference data signal into the fluctuation condition model 9 generated by the learning unit 10. By doing so, it is possible to output the second fluctuation condition inferred from the fluctuation interval length and the inference data signal. The second fluctuation condition is a condition in which the motor can fluctuate during the fluctuation section, that is, the fluctuation described in the machining program in order to suppress the chatter vibration of the motor within the allowable temperature range of the motor during the fluctuation section. It is a condition superimposed on the motor operating condition outside the section. The second fluctuation condition is output to the drive control unit 6 by the fluctuation condition setting unit 8 as a fluctuation condition signal.

As the learning algorithm used by the learning unit 10, known algorithms such as supervised learning and reinforcement learning can be used. As an example, a neural network, which is a supervised learning method, can be applied.

Here, the fluctuation condition model 9, the learning data and the inference data used for the input / output of the model, and the fluctuation time information will be described in detail. The fluctuation condition model 9 is the neural network shown in FIGS. 2 and 3. FIG. 2 shows the input / output relationship of the fluctuation condition model 9 when the learning unit 10 learns the fluctuation condition model 9 using the learning data, and FIG. 3 shows the inference data and the fluctuation section length of the fluctuation condition setting unit 8 in FIG. Represents the input / output relationship of the fluctuation condition model 9 when setting the fluctuation condition using. A neural network is composed of an input layer composed of one or more nodes, an intermediate layer (hidden layer) composed of one or more nodes, and an output layer composed of one or more nodes. The number of intermediate layers may be one layer or two or more layers.

The input / output relationship of the variable condition model 9 in the case of learning by the learning unit 10 will be described with reference to FIG. In FIG. 2, the input of the variable condition model 9 is the tool number and the work number representing the machining condition information, the spindle speed, the feed rate and the depth of cut, the motor model number and the temperature time constant representing the motor specification information, and the motor state information. The maximum motor temperature representing the above and the actual fluctuation time representing the fluctuation time information. Here, the values acquired when chatter vibration does not occur are given to the spindle rotation speed, the feed rate, and the depth of cut, which are the machining condition information. The spindle speed and feed rate are given values in a constant speed state before superimposing the fluctuating waveform.

The motor model number and temperature time constant, which are the motor specification information, are given information specific to the motor regardless of the machining program 12. The maximum motor temperature, which is the motor state information, is given the maximum value of the motor temperature while the spindle motor 4 is fluctuating. For the actual fluctuation time, which is the fluctuation time information, the actual time during which the spindle motor 4 continues to fluctuate is used. The fluctuation condition setting unit 8 calculates the actual fluctuation time by calculating the time length at which the continuation possible signal is output from the continuation determination signal.

Further, in FIG. 2, the output of the fluctuation condition model 9 is the fluctuation amplitude and the fluctuation frequency, which are the fluctuation condition information. Further, in FIG. 2, there is a teacher signal for learning, and the fluctuation amplitude and the fluctuation frequency included in the learning data are given to the fluctuation condition model 9. The learning unit 10 compares the output of the fluctuation condition model 9 with the teacher signal included in the learning data, and optimizes the weighting coefficient inside the fluctuation condition model by a known method such as an error back propagation method. When the learning unit 10 performs the learning process of the fluctuation condition model 9 using the above learning data, chatter vibration can be suppressed when chatter vibration occurs under a combination of a certain processing condition and the motor specifications. The relationship between the condition and the motor state and the fluctuation time corresponding to the fluctuation condition can be constructed in the fluctuation condition model 9.

Next, the input / output relationship of the fluctuation condition model 9 when the fluctuation condition setting unit 8 calculates the second fluctuation condition will be described with reference to FIG. In FIG. 3, a part of the input of the variable condition model 9 is the tool number and the work number representing the machining condition information, the spindle rotation speed, the feed rate and the depth of cut, and the motor model number and the temperature representing the motor specification information as in FIG. It is a time constant. Other inputs are different from the case of FIG. 2, in which the motor allowable temperature is input as the motor state information and the fluctuation section length is input as the fluctuation time information. The fluctuating section length is the length of time for fluctuating the spindle rotation speed by the fluctuating section command described in the machining program 12. The fluctuation section length is calculated by the fluctuation condition setting unit 8, and can be calculated from the feed rate and the machining path in the machining program 12 included in the fluctuation section signal.

Further, in FIG. 3, the output of the fluctuation condition model 9 is the fluctuation condition information as in FIG. Since learning is not performed in the processing of the variable condition setting unit 8, the teacher signal existing in FIG. 2 does not exist. Further, here, the fluctuating section length is calculated from the feed rate and the machining path in the machining program 12 included in the fluctuating section signal, but the method is not limited to this method. For example, the command generation unit 11 may obtain the fluctuation section length and transmit the fluctuation section length as a fluctuation section signal to the temperature prediction unit 7 and the fluctuation condition setting unit 8.

Chatter vibration is generated by performing the inference processing of the fluctuation condition model 9 using the inference data received from the drive control unit 6 by the fluctuation condition setting unit 8 and the fluctuation section length received from the temperature prediction unit 7. For the combination of the machining conditions and the motor specifications, it is possible to calculate the second fluctuation condition that does not exceed the motor allowable temperature within the fluctuation section described in the machining program 12.

The processing flow of the numerical control device 1 of the first embodiment described above will be described with reference to FIGS. 4 and 5. FIG. 4 is a processing flow of the learning stage executed by the learning unit 10 in order to construct the variation condition model 9, and FIG. 5 is an inference performed by the variation condition setting unit 8 and using the generated variation condition model 9. It is a step processing flow. In FIGS. 4 and 5, the same processing points are assigned the same numbers. The numerical control device 1 executes the processing flow of FIG. 4 before the learning of the variable condition model 9 is completed, and executes the processing flow of FIG. 5 after the learning of the variable condition model 9 is completed. Hereinafter, the processing flow at the learning stage will be described with reference to FIG. The numerical control device 1 generates the variable condition model 9 by executing the following processes for various machining programs 12.

In step S111, the vibration determination unit 5 receives the sensor signal from the sensor 3, determines the presence or absence of chatter vibration of the machine tool 2, and transmits the vibration detection signal, which is the determination result, to the drive control unit 6. Here, a known method may be used for the determination performed by the vibration determination unit 5. For example, it is determined that chatter vibration has occurred when the vibration amplitude in a predetermined time region exceeds the determination threshold value. As another example, it may be determined that chatter vibration has occurred when the component of the maximum peak in a predetermined frequency region exceeds the determination threshold value. If chatter vibration does not occur, this flow ends, and if chatter vibration occurs, the process proceeds to S112.

In step S112, the drive control unit 6 changes the spindle speed of the spindle motor 4 by using the fluctuation condition received from the command generation unit 11. The spindle speed of the spindle motor 4 is a control condition of the spindle motor 4. That is, the drive control unit 6 changes the control conditions of the spindle motor 4. In step S113, the learning unit 10 collects the learning data signal transmitted from the drive control unit 6 and builds a fluctuation condition model. The learning data signal collected by the learning unit 10 includes processing condition information when it is determined that chatter vibration has not occurred, motor specification information, and the motor temperature predicted by Equation 1 is equal to or lower than the motor allowable temperature. The fluctuation condition information when the chatter vibration is successfully suppressed, the motor state information while the fluctuation condition is superimposed on the machining condition, and the fluctuation time information which is the actual fluctuation time in which the fluctuation condition is continued. .. The learning unit 10 collects the learning data signal and learns the input / output relationship of the fluctuation condition model 9 using the data extracted from the signal.

In step S114, the learning unit 10 generates the fluctuation condition model 9 from the collected learning data. The learning unit 10 receives a learning data signal from the drive control unit 6, and suppresses chatter vibration when the motor temperature is equal to or lower than the motor allowable temperature based on the combination of processing condition information, motor specification information, motor state information, and fluctuation time information. The variable conditions that can be created are learned, and the variable condition model 9 is generated as the learning result. Here, the learning data signal is data in which machining condition information, motor specification information, motor state information, fluctuation time information, and fluctuation condition information are associated with each other. The input / output relationship is learned so that the fluctuation condition can be generated from the motor state information, the motor specification information, and the fluctuation section length by a known method such as the error back propagation method.

In the present embodiment, the learning unit 10 is provided inside the numerical control device 1 and is used for learning variable conditions. For example, the learning unit 10 is used as an external device separate from the numerical control device 1 via a network. It may be connected to the numerical control device 1. Further, the learning unit 10 may exist on the cloud server. Further, although the fluctuation condition setting unit 8 has been described as outputting the second fluctuation condition using the fluctuation condition model 9 learned by the learning unit 10 of the numerical control device 1, the network is connected from the outside of the numerical control device 1. The fluctuation condition model 9 may be acquired through the system, and the second fluctuation condition may be output based on the fluctuation condition model 9.

The processing flow at the inference stage will be described below with reference to FIG. Steps S111, S112, and S115 are the same as in FIG. In step S123, the temperature prediction unit 7 calculates the fluctuation section length, and uses Equation 1 to determine whether the fluctuation of the spindle motor 4 can be continued within the fluctuation section. That is, it is determined whether or not the motor temperature predicted by using Equation 1 is within the allowable motor temperature. If it is determined that the fluctuation within the fluctuation section can be continued, that is, if the predicted motor temperature within the fluctuation section is within the allowable motor temperature, the process proceeds to step S115. If this is not the case, that is, if the predicted motor temperature in the fluctuation section exceeds the motor allowable temperature, the process proceeds to step S124.

In step S124, the fluctuation condition setting unit 8 calculates the second fluctuation condition from the learning data signal using the fluctuation condition model 9 generated by the learning unit 10. In the variable condition model 9, the input / output relationship in which the learning by the learning unit 10 is completed is constructed. The fluctuation condition setting unit 8 calculates the fluctuation condition corresponding to the inference data by giving the inference data and the fluctuation time information to the fluctuation condition model 9. The fluctuation condition setting unit 8 receives processing condition information when it is determined that chatter vibration is occurring, motor specification information, and motor state information indicating the allowable motor temperature from the drive control unit as inference data signals. get. The calculated fluctuation condition is the second fluctuation condition. The fluctuation condition setting unit 8 transmits the second fluctuation condition to the drive control unit 6. The drive control unit 6 fluctuates the spindle rotation speed of the spindle motor 4 by superimposing this fluctuation condition on the rotation speed waveform before the fluctuation. In step S115, the drive control unit 6 ends the fluctuation of the spindle motor 4 at the end of the fluctuation section.

FIG. 6 is a diagram showing an example of the hardware configuration of the numerical control device according to the first embodiment. As shown in FIG. 6, the numerical control device 1 includes a computer including a processor 101, a memory 102, and an interface circuit 103.

The processor 101, the memory 102, and the interface circuit 103 can send and receive information to and from each other by, for example, the bus 104. The variation condition model 9 is stored in the memory 102. The drive control unit 6 is realized by the interface circuit 103. The processor 101 executes functions such as a vibration determination unit 5, a temperature prediction unit 7, a fluctuation condition setting unit 8, a learning unit 10, and a command generation unit 11 by reading and executing a program stored in the memory 102. The processor 101 is, for example, an example of a processing circuit, and is a CPU (Central Processing Unit), a DSP (Digital Signal Processor), and a system LSI (Large).
Includes one or more of Scale Integration).

The memory 102 includes one or more of RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), and EEPROM (registered trademark) (Electrically Erasable Programmable Read Only Memory). Including. The memory 102 also includes a recording medium on which a computer-readable program is recorded. Such recording media include one or more of non-volatile or volatile semiconductor memories, magnetic disks, flexible memories, optical disks, compact discs, and DVDs (Digital Versatile Discs). The numerical control device 1 includes an ASIC (Application Specific Integrated Circuit) and an FPGA (Field).
It may include integrated circuits such as Programmable Gate Array).

Generally, in order to suppress chatter vibration, it is necessary to raise the fluctuation amplitude and the fluctuation frequency sufficiently. However, if the fluctuation amplitude and the fluctuation frequency are too high, the load on the motor due to acceleration / deceleration increases and the motor becomes overheated, so that normal operation cannot be maintained for the motor. Here, the relationship between the length of the fluctuation section, the settable range of the fluctuation condition, and the tendency of the motor temperature to rise will be described with reference to FIGS. 7 and 8. FIG. 7 is a diagram showing a difference in the range in which the fluctuation condition can be set due to the difference in the length of the fluctuation section. FIG. 7A is a range of fluctuation conditions that can be set when the fluctuation section is long, and FIG. 7B is a range of fluctuation conditions that can be set when the fluctuation section is short. In both figures, if a fluctuation condition outside the variable condition settable range is given, the load on the motor increases, and the motor falls into an overheated state. The fluctuation condition setting range b when the fluctuation section of FIG. 7B is short is wider than the fluctuation condition setting range a when the fluctuation section of FIG. 7A is long. This is because when the fluctuation section is short, the time for the motor temperature to rise is also short, so that the fluctuation condition with a higher motor load can be set as compared with the case of FIG. 7A in which the fluctuation section is long.

The difference in the upward tendency of the motor temperature when the spindle speed fluctuates under the fluctuation condition a and the fluctuation condition b shown in FIG. 7 will be described with reference to FIG. FIG. 8A shows the spindle speed and the motor temperature when the spindle speed fluctuates under the fluctuation condition a when the fluctuation section is long, and FIG. 8B shows the spindle under the fluctuation condition b when the fluctuation section is short. It is the spindle speed and the motor temperature when the rotation speed fluctuates. In both figures, the spindle speed fluctuates between the chatter vibration detection time and the end time of the fluctuation section. In FIG. 8A, since the actual fluctuation section is long, the temperature rise is a gradual fluctuation condition, and in FIG. 8B, the actual fluctuation section is short, so that the temperature rise is a steep fluctuation condition. In this way, the numerical control device 1 collects the learning data when the chatter vibration is suppressed within the range where the motor does not become overheated, and constructs the fluctuation condition model 9. Therefore, the allowable range of the motor temperature is used by using the inference data. It is possible to calculate the fluctuation condition within the variable condition settable range.

As described above, the numerical control device 1 of the first embodiment detects chatter vibration in the machine tool 2 and fluctuates the spindle motor 4 under the fluctuation conditions described in the machining program 12. Let them learn the fluctuation conditions. Further, it is determined whether or not the fluctuation of the spindle motor 4 can be continued within the fluctuation section based on the motor temperature, and if it is determined that the fluctuation cannot be continued, the maximum motor calculated from the fluctuation condition model 9 constructed by learning. Change to the second fluctuation condition so that the temperature falls within the allowable motor temperature. Therefore, the numerical control device 1 of the first embodiment has the effect that the spindle motor 4 can be automatically changed to the second fluctuation condition effective for suppressing chatter vibration without becoming overheated. Play.

The numerical control device 1 learns by using the maximum motor temperature for the motor state information at the time of learning, and infers using the motor allowable temperature for the motor state information at the time of inference. As a result, the numerical control device 1 can calculate a fluctuation condition in which the motor temperature does not exceed the allowable motor temperature during the fluctuation from the fluctuation condition model 9 at the time of inference, and can fluctuate the motor under the fluctuation condition. This fluctuation condition is calculated from the machining condition information, the motor specification information, and the fluctuation time information in addition to the motor state information.

Therefore, regardless of the command content of the spindle motor 4 described in the machining program 12, it is possible to automatically calculate the fluctuation conditions that can suppress the vibration within the fluctuation section according to the specifications of the motor. As a result, the numerical control device 1 controls the motor under a high load fluctuation condition in which the motor temperature tends to rise when the fluctuation section is short, and a low load fluctuation in which the motor temperature does not easily rise when the fluctuation section is long. The motor can be controlled under the conditions. That is, the numerical control device 1 has an effect that chatter vibration can be suppressed under appropriate load fluctuation conditions according to the fluctuation section length.

Further, in the first embodiment, the numerical control device 1 generates a command for superimposing a fluctuation waveform on one spindle motor, but the same applies even when the rotation speeds of both the spindle and the feed shaft are varied. It can be effective. Further, in the first embodiment, the numerical control device 1 has a configuration in which vibration determination is performed on one sensor 3, but even two or more sensors 3 can perform processing in the same configuration. .. When two or more sensors 3 are set in the machine tool 2, it may be determined that vibration is generated when the output value of at least one sensor 3 exceeds the determination threshold value.

Further, the learning unit 10 may create the fluctuation condition model 9 according to the learning data generated in not only the numerical control device 1 but also other numerical control devices. In this case, it is also possible to switch the numerical control device for collecting the learning data for each machining program. Further, the fluctuation condition model 9 created by the numerical control device 1 may be transmitted to another numerical control device to relearn the fluctuation condition model 9.

Further, in the numerical control device 1 of the first embodiment, the fluctuation condition is a combination of the fluctuation amplitude and the fluctuation frequency, but the fluctuation amplitude and the fluctuation period may be combined.

Embodiment 2.
The first embodiment is configured to suppress chatter vibration generated in the fluctuation section commanded by the machining program 12. That is, the fluctuation section was specified in advance from the outside. In the present embodiment, the machining program 12b does not describe the fluctuation section, and the section in which the chatter vibration occurs when the machining program 12b is executed one or more times is stored, and when the machining program 12b is executed thereafter, the section is stored. By changing the rotation speed of the motor with respect to the memorized section, chatter vibration is suppressed.

FIG. 9 is a block diagram showing a configuration of the numerical control device 1b according to the second embodiment of the present disclosure. In FIG. 9, the same reference numerals are given to the components that are the same as or equivalent to the components of the first embodiment shown in FIG. In the second embodiment, the vibration storage unit 40 has an additional configuration as compared with the first embodiment. Further, the processing contents of the vibration determination unit 5b, the drive control unit 6b, the command generation unit 11b, the fluctuation condition setting unit 8b, and the temperature prediction unit 7b are different from the corresponding processes of the first embodiment, and the machining program 12b is also executed. It is different from that of Form 1. Here, a part different from the first embodiment will be mainly described.

Similar to the first embodiment, the machining program 12b includes machining conditions and machining paths for machining the work into a predetermined shape, and variable conditions for suppressing chatter vibration when chatter vibration occurs. is described. However, unlike the first embodiment, the machining program 12b does not have a description of the variable section command. The command generation unit 11b analyzes the command described in the machining program 12b and transmits an operation command signal to the drive control unit 6b. Further, the command generation unit 11b transmits the fluctuation condition described in the machining program 12b to the drive control unit 6b. The vibration determination unit 5b receives the sensor signal as in the first embodiment, and determines the presence or absence of chatter vibration in the machine tool 2. The transmission destination of the resulting vibration detection signal is different from that of the first embodiment, and is the drive control unit 6b and the vibration storage unit 40 described later.

The vibration storage unit 40 receives and stores the vibration detection signal output by the vibration determination unit 5b. Further, the vibration storage unit 40 outputs the fluctuation section signal to the drive control unit 6b, the temperature prediction unit 7b, and the fluctuation condition setting unit 8b after the numerical control device 1b executes the operation of the machining program 12b one or more times. The fluctuation section signal is a vibration detection signal stored inside the vibration storage unit 40, and is a signal capable of calculating the time length of the section in which the occurrence of chatter vibration is determined.

The temperature prediction unit 7b receives the fluctuation section signal from the vibration storage unit 40, and uses Equation 1 as in the temperature prediction unit 7 of the first embodiment to determine whether or not the maximum motor temperature in the fluctuation section exceeds the allowable motor temperature. Is determined, and the determination result is output to the fluctuation condition setting unit 8b as a continuation determination signal. Here, the vibration storage unit 40 sends a fluctuation section signal capable of calculating the time length of the section in which the occurrence of chatter vibration is determined to the drive control unit 6b, the temperature prediction unit 7b, and the fluctuation condition setting unit 8b. Although transmitted, the vibration storage unit 40 calculates the time length of the section in which the occurrence of chatter vibration is determined, and the time length of the section with vibration is sent to the drive control unit 6b, the temperature prediction unit 7b, and the fluctuation condition setting unit 8b. May be sent.

The fluctuation condition setting unit 8b receives the continuation determination signal from the temperature prediction unit 7b, and receives the fluctuation section signal from the vibration storage unit 40. Further, the fluctuation condition setting unit 8b receives the inference data signal from the drive control unit 6b in the same manner as the fluctuation condition setting unit 8 of the first embodiment. Here, the machining condition information, the motor specification information, and the motor state information, which are a part of the inputs given to the variable condition model 9, are given from the inference data. The fluctuation time information, which is the remaining input, is given the fluctuation section length calculated from the fluctuation section signal received from the vibration storage unit 40. The drive control unit 6b receives the fluctuation section signal from the vibration storage unit 40, and if it is within the fluctuation section, the drive control unit 6b receives the first fluctuation condition or the second fluctuation condition setting unit 8b received from the command generation unit 11b. The motor is fluctuated under fluctuating conditions. Motor fluctuations continue until the end of the fluctuation section.

As described above, the numerical control device 1b of the second embodiment stores the section determined that chatter vibration is generated when the same machining program 12b is executed a plurality of times, and in the subsequent machining. , The motor is configured to fluctuate with respect to the memorized fluctuation section. With this configuration, since the section in which the spindle motor 4 fluctuates coincides with the section in which chatter vibration occurs in actual machining, the length of the section in which the spindle motor 4 fluctuates can be kept to the minimum necessary. In other words, by minimizing the fluctuation interval, the motor can output a larger amplitude and fluctuation frequency. This has the effect of suppressing chatter vibration under variable conditions with a high vibration suppression effect.

Embodiment 3.
In the second embodiment, the section in which the chatter vibration is generated is stored, and then when the same machining program 12b is executed, the motor is changed with respect to the section. In the third embodiment, a configuration will be described in which a section in which chatter vibration is generated and continues is predicted by a machining simulation, and the fluctuation of the motor is continued until the time when the chatter vibration is predicted to end.

FIG. 10 is a block diagram showing the configuration of the numerical control device 1c according to the third embodiment of the present disclosure. In the third embodiment, the vibration storage unit 40 and the sensor 3 are not provided, and the simulation unit 41 is an additional component as compared with the second embodiment. Further, the processing contents of the vibration determination unit 5c, the drive control unit 6c, and the command generation unit 11c are different from the processing of the same name in the second embodiment. In FIG. 10, the same reference numerals are given to the components that are the same as or equivalent to the components of the second embodiment shown in FIG. Here, the part related to the third embodiment will be mainly described.

The command generation unit 11c analyzes the machining program 12b and transmits the machining condition information to the simulation unit 41 described later. Other processing is the same as that of the command generation unit 11b of the second embodiment. Here, the machining program 12b is the same as the machining program 12b of the second embodiment. The simulation unit 41 receives a machining condition signal from the command generation unit 11c. The machining condition signal is a signal representing machining condition information described in the machining program 12b.

Inside the simulation unit 41, the dynamic compliance of each of the tool and the work and the specific cutting resistance between the tool and the work are set. The simulation unit 41 executes the machining simulation 41 according to the machining condition information using these information, and reproduces the operation operation, the machining process, and the vibration in the machine tool 2 when the machining program 12b is executed.

The simulation unit 41 transmits the vibration of the machine tool 2 calculated by executing the simulation to the vibration determination unit 5c as a simulation signal. The vibration determination unit 5c determines the presence or absence of chatter vibration with respect to the simulation signal received from the simulation unit 41 as in the first and second embodiments, and transmits the determination result as a vibration detection signal to the simulation unit 41. To do. Based on the vibration detection signal, the simulation unit 41 transmits information representing a section where chatter vibration is determined to occur as a fluctuation section signal to the temperature prediction unit 7b, the fluctuation condition setting unit 8b, and the drive control unit 6c.

As described above, the numerical control device 1c of the third embodiment predicts a section in which chatter vibration occurs by executing a machining simulation based on the machining condition information described in the machining program, and the section thereof. It is a configuration that fluctuates the motor in. As a result, even in a form in which the sensor 3 or the vibration storage unit 40 is not used, it is possible to suppress chatter vibration under variable conditions that do not overheat the motor.

The numerical control device 1c of the third embodiment is configured to execute the machining simulation from the machining condition information, but it is acquired from the spindle motor 4 instead of the spindle rotation speed and the machining path described in the machining program 12b. The machining simulation may be executed based on the feedback information that can be obtained. In this case, since the simulation can be executed while reflecting the actual movements of the tool and the workpiece, the chatter vibration can be determined with high accuracy.

The configuration shown in the above embodiments shows an example of the contents of the present disclosure, and can be combined with another known technique, the combination of the above embodiments, and the gist of the present disclosure. It is also possible to omit or change a part of the configuration as long as it does not deviate from.

1 Numerical control device, 2 Machine tools, 3 Sensors, 4 Spindle motors, 5 Vibration judgment unit, 6
Drive control unit, 7 temperature prediction unit, 8 fluctuation condition setting unit, 9 fluctuation condition model, 10 learning unit, 11 command generation unit, 12 machining program.

Claims (10)

A command generator that analyzes the machining program and generates an operation command signal, which is a control condition for controlling the motor of the machine tool.
A vibration determination unit that determines the presence or absence of vibration of the machine tool,
When the vibration determination unit determines that vibration is present, a drive control unit that controls the motor based on the first fluctuation condition during the fluctuation section, and a drive control unit.
The actual fluctuation time, which is the time when the motor is controlled based on the first fluctuation condition, and the maximum value of the temperature of the motor during the actual fluctuation time are input, and the maximum value of the temperature of the motor is the maximum value of the motor. A learning unit that generates a fluctuation condition model so as to estimate a second fluctuation condition that can suppress the vibration of the machine tool within an allowable temperature.
Numerical control device equipped with. A command generator that analyzes the machining program and sends an operation command signal, which is a control condition for controlling the motor of the machine tool.
A vibration determination unit that determines the presence or absence of vibration of the machine tool,
When the vibration determination unit determines that vibration is present, a drive control unit that controls the motor based on the first fluctuation condition during the fluctuation section, and a drive control unit.
The actual fluctuation time, which is the time when the motor is controlled based on the first fluctuation condition, and the maximum value of the temperature of the motor during the actual fluctuation time are input, and the maximum value of the temperature of the motor is the maximum value of the motor. The length of the fluctuation section and the allowable temperature of the motor are input to the fluctuation condition model learned to estimate the second fluctuation condition capable of suppressing the vibration of the machine tool within the allowable temperature, and the fluctuation condition model is used. A fluctuation condition setting unit that outputs the estimated second fluctuation condition to the drive control unit, and a fluctuation condition setting unit.
Numerical control device equipped with. The numerical control device according to claim 1 or 2, wherein the command generation unit transmits the first fluctuation condition to the drive control unit. The numerical control device according to any one of claims 1 to 3, wherein the length of the fluctuation section is information described in the machining program. A vibration storage unit that stores the vibration determination result by the vibration determination unit when the machining program is executed, and a vibration storage unit.
Based on the vibration determination result by the vibration determination unit, the temperature determination unit that determines whether or not the control under the first fluctuation condition can be continued in the fluctuation section, and the temperature determination unit.
The numerical control device according to any one of claims 1 to 3, further comprising. Machining conditions for controlling the motor are described in the machining program.
It also has a simulation unit that calculates the estimated vibration value of the machine tool from the machining conditions.
The vibration determination unit determines whether or not vibration has occurred in the machine tool using the vibration estimation value.
Based on the result of the determination by the vibration determination unit, the temperature determination unit that determines whether or not the control under the first fluctuation condition can be continued in the fluctuation section, and the temperature determination unit.
The numerical control device according to any one of claims 1 to 3, further comprising. The numerical value according to any one of claims 1 to 6, wherein the first fluctuation condition or the second fluctuation condition is a combination of either one of a fluctuation frequency and a fluctuation period and a fluctuation amplitude. Control device. The temperature determination unit is characterized in that by comparing the temperature of the motor with the permissible temperature of the motor, it is determined whether or not the control under the first fluctuation condition can be continued in the fluctuation section. The numerical control device according to any one of claims 1 to 7. The numerical control device according to any one of claims 1 to 8, wherein the temperature of the motor is a value acquired by a sensor or a value predicted by the temperature determination unit. The processing conditions for controlling the motor, the motor specification information which is the specification information of the motor, the motor state information indicating the state of the motor, and the control of the motor under the first fluctuation condition in the fluctuation section can be continued. Acquire training data including variable time information that is the time length,
A learning unit that uses the learning data to generate a fluctuation condition model for inferring a second fluctuation condition from the processing conditions, the motor specification information, the motor state information, and the length of the fluctuation section. A learning device equipped with.

Images