JPH086644A - Numerical controller - Google Patents

Abstract

PURPOSE:To stabilize the behavior of a controlled system and improve the machining precision of a work by adding a correction torque value calculated in a machining cycle as a feedforward control value to a torque command value in a next machining cycle. CONSTITUTION:A correction torque value arithmetic part 3 when inputting a position command value SP, a full closed position detected value SPDF, a correction torque specification value ST', and a speed detected value SVD calculates the correction torque value STFCI in each machining cycle and transfers it to a main controller 4. The main controller 4 multiplies the correction torque value STFCO after filter processing and phase matching by a feedforward gain Kfc and an adder 16 adds the result to the torque command value ST. Therefore, variation in torque can be estimated in each machining cycle at the time of repetitive control and the position error torque estimated value is added as the feedforward control value to the torque command value together with a torque variation estimated value, so the behavior of the controlled system is stabilized and the machining precision of the work is stabilized.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a numerical control device for machine tools, and more particularly, it has a function of improving machining accuracy by correcting a torque command value during repetitive control and stabilizing the behavior of a controlled object. The present invention relates to a feed axis numerical control device.

[0002]

2. Description of the Related Art FIG. 8 is a block diagram showing the outline of a numerical controller for controlling a feed axis of a machine tool. Position command value SP and position detection value SPD output from main controller 4
Then, the subtracter 14 calculates the position error value. The position control unit 1 multiplies the position error value by a position loop gain and outputs it as a speed command value SV to the subtractor 15. Further, the speed detection value SVD is input to the subtractor 15 to calculate the speed error, and the speed control unit 2 calculates the torque command value ST according to the speed error. Then, the power amplifier 8 drives the motor 9 with a current according to the torque command value ST. A position detector 12 is attached to the motor 9, and a separate position detector 13 is attached to the table 10. Each position detection value SP
A position detection unit 6 that calculates a position detection value SPD from DS and SPDF and a speed detection unit 7 that calculates a speed detection value SVD from the position detection value SPDS are provided. The fully closed position detection value SPD is used as a means for improving the processing accuracy.
The position error calculation unit 8 calculates the position error value between F and the position command value SP.
1. The position command value SP is added from the position command value SP by adding the corrected position command value SPC, which is a numerical value obtained by numerically processing (filtering or phase matching) the position error value. SP + position command value correction amount SPC = corrected position command SP ′ is commanded. In addition,
This correction step is not performed every machining cycle during machining, but is performed off-line with the control system for several machining cycles before machining.

[0003]

However, in the case of repeatedly controlling the periodic target value by the above-mentioned conventional numerical control device, the ripples and torque of the speed detecting portion (mainly fluctuations in the velocity detecting portion after correcting the position command value). (Sliding torque, inertia fluctuation, disturbance due to instability of motor control) appear as trace errors on the surface of the work piece, which causes a decrease in surface roughness such as ripple marks and a decrease in processing accuracy. is there. In addition, since the correction is performed on the position control loop which is a major loop (long control cycle), it is less effective against the disturbance generated in the speed control loop and the current control loop which are minor loops (short control cycle). However, there is a drawback that a disturbance component generated in a major loop is superimposed on the position correction value itself, and that the position command value is corrected off-line, so that unnecessary man-hours are required for the correction. The present invention has been made in view of the above-described circumstances, and an object of the present invention is to stabilize the behavior of a controlled object and improve the machining accuracy of a workpiece by correcting the torque command value when performing repetitive control. It is to provide a numerical control device for a machine tool that can perform.

[0004]

To achieve the above object, the invention according to claim 1 calculates a position error value from a position command value from a main controller and a position detection value from a controlled object. A position control unit for performing position control, the position error value,
A speed control unit that performs speed control by calculating a speed error value from the speed detection value calculated from the position detection value, and a power amplification unit that converts a torque command value output from the speed control unit into a current command value. And a position detector that detects the position of the controlled object, and in a numerical controller that repeatedly controls a periodic target value, the position command value, the position detection value, the speed detection value, and the torque command. And a correction torque value calculating means for calculating a correction torque value for each processing cycle, and the correction torque value calculated in the previous processing cycle is used as a feedforward control value in the next processing cycle. The feature is that it is added to the command value. According to a second aspect of the present invention, in the numerical control device according to the first aspect, the correction torque value computing means has a disturbance torque estimating means for estimating a torque deviation value from the torque command value and the speed detection value. Is characterized by. According to a third aspect of the present invention, in the numerical control device according to the first aspect, the correction torque value calculating means estimates a torque deviation value from a position error value between the position command value and the position detection value. It is characterized by having a torque estimating means. According to a fourth aspect of the present invention, in the numerical control device according to the first aspect, the correction torque value computing means has a function generating section for generating a function based on the position error value. According to a fifth aspect of the present invention, in the numerical control device according to the first aspect, the correction torque value computing means estimates a first torque deviation value from the torque command value and the speed detection value, and the position command. And a second torque deviation value estimated from a position error value between the detected value and the position detection value, and a numerical processing unit for calculating the correction torque value.

[0005]

In the present invention, during the repetitive control, the torque fluctuations (mainly sliding torque, inertia fluctuations,
(Disturbance due to instability of motor control) can be estimated, and a function is generated from a position error due to control delay of the servo system and ripple of the detector to obtain a position error torque estimated value. Since the feedforward control value is added to the torque command value, the behavior of the controlled object can be stabilized. Also, a series of correction steps can be performed online.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A numerical controller according to the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an example of the numerical control device of the present invention in association with FIG. The blocks having the same numbers and the signals having the same symbols used in FIG. 8 have the same functions and contents. Therefore, the description is omitted because it is the same as in FIG. 8 except the means for improving the processing accuracy. The full-closed position detection value SPDF and the position command value SP detected by the position detector 13 attached to the table 10 are input to the correction torque value calculation unit 3. Further, the semi-closed position detection value SPDS detected by the position detector 12 attached to the motor 9 is used to feed the speed detection value SVD via the speed detection section 7 and the torque command value ST calculated by the speed control section 2. The correction torque value STFCO multiplied by the feedforward gain Kfc by the forward gain correction unit 5 is input to the adder 16, and the calculated correction torque command value ST ′ is input. The feedforward gain Kfc is variable between 0 and 100%,
The value of the correction torque value STFCO added to the torque command value ST as the feedforward control value can be adjusted.
In this way, the correction torque value calculation unit 3 receives the position command value SP, the fully closed position detection value SPDF, the correction torque command value ST ′, and the speed detection value SVD, and then outputs the correction torque value STFCI for each machining cycle. The correction torque value calculation means calculates the correction torque value STFCI and transfers the correction torque value STFCI to the main controller 4. Correction torque value STFCO filtered and phase-matched (see FIG. 6) by main controller 4
Is multiplied by the feedforward gain Kfc and added by the torque command value ST and the adder 16. There are two types (STFCO, STFCI) in the notation of the correction torque value because they are output as STFCO after the STFCI is filtered and phase-matched in the main control unit, so that they are distinguished for convenience.

FIG. 2 is a block diagram showing the inside of the correction torque value calculation unit 3 of the present invention. The internal processing in the correction torque value calculation unit 3 is roughly divided into two blocks. One is a disturbance torque estimating unit 2 as a disturbance torque estimating means for estimating a torque deviation value from a torque command value, which is a correction torque value ST ′ and a speed detection value SVD in this embodiment.
2 is the other, and the other is the position error value torque estimating unit 21 as a position error value torque estimating means for estimating the torque deviation value from the position error value between the position command value SP and the fully closed position detection value SPDF. First, the disturbance torque estimation unit 22 adds the torque coefficient K to the input corrected torque command value ST ′.
A value obtained by multiplying the actual torque value by multiplying t and then multiplying the speed detection value SVD calculated by the speed detection unit 7 by the sliding friction coefficient F + inertia J · Laplace operator S (S means time derivative). By subtracting with the subtractor 226, the sliding friction torque fluctuation, the torque fluctuation due to inertia, and the unstable torque for motor control can be estimated. If a signal containing a frequency higher than the servo performance is used as it is, the control system becomes unstable. Therefore, the low-pass filter (LPF) 22
Processing is performed at 3. Further, in the torque constant calculation unit 224, division is performed by the torque constant Kt in order to align the dimension with the torque command value ST '. Further, it is multiplied by a gain (1−Kd) to output a disturbance torque estimated value STOB to the numerical processing unit 23. In addition, the full-closed position detection value SPD input to the position error value torque estimation unit 21.
Since F normally contains a noise component, the noise component is removed by the low pass filter (LPF) 211, and the phase matching by the filter is performed by the phase matching unit 212. The noise-free fully closed position detection value SPDF and the position command value SP are input to the subtractor 215 to obtain the position error value SD. The position error torque estimated value STD having the same dimension as the torque command value ST is output from the function generation unit 213 based on the position error value SD, multiplied by the gain Kd, and output as the position error torque estimated value STDI to the numerical processing unit 23.
Note that the torque values STOB and STDI output from the disturbance torque estimation unit 22 and the position error value torque estimation unit 21 are multiplied by the gains (Kd and 1−Kd), which means that the outputs from the estimation units are weighted. This is to do it. For example,
If the position error torque estimated value STDI is emphasized, K
When d = 80%, the gain of the disturbance torque estimated value STOB is 1−0.8 = 0.2; that is, 20%, and the position error torque estimated value STDI: disturbance torque estimated value STOB is weighted at a ratio of 8: 2. It Next, the numerical processing unit 23
The estimated disturbance torque value STO as the first torque deviation value
B and the position error torque estimated value STDI as the second torque deviation value are input, and the corrected torque value STFCI is output. The position error torque estimated value STDI often includes the disturbance torque estimated value STOB, and the Kd is usually 50.
Use at% or more. The gain Kd is adjusted depending on the situation of the controlled object and the purpose of the servo adjuster.

FIG. 3 is a block diagram showing the configuration of the numerical processing unit 23 shown in FIG. The disturbance torque estimated value STOB and the position error torque estimated value STDI are input to the numerical processing unit 23, and the respective values are input to the adder 35 and the subtractor 36. The output from the adder 35 is the sign determiner 33.
Is output to the calculator 37. The output from the subtractor 36 is converted into an absolute value and output to the switch 34. The sign determiner 33 determines whether the P side of the switch 34, N
Switch side. The condition for switching by this code determination can be set arbitrarily. When the switch 34 is on the P side, the calculator 37 functions as an adder, and when it is on the N side,
The calculator 37 has a function of working as a subtractor. Calculator 37
Output is halved, then the corrected torque value STFC
It is output as I.

FIG. 4 shows an example of the correction torque calculation processing result by the numerical processing unit 23. Control position (time) on the X-axis, Y
The axis shows torque. Description will be given below with reference to the drawings. In FIG. 4A, the disturbance torque estimation value STOB output from the disturbance torque estimation unit 22 and the position error value torque estimation unit 2 are shown.
The output from 1 is the position error torque estimate STDI. FIG. 4B shows the conditions of the sign determiner 33 of FIG. 3 as follows: when the disturbance torque estimated value STOB + position error torque estimated value STDI ≦ 0, the switch 34 is set to the N side, and the disturbance torque estimated value STOB + position error torque estimated value. STDI>
In the case of 0, the correction torque value STFCI when the switch 34 is set to be switched to the P side is shown. As can be seen from the above, in the case of the above condition, the one having the larger absolute value of each estimated torque value at the same position (time) is selected. This condition is often selected when the correction amount is dynamically applied. FIG. 4C shows the condition of the sign determiner 33 of FIG. 3 as the disturbance torque estimated value STOB + position error torque estimated value STDI.
When> 0, the switch 34 is set to the N side, and when the disturbance torque estimated value STOB + position error torque estimated value STDI ≦ 0, the corrected torque value STFCI is set when the switch 34 is set to the P side. Under this condition, the smaller one of the absolute values of the estimated torque values at the same position (time) is selected. FIG. 4D shows a correction torque value STFCI when the switch 34 of FIG. 3 is opened (the code determination unit 33 is set so as not to switch to the N side and the P side). When this condition is selected, a value of (disturbance torque estimated value STOB + positional error torque estimated value STDI) / 2 is output, and therefore, an intermediate value between the disturbance torque estimated value STOB and the position error torque estimated value STDI is set as the correction torque value STFCI. Can be used.

FIG. 5 shows the position error value torque estimating unit 21 of FIG.
3 is a block diagram showing the configuration of a function generating unit 213 in FIG.
The function generating section 213 generates a function based on the position error value. First, the position error value SD is input to the low pass filter 41 and the integrator 45. The position error value SD is processed by the low-pass filter 41 to remove noise and reduce the frequency higher than the servo performance. Further, the position error value SD is time-integrated by the integrator 45 to remove the steady deviation, and added by the adder 46. It Since the control cycle of the speed and current control loop that handles the torque command value ST is shorter than that of the position control loop that handles the normal position error value SD, the position error value SD is interpolated by the interpolator 42 (a spline function is used as an example of the interpolation process. The value that can be handled by the speed and current control loop by performing numerical interpolation (using the same control cycle)
And Next, the Laplace operator S is multiplied by the square (two-fold differentiation of time), the phase matching unit 44 performs phase matching, and the position error torque estimated value STD is output.

FIG. 6 is a block diagram showing the structure of the main control unit 4 which inputs the correction torque value STFCI and outputs the correction torque value STFCO. The main controller 4 determines the position command value S
A memory 55 for storing P is provided, from which the position command value S
Output P. Further, the correction torque value STFCI calculated by the correction torque value calculation unit 3 is input to the buffer memory 54. The buffer memory 54 is used for performing the processing by the low pass filter 53 and the processing by the phase matching unit 52. The correction torque value STFCO after the low-pass filter processing and the phase matching processing is performed is stored in the memory 51.
The correction torque value S corresponding to the position command value SP is stored in
TFCO is output.

FIG. 7 shows the effect of using the numerical controller according to this embodiment in comparison with the case of using the conventional numerical controller. The X-axis shows the control position (time) and the Y-axis shows the position error value between the fully closed detection position of the controlled object and the position command value. FIG. 7A shows the position error of the controlled object when the conventional numerical control device is used, and the maximum is about ± 8 μm.
A position error value of about m is generated. There is also a large pulsation. In this case, pulsation similar to the position error value appears on the surface of the work piece, which often causes a problem. The cause of this is considered to be that the pulsation of the position detector 12 used for velocity feedback has a great influence. The speed detection value SVD used for speed feedback is the position detection value SPD.
Since S is created by differentiating, when the speed detection value SVD has a pulsation, the amplitude of the pulsation increases by performing the differentiation.
It is speculated that particularly high-order pulsation becomes large, but frequency components of servo performance or higher are filtered by the controller, and ripples of several tens of cycles per one rotation of the detector are actually a problem in many cases. From this, it can be said that the effect is small even if the position error value SD is directly corrected with respect to the position command value SP. FIG. 7B shows the position error of the controlled object when the numerical control device of the present invention is used. It can be seen that the position error value can be suppressed to about ± 1 μm. Moreover, pulsation is also reduced.

The present invention is not limited to the present embodiment shown in FIGS. 1 to 6, and the following modifications may be made without departing from the spirit of the present invention. (1) The correction torque value of the present embodiment is not transferred from the main control device, has a memory in the correction torque value calculation unit 3, and the correction torque value synchronized with the position command value SP from the main control device 4 is from the memory. STFCO may be output. (2) In the present embodiment, the description has been given of the discrete control device, but a continuous control device may be used. (3) In the description of the present embodiment, the low-pass filter 41 is used in the processing in the function generator 213, but the processing of this portion may use a method such as weighted averaging or moving averaging in the discrete system control device. (4) In the numerical processing in the correction torque value calculation unit 3 of this embodiment, a method of directly calculating the sum or difference of the position error torque estimated value STDI and the disturbance torque estimated value STOB may be used. (5) In this embodiment, the correction torque value STFCO is updated during each machining cycle during the actual machining, but the correction torque value STFCO is calculated before machining so that the position error value SD becomes smaller for each machining cycle. However, the torque feedforward control value (= correction torque value STFCO) may be treated as a fixed value.

[0014]

As described above, according to the numerical control device of the present invention, torque fluctuations (mainly sliding torque, inertia fluctuations, instability of motor control, etc.) in each machining cycle during repetitive control. (Disturbance caused by, etc.) can be estimated, and a function is generated from the position error due to the control delay of the servo system and the ripple of the speed detection unit to obtain the position error torque estimated value, which is fed to the torque command value together with the torque fluctuation estimated value. Since it is added as the forward control value, the behavior of the controlled object can be stabilized, so that the machining accuracy of the work can be improved. Further, since a series of correction steps can be performed online, the number of man-hours can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of a numerical control device according to the present invention.

FIG. 2 is a block diagram showing an example of a correction torque calculation unit of the numerical control device according to the present invention.

FIG. 3 is a block diagram showing a configuration of a numerical processing unit shown in FIG.

FIG. 4 is a diagram showing an example of a correction torque calculation processing result in the present embodiment.

5 is a block diagram showing a configuration of a function generating section shown in FIG.

FIG. 6 is a block diagram showing a configuration of a main controller in the present embodiment.

FIG. 7 is a diagram showing an example of the effects of the conventional numerical control device and the numerical control device of the present invention.

FIG. 8 is a block diagram showing a configuration of a conventional numerical control device.

[Explanation of symbols]

1 position control unit 2 speed control unit 3 correction torque value calculation unit 4 main control unit 5 feedforward gain correction unit 6 position detection unit 7 speed detection unit 8 power amplification unit 9 motor 10 table 11 ball screw 12, 13 position detector 14, 15, 36, 215, 226 Subtractor 16, 35, 46 Adder 21 Position error value torque estimation unit 22 Disturbance torque estimation unit 23 Numerical processing unit 33 Sign determination unit 34 Switch 37 Operation unit 41, 53, 211, 223 Low pass filter 42 Interpolator 44, 52, 212 Phase matching device 45 Integrator 51, 55 Memory 54 Buffer memory 61, 62, 63, 64, 65 Torque waveform 71, 72 Position error value waveform 81 Position deviation calculator 213 Function generator 224 Torque Constant calculation unit SP Position command value SP 'Position command value after correction SV Speed command Value ST Torque command value ST 'Correction torque command value SPD Position detection value SVD Speed detection value SPDS Semi-closed position detection value SPDF Full closed position detection value STFCI Correction torque value STFCO Correction torque value SD Position error value STD Position error torque estimated value STOB Disturbance torque estimated value STDI Position error torque estimated value (gain Kd x ST
D) SPC position error correction amount

Claims (5)

[Claims] 1. A position control unit for performing position control by calculating a position error value from a position command value from a main control unit and a position detection value from a controlled object, a position error value and a position detection value. A speed control unit that calculates a speed error value from the detected speed value and performs speed control, a power amplification unit that converts the torque command value output from the speed control unit into a current command value, and the position of the controlled object. In a numerical controller having a position detector for detecting, and performing repetitive control of a periodic target value, a torque value is calculated from the position command value, the position detection value, the speed detection value, and the torque command value. A correction torque value calculating means for correcting and calculating a correction torque value for each machining cycle is provided, and the corrected torque value calculated in the previous machining cycle is added to the torque command value as a feedforward control value in the next machining cycle. Features Numerical control device. 2. The numerical control device according to claim 1, wherein the correction torque value calculation means has a disturbance torque estimation means for estimating a torque deviation value from the torque command value and the speed detection value. 3. The correction torque value computing means has a position error value torque estimating means for estimating a torque deviation value from a position error value between the position command value and the position detection value. Numerical control device described. 4. The numerical control device according to claim 1, wherein the correction torque value calculating means has a function generating section for generating a function based on the position error value. 5. The correction torque value calculating means estimates from a first torque deviation value estimated from the torque command value and the speed detection value, and a position error value between the position command value and the position detection value. The numerical control device according to claim 1, further comprising a numerical processing unit that calculates the correction torque value based on the second torque deviation value.