Machine tool control device

Abstract

This machine tool control device provides a technology that is capable of avoiding unstable learning control and realizing stable processing. A machine tool control device 10 comprises: a program storage unit 30 for storing a processing program; a position command generation unit 11 for generating a position command from the processing program; a motor speed calculation unit 14 for acquiring a motor speed calculated from the processing program or a motor speed exhibited during idle operation of a motor 1; a learning controller 13 for executing learning control which involves associating a motor angle with a position deviation representing a difference between the position command and the actual position of a driven object that is driven by the motor 1, storing the same, and outputting a correction amount synchronized with the motor angle so that the position deviation is brought closer to zero on the basis of a prescribed learning control parameter; and a stability estimation unit 15 for evaluating stability on the basis of the motor speed that changes over time in a section where the learning control is executed and the learning control parameter.

Description

Machine tool control device

[0001] The present disclosure relates to a control device for a machine tool.

[0002] Conventionally, in a machine tool control device, a technique is known in which a correction amount is output so that the difference between a position command and the actual position of a driven body driven by a servo motor approaches zero (for example, Patent Documents 1 and 2).

[0003] Patent Document 1 describes a servo control device that has the function of automatically adjusting a learning controller, which enables the selection of learning parameters that are suited to machine characteristics and operating conditions by visualizing the convergence of learning and disturbance response.

[0004] Patent Document 2 describes a servo motor control device that can achieve high precision by applying angle-synchronized learning control, even when drilling holes with variable circle diameters or machining free-form closed curve shapes.

[0005] JP 2017-084104 A JP 2016-031735 A

[0006] For operations in which the same operation is repeated for a workpiece at predetermined intervals, learning control is effective, which associates the position deviation for one cycle with the motor angle and stores it in memory, and outputs correction data for each cycle so that the position deviation approaches 0. However, while learning control based on the motor angle is advantageous for following commands synchronized with the motor rotation and suppressing disturbances synchronized with the motor rotation, it has the disadvantage of being prone to becoming unstable by amplifying disturbances not synchronized with the motor rotation and high-frequency noise.

[0007] To address this issue, it is important to use a low-pass filter or adjust the division number (a parameter for the angular resolution of the memory) to moderately limit the response band of the learning control, which is essential for obtaining good machining results. These parameter adjustments are usually performed appropriately by those skilled in the art. However, depending on the machining, such as constant peripheral speed control, the motor rotation speed may change during operation. In such operations, the response band of the learning control changes in synchronization with the motor rotation speed, which can make the learning control unstable.

[0008] The present disclosure has been made in consideration of the above-mentioned problems, and aims to provide a technique for avoiding unstable learning control in a machine tool control device and achieving stable machining.

[0009] The present disclosure relates to a control device for a machine tool that controls at least one motor based on a machining program, comprising: a program memory unit that stores the machining program; a position command generation unit that generates a position command from the machining program; a motor speed calculation unit that acquires the motor speed calculated from the machining program or the motor speed when the motor is running idle; a learning controller that stores a motor angle and a position deviation that is the difference between the actual position of a driven body driven by the motor and the position command in association with each other, and performs learning control that outputs a correction amount synchronized with the motor angle so as to bring the position deviation closer to zero based on predetermined learning control parameters; and a stability estimation unit that evaluates stability based on the motor speed that changes over time in the section where the learning control is performed and the learning control parameters.

[0010] According to the present disclosure, it is possible to provide a technique for avoiding unstable learning control in a machine tool control device and realizing stable machining.

[0011] Fig. 1 is a functional block diagram of a control device for a machine tool according to a first embodiment; Fig. 2 is a graph showing an example of the relationship between the allowable range of the control band and the response band of learning; Fig. 3 is a table showing an example of time series data of an estimated motor speed; Fig. 4 is a graph showing an example of a frequency distribution of the estimated motor speed; Fig. 5 is a flowchart showing the flow of a stability estimation process; Fig. 6 is a functional block diagram of a control device for a machine tool according to a second embodiment;

[0012] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In the description of the second and subsequent embodiments, the same reference numerals will be used to designate components common to the first and second embodiments, and the description thereof will be omitted as appropriate.

[0013] 1 is a functional block diagram of a control device 10 for a machine tool according to a first embodiment. The machine tool is, for example, a machining device or a robot that operates based on content specified in a machining program or content input by manual operation.

[0014] The control device 10 controls the operation of the machine tool by controlling at least one motor 1 that drives a driven object based on a machining program. The motor 1 that is the control target of the control device 10 is, for example, an electric motor such as a servo motor. The number of motors 1 that are the control target of the control device 10 may be one or more. For example, the machine tool that is the control target may also be a machine tool that further includes another motor (spindle motor) that operates in synchronization with the motor (servo motor).

[0015] The control device 10 of this embodiment includes a program memory unit 30, a position command generation unit 11, an encoder 20, a subtractor 21, an adder 22, a position and speed control unit 12, a learning controller 13, a motor speed calculation unit 14, and a stability estimation unit 15.

[0016] The program storage unit 30 is composed of a ROM (Read Only Memory), a RAM (Random Access Memory), a non-volatile memory, a hard disk drive, etc. The program storage unit 30 stores various information for executing machining operations of the machine tool. The various information includes a machining program that determines the content of the machining operation. In this embodiment, the program storage unit 30 is disposed in a numerical control device (CNC: Computerized Numerical Control), but the location where it is disposed is not particularly limited, and it may be disposed outside the numerical control device.

[0017] In this embodiment, the machining program stored in the program storage unit 30 is read out by the position command generation unit 11 and also by the motor speed calculation unit 14 .

[0018] The position command generator 11 generates a position command for operating the motor 1 based on the contents set in the machining program. The position command is generated, for example, to change the pulse frequency in order to change the speed of the motor 1. The position command generator 11 is realized, for example, by a computer numerical control device (CNC). The position command generator 11 outputs the generated position command to the subtractor 21.

[0019] The encoder 20 is a position detector that is disposed on the feed shaft of the motor 1 and detects the amount of rotation of the motor. The position information detected by the encoder 20 is output to the subtractor 21 as position feedback (actual position), and is also output to the learning controller 13 as a motor angle indicating the rotation angle of the motor 1.

[0020] The subtractor 21 is a position deviation acquisition unit that calculates a position deviation, which is the difference between the position command (movement command) created by the position command generation unit 11 and the position feedback from the motor 1. The subtractor 21 outputs the calculated position deviation to the adder 22 and also to the learning controller 13.

[0021] The adder 22 is a position deviation correction unit that generates a composite command by adding the position deviation output from the subtractor 21 and the correction amount calculated by the learning controller 13. The adder 22 outputs the generated composite command to the position and speed control unit 12.

[0022] The position and speed control unit 12 performs position control, speed control, and current control based on the command output from the adder 22 to drive and control the motor 1. This drive control drives the driven object via the motor 1. The motor 1 also outputs the motor angle to the learning controller 13.

[0023] The learning controller 13 stores in memory the position deviation for a predetermined period (e.g., one revolution) in association with the actual position (motor angle) based on predetermined learning control parameters. The learning control parameter is, for example, a division number. The division number here refers to the angular resolution of the memory for storing the motor angle and position deviation in association with each other. The actual position may be any motor angle or position information synchronized with the driven body, and may be, for example, information indicating the motor angle of a servo motor or the motor position of a spindle motor.

[0024] The learning controller 13 executes learning control to set a correction amount for each period based on the relationship between the position error and the actual position so as to bring the position error closer to zero. The learning controller 13 calculates a correction amount synchronized with the motor angle (actual position) input from the encoder 20. As described above, the correction amount calculated by the learning controller 13 is added to the position error by the adder 22. The learning controller 13 also outputs learning control parameters to the stability estimation unit 15, which will be described later.

[0025] The motor speed calculation unit 14 calculates and acquires the motor speed from the machining program. The motor speed calculation unit 14 calculates an estimated motor speed that includes at least the section during learning control by the learning controller 13, and outputs the estimated motor speed to the stability estimation unit 15.

[0026] The stability estimation unit 15 evaluates the stability of the learning control before the start of machining based on the estimated motor speed input from the motor speed calculation unit 14 and the learning control parameters input from the learning controller 13. The estimated motor speed in this embodiment is the motor speed that changes over time in the section where the learning control is executed.

[0027] Here, the evaluation criteria for stability will be explained. Fig. 2 is a graph showing an example of the relationship between the allowable range of the control band and the learning response band. Fig. 2 shows the relationship between the response band (actual band) of learning control when a machining program to be subjected to learning control is executed and the allowable band in which learning control is stably executed, in a graph with the horizontal axis representing machining time (s) and the vertical axis representing band (Hz).

[0028] The response band of learning control is correlated with the division number x reference axis motor speed. As shown in Figure 2, when the response band of learning control is within the allowable range of the control band, the position error is stabilized by the correction of learning control. On the other hand, when the response band of learning control exceeds the threshold of the band where learning becomes unstable, the position error increases.

[0029] The control band (actual band) in which learning control can efficiently suppress position deviation is calculated as shown in Equation 1. The number 60000 in Equation 1 is a value for converting RPM, which indicates the number of rotations per minute, into rotation speed per ms. The learning band in Equation 1 is the cutoff frequency of the LPF (Low Pass Filter), the division number is the angular resolution of the memory, and N is the rotation speed (motor speed) of the reference axis [min －１ The learning band and the number of divisions are both parameters that are set in advance.

[0030]

[0031] Equation 1 shows that when the number of divisions is constant, fluctuations in the motor speed of the reference axis motor cause fluctuations in the actual band. Fluctuations in the actual band cause the amount of correction for the actual position output by the learning controller 13 to become inappropriate. For example, when the motor speed increases, the actual band also increases, resulting in excessive correction and unstable control. Conversely, when the motor speed decreases, the actual band also decreases, resulting in insufficient correction and reduced accuracy.

[0032] Solving Equation 1 for the division number results in Equation 2. In Equation 2, if it is assumed that the learning band and the control band are the same, which is an ideal condition, Equation 2 can be simplified to Equation 3. If the motor speed is known from Equation 3, it is possible to calculate a learning control parameter (division number) that satisfies the condition.

[0033]

[0034]

[0035] Next, a description will be given of a specific method of stability evaluation by the stability estimation unit 15. The stability estimation unit 15 of this embodiment can select an evaluation method from either a method of evaluating stability based on a maximum speed value included in time-series data for a section where learning control is executed, or a method of evaluating stability based on a frequency distribution of the motor speed for a section where learning control is executed.

[0036] A method for evaluating stability based on the maximum speed value included in the time-series data for the section where learning control is executed will be described with reference to Fig. 3. Fig. 3 is a table showing an example of time-series data for estimated motor speed. Fig. 3 also shows the change over time in the estimated motor speed for the section where learning control is performed by the learning controller 13. In this example, the motor speed of 3200 [rpm] on the reference axis at an elapsed time of 30000 [ms] is the maximum speed value in the time-series data for estimated motor speed.

[0037] The stability estimation unit 15 evaluates whether the optimum value of the number of divisions at the maximum speed deviates from a predetermined allowable range (threshold) based on the number of divisions of the learning control parameter. If the optimum value of the number of divisions at the maximum speed deviates from the predetermined range, the stability estimation unit 15 evaluates that the stability of the learning control has deteriorated.

[0038] Next, a method for evaluating stability based on the frequency distribution of the motor speed in the section where learning control is executed will be described with reference to Fig. 4. Fig. 4 is a graph showing an example of the frequency distribution of the estimated motor speed. Fig. 4 also shows a histogram of the frequency distribution in the section where learning control is executed by the learning controller 13. In this example, the most frequent speed value of the estimated motor speed is 1000 to 1500 rpm.

[0039] The stability estimation unit 15 evaluates whether the appropriate value of the division number of an estimated motor speed that appears frequently in the frequency distribution of the motor speed deviates from a predetermined range (threshold value) based on the division number of a preset learning control parameter. If the appropriate value of the division number for the most frequent speed value deviates from the predetermined range, the stability estimation unit 15 evaluates that the stability of the learning control has deteriorated.

[0040] In the stability evaluation process of this embodiment, when the stability estimation unit 15 determines that the stability of the learning control has deteriorated, it outputs an instruction to the learning controller 13 to change the division number so that the appropriate value of the division number of the estimated motor speed falls within a threshold value. Next, the flow of the stability evaluation process based on the motor speed will be described with reference to Figure 5. Figure 5 is a flowchart showing the flow of the stability evaluation process.

[0041] In step S1, a threshold value for the control band is set based on a preset division number (learning control parameter). The division number is set by the operator of the machine tool. As shown in FIG. 2, the threshold value is set based on the range of the band in which learning becomes unstable. The threshold value is set, for example, based on a value obtained by adding a predetermined margin to the vibration frequency expected to be generated by machining. For example, in gear generating processing, when a rotating tool has multiple blades, harmonic vibrations are generated by multiplying the frequency of vibrations caused by the rotation of the tool shaft motor by the number of blades. Therefore, the threshold value can be a value obtained by adding a predetermined margin to this harmonic frequency. As another example, in jig grinding, the tool oscillates along the linear axis, so the threshold value can be a value obtained by adding a predetermined value to the oscillation frequency of the linear axis instead of the motor speed. As yet another example, if the frequency of a disturbance is known, it can be directly used as the threshold value. For example, the frequency of mechanical resonance or chatter vibration of the machine tool may be used, or a disturbance frequency asynchronous with motor rotation obtained in advance by frequency characteristic analysis of the machining results may be used as the threshold value.

[0042] In step S2, the motor speed calculation unit 14 acquires an estimated motor speed, which is an estimate of the motor speed of the reference axis, before machining. The motor speed calculation unit 14 continuously acquires the estimated motor speed for the section where learning control is performed as the machining time changes. In the first embodiment, the estimated motor speed and machining time are acquired by look-ahead from the position commands of the machining program. Furthermore, if the machine tool has multiple motor axes, a person skilled in the art will set the learning axis and reference axis in advance using predetermined parameters, and the reference axis is identified from those parameters, and the motor speed of that reference axis is acquired. The motor speed calculation unit 14 in this embodiment acquires either the estimated motor speed of the maximum speed value in the time-series data (see FIG. 3) or the estimated motor speed with a high frequency of appearance in the frequency distribution of motor speeds (see FIG. 4).

[0043] In step S3, the stability estimation unit 15 calculates an appropriate value for the number of divisions based on the estimated motor speed acquired by the motor speed calculation unit 14. The appropriate value for the number of divisions can be calculated, for example, using Equation 3. As described above, the estimated motor speed used to calculate the appropriate value for the number of divisions is the estimated motor speed of the maximum speed value in the time-series data (see FIG. 3) or the estimated motor speed that appears most frequently in the frequency distribution of motor speeds (see FIG. 4).

[0044] In step S4, stability evaluation is performed by the stability estimation unit 15. In this stability evaluation, the stability estimation unit 15 determines whether the appropriate value of the division number calculated in step S3 falls within a threshold value (tolerance range) based on the division number (learning control parameter) set in advance in the learning controller 13.

[0045] If the optimum value of the number of divisions calculated in step S3 falls within a threshold value based on a preset number of divisions, the stability condition is met and the process ends (step S4; Yes). If the optimum value of the number of divisions calculated in step S3 does not fall within a threshold value based on a preset number of divisions, the stability condition is not met and the process proceeds to step S5 (step S4; No).

[0046] In step S5, the stability estimation unit 15 outputs information instructing a change in the number of divisions as information based on the evaluation result to the learning controller 13. The information instructing a change in the number of divisions includes, for example, numerical information for changing the number of divisions so that the response band of learning does not become too large.

[0047] In step S6, the learning controller 13 changes the division number. This change in the division number brings the appropriate value of the division number based on the estimated motor speed into the threshold value, and the condition for stable execution of learning control is met. After the processing of step S6, the process of evaluating stability ends. Note that if an appropriate division number cannot be set, a configuration may be adopted in which the operator is notified that the division number could not be set. The operator can be notified, for example, by a screen display unit such as a display.

[0048] [Second embodiment] Fig. 6 is a functional block diagram of a control device 10a for a machine tool according to a second embodiment. As shown in Fig. 6, in the second embodiment, the machining program stored in the program storage unit 30 is not read out by the motor speed calculation unit 14, but is read out only by the position command generation unit 11.

[0049] Further, the encoder 20 outputs information for calculating the estimated motor speed to the motor speed calculation unit 14 in addition to the subtractor 21 and the learning controller 13. The motor speed calculation unit 14 acquires the actual motor speed when the motor 1 is idled as the estimated motor speed.

[0050] In this way, the motor speed calculation unit 14 of the second embodiment acquires the actual motor speed obtained by idling the motor 1 before machining as the estimated motor speed after machining starts, instead of the position command of the machining program. Note that the other configurations are the same as those of the first embodiment. The configuration of this second embodiment can also achieve the same effects as the above embodiments.

[0051] In the above embodiment, the stability of the learning control is evaluated by calculating an appropriate value for the number of divisions from the estimated motor speed and comparing this number of divisions with a number of divisions that is a learning control parameter that is preset in the learning controller 13, but the present invention is not limited to this configuration. For example, it is also possible to calculate the motor speed from a preset number of divisions using Equation 3 or the like, and evaluate that stability has deteriorated if the estimated motor speed deviates from an allowable range based on this motor speed.

[0052] In the above embodiment, the learning control parameter is the division number, but this is not limiting. For example, since the actual band is made up of the learning band and the division number as shown in Equation 1, the learning control parameter can also be the learning band.

[0053] As described above, the control device 10 (10a) of the machine tool of this embodiment comprises a program memory unit 30 that stores a machining program, a position command generation unit 11 that generates a position command from the machining program, a motor speed calculation unit 14 that acquires the motor speed calculated from the machining program or the motor speed when the motor 1 is running idle, a learning controller 13 that stores the motor angle and the position deviation that is the difference between the actual position of the driven body driven by the motor 1 and the position command in association with each other, and executes learning control that outputs a correction amount synchronized with the motor angle so as to bring the position deviation closer to zero based on predetermined learning control parameters, and a stability estimation unit 15 that evaluates stability based on the motor speed that changes over time in the section where learning control is executed and the learning control parameters.

[0054] This makes it possible to determine before machining whether the division number set in the learning controller 13 is appropriate based on an estimated motor speed that is a pre-estimate of the rotation speed of the motor 1 after machining has started. This makes it possible to avoid situations where the correction amount is excessive or insufficient, such as an excessive response band (actual band) during machining, due to an inappropriate division number set in the learning controller 13. This makes it possible to avoid situations where learning control is executed in an unstable state, and to achieve stable machining.

[0055] In this embodiment, the learning control parameter is set as a division number that indicates the angular resolution of a memory for storing the motor angle and position deviation in association with each other, and the stability estimation unit 15 calculates an appropriate value for the division number from the motor speed and evaluates whether the appropriate value for the division number deviates from a predetermined range based on a preset division number. This makes it possible to use the angular resolution (division number) to evaluate the stability of the learning control in an appropriate and simple manner.

[0056] In this embodiment, the motor speed calculation unit 14 calculates the maximum motor speed in the section where learning control is performed, and the stability estimation unit 15 evaluates stability based on whether the appropriate value for the number of divisions at the maximum motor speed in the time-series data of the motor speed falls within a predetermined range. As a result, the number of divisions is set based on the highest motor speed, ensuring stability throughout the entire section where learning control is performed.

[0057] In this embodiment, the stability estimation unit 15 evaluates stability based on whether the appropriate value of the division number of the motor speed that appears most frequently in the frequency distribution of the motor speed in the section where learning control is executed falls within a predetermined range. This effectively improves the stability of learning control in the range of motor speeds that appear most frequently. Compared to when the division number of the motor speed of the maximum speed is used, the accuracy of learning control can be ensured even in low-speed ranges.

[0058] Furthermore, in this embodiment, when the stability estimation unit 15 determines that the stability of the learning control has deteriorated, it instructs the learning controller 13 to change the division number. As a result, when it is estimated that the stability is deteriorating, the division number is automatically changed so that the actual band falls within the allowable range, thereby reliably preventing the position error from being corrected with an excessive or insufficient correction amount.

[0059] The present disclosure is not limited to the above-described embodiments, and includes modifications and improvements within the scope of achieving the object of the present disclosure.

[0060] The following supplementary note is further disclosed regarding the above embodiment and modified examples. (Supplementary Note 1) A control device (10, 10a) for a machine tool that controls at least one motor (1) based on a machining program, comprising: a program storage unit (30) that stores the machining program, a position command generation unit (11) that generates a position command from the machining program, a motor speed calculation unit (14) that acquires a motor speed calculated from the machining program or the motor speed when the motor (1) is running idle, a learning controller (13) that stores a motor angle and a position error that is the difference between an actual position of a driven body driven by the motor (1) and the position command in association with each other, and performs learning control that outputs a correction amount synchronized with the motor angle so as to bring the position error closer to zero based on predetermined learning control parameters, and a stability estimation unit (15) that evaluates stability based on the motor speed that changes over time in an interval where the learning control is performed and the learning control parameters.

[0061] (Supplementary Note 2) In the above-mentioned machine tool control device (10, 10a), the learning control parameter is set as a division number indicating the angular resolution of a memory for storing the motor angle and position deviation in association with each other, and the stability estimation unit (15) calculates an appropriate value for the division number from the motor speed and evaluates whether or not the appropriate value for the division number deviates from a predetermined range based on a preset division number.

[0062] (Supplementary Note 3) In the above-described machine tool control device (10, 10a), the motor speed calculation unit (14) calculates a maximum speed value of the motor in the section where the learning control is executed, and the stability estimation unit (15) evaluates stability based on whether or not the appropriate value of the number of divisions at the maximum motor speed value of the time-series data of the motor speed falls within a predetermined range.

[0063] (Supplementary Note 4) In the above-mentioned machine tool control device (10, 10a), the stability estimation unit (15) evaluates stability based on whether the appropriate value of the division number of the motor speed that appears frequently in the frequency distribution of the motor speed in the section where the learning control is executed falls within a predetermined range.

[0064] (Supplementary Note 5) In the above-described machine tool control device (10, 10a), the stability estimation unit (15) instructs the learning controller (13) to change the learning control parameters based on the evaluated stability.

[0065] REFERENCE SIGNS LIST 10, 10a Machine tool control device 11 Position command generation unit 12 Position and speed control unit 13 Learning controller 14 Motor speed calculation unit 15 Stability estimation unit 20 Encoder (position detection unit) 21 Subtractor (position deviation acquisition unit)

Claims

1. A control device for a machine tool that controls at least one motor based on a machining program, the control device including: a program storage unit that stores the machining program; a position command generation unit that generates a position command from the machining program; a motor speed calculation unit that obtains a motor speed calculated from the machining program or a motor speed when the motor is idling; a learning controller that stores the motor angle in association with a position deviation, which is a difference between the actual position of a driven body driven by the motor and the position command, and outputs a correction amount synchronized with the motor angle so as to bring the position deviation closer to zero based on a predetermined learning control parameter; and a stability estimation unit that evaluates stability based on the motor speed that changes over time in a section where the learning control is executed and the learning control parameter.

2. The learning control parameter is set as the number of divisions indicating the angular resolution of a memory for storing the motor angle in association with the position deviation, and the stability estimation unit calculates an appropriate value of the number of divisions from the motor speed and evaluates whether or not the appropriate value of the number of divisions deviates from a predetermined range based on a preset number of divisions. The control device for a machine tool according to claim 1.

3. The motor speed calculation unit calculates a maximum value of the speed of the motor in a section where the learning control is executed, and the stability estimation unit evaluates stability based on whether or not an appropriate value of the number of divisions at the maximum value of the motor speed in the time series data of the motor speed falls within a predetermined range. The control device for a machine tool according to claim 2.

4. The stability estimation unit evaluates stability based on whether or not an appropriate value of the number of divisions of the motor speed having a high appearance frequency in the frequency distribution of the motor speed in a section where the learning control is executed falls within a predetermined range. The control device for a machine tool according to claim 2.

5. The stability estimation unit instructs the learning controller to change the learning control parameter based on the evaluated stability. The control device for a machine tool according to claim 1 or 2.

Images