JPS6119540A - Compensating method of feed screw pitch error in numerically controlled machine tool - Google Patents

Abstract

PURPOSE:To compensate a feed screw pitch error very accurately in a numerically controlled machine tool by adding compensatory amount at a point, to be compensated, of a moving body to a command signal to control a servo-motor. CONSTITUTION:In the compensating method of feed screw pitch error, when the total traveling stroke length SL of a moving body 12 is input from an input unit 3 to CPU1 of a NC device 2 CPU1 processes a step SP10 (computation process for a space (CI) to be compensated. Secondly, according to positional data, the magnitude of error in the positional data, and the total traveling stroke of the moving body (a table) 12, the point to be compensated is set and a compensatory amount at the point to be compensated is computed and stored in the memory in compliance with a memory capacity. Consequently, a memory domain prepared for the compensatory amount in the memory can be utilized to its maximum and efficiently. The pitch error of a feed screw 14 can be therefore compensated very accurately.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is a numerically controlled machine tool in which a correction amount for pitch error of a feed screw is stored in advance, and when a movable object is moved, the correction amount is extracted and the pitch error of a feed screw is adjusted. The present invention relates to a feed screw pitch error correction method for correcting pitch errors of.

Conventionally, as a correction method of type q, a correction amount table in the device is stored at each measurement point, and when the machine comes to the correction position, the corresponding correction amount is retrieved from the correction amount table. A method of correcting the pitch error of the feed screw is known (see Japanese Patent Publication No. 59-11125). In the correction amount table used in this method, the correction position from the machine origin and the correction amount at that position (as shown in Table 1) are sequentially input and stored.

Table 1 However, when setting the correction amount in the correction amount table in this way, if the number of correction positions (measurement positions) is small, the correction amount at each correction position will become large and the pitch error correction will be rough. Next, there is the problem that highly accurate correction cannot be performed. Mayu, a numerical control device usually has one axis.
pF) Since a memory area is provided in which 100 to 200 correction points can be set, if you try to maximize the use of this memory area to increase the correction accuracy, you will have to input as many correction positions and As the amount of correction increases, the number of operations for calculating the amount of correction from measurement data and inputting the correction position and amount of correction increases, leading to problems in that the amount of work increases and calculation and input errors are more likely to occur.

The present invention has been made in view of the above circumstances, and its purpose is to easily input pitch error data of a feed screw and to enable highly accurate pitch error correction in a numerically controlled machine tool. An error correction method is provided.

In order to achieve the above object, the present invention approximates the pitch error data of nine feed screws measured in advance by a straight line, and calculates the position of the bending point of the straight line, the amount of deviation at the position, and the total movement of the moving body. input the stroke length into the numerical control device, set correction points for each of the above positions according to the above positions, the error length, and the total movement stroke length, and perform correction 1 at the correction points.
This is a process in which the amount of correction at the correction point is added to the command signal each time the moving object reaches the correction point to control the servo motor. This is what you do.

Hereinafter, one embodiment of the present invention will be described based on the drawings.

The first factor is the CPU (Central Processing Unit) of the numerical control device 2, and the CPU 1 is designed to perform various processes according to programs inputted from the input section 3 on tape. Mayu crty 1 includes a point register (digit register) 4 and a storage device 5. and the counter circuit 6 are connected. This point register 4 is the CPU
The storage device 5 has a correction amount table T that stores correction numbers N and corresponding correction amounts CP. It is set. Further, the counter circuit 6 is a CPU
Count each time a signal is input from 1 and set it to a predetermined value -i4
This is the book that notifies CPU 1 of the completion of counting. Furthermore, the % CPU 1 is connected to an interpolator 7, and this interpolator 7 performs interpolation (linear interpolation, circular interpolation, etc.) based on the data output by the CPU 1 and outputs a command pulse PA to the subtracter 8. Output. This subtracter 8 is configured to calculate the difference between the command pulse PA and the feedback pulse PB from the synchronization circuit 9 and output the difference to the D/A converter (digital-to-analog converter) 10. It is configured. The continuous signal converted from the pulse signal by the D/A converter 10 is amplified by the amplifier 11 and input to the servo motor M. Furthermore, the next encoder PG connected to the servo motor M outputs a position detection pulse PD to the synchronization circuit 9, and at the same time outputs a position detection pulse PD to the synchronization circuit 9.
A rotation signal (pulse signal IPC) is output for each rotation of the encoder PG for one interrupt gradient.

The servo motor M is designed to rotate or stop the table of the machine tool body (the feed screw 14 provided at the bottom of the moving body 112 and meshing with the nine nut 13). At times, the relationship between the encoder PG and the table 12 is set such that the rotation signal PC of the encoder PG is outputted.

Next, the feed screw pitch error correction method according to the present invention will be explained using the apparatus configured as described above.

(1) First, when the total movement stroke length SL of the moving body 12 is input from the input section 3, the CPU 1 performs the process f of step 5pio (calculation process of the correction interval CI) shown in FIG. Here, the storage point PN is determined by the size of the correction amount area (storage area for storing the correction amount) set in the storage device 5. And, for example, memory point PN =
200° full movement stroke length SL = 500 (m
), then correction interval CI = 5007200 = 2.5
(b) becomes.

(2) On the other hand, move the table 12 from the origin in advance and measure at predetermined intervals (50 cm intervals in the first example), and based on the pitch error amount, The pitch error curve shown in FIG. 2 (b) is drawn and approximated by a straight line as shown in FIG. In this case, the pitch error of the feed screw is the second
Since most of the curves shown in Figure 2(A) have thick undulating changes, it is possible to approximate the curves shown in Figure 2(B) with several straight lines.

(3) Then, as shown in FIG. 2(b), the position data of the bending point of the straight-line approximated pitch error curve (Q distance from the origin) and the amount of error at that point are input from the input section 3. For example, in Figure 2 (b), p3 (150,
10), P5 (250,5), P6 (3o o,
s ), P, (350,-1), Plo(500,-1
0) Input the data of 5 points.

(41P3.P5 * P6 + P7 a Plo
When the position data of the five points and the error amount at the point are input, the CPU 1 first calculates the error amount (corrected pitch amount) CPNf between the input position data and the next position data according to step SP 20 shown in Figure 1. do. For example, between the 26th point and the P7 point, the previous error amount PPN = 5
(μ)・・・・・・P point error amount, current error 1NPN
=-1(μ)...The amount of error is 27 points, so
The corrected pitch amount CPN =5-(-1)=6(μ).

(5) Furthermore, the distance between two points (section data) is calculated by PID (see step 5P21). For example, between points P6 and P7, current position data NPD = 350 (u
'I...P7 point position data, previous position data PPD = 300 (arrogance)...P6 point position data, so interval data PID = 350-30
0=50 (pause).

(6) Next, calculate the number of section points IPN indicating the number of input points that can be set between each position data [Step 5P
22). For example, in the case between %P6 point and P7 point, interval data PID = 50 (ya), correction interval CI
= 2.5 (aua), so the number of section points IP
N=50/2.5=20. And position number P
Set N to 0.

(At 71KR'8, the position number PN and the section point number IPN are compared as shown in step 8P23.

Then, the following process (correction data setting process step 5P2) is performed until the position number PN matches the section point number IPN.
4 to 30). For example, % P6 points and P7
In the case between points, since the number of points in one section IPN=20, the loop LPe shown in FIG. 4 is repeated 20 times and the process is completed.

(8) If the position number PN does not match the section point number IPN in step 5P23, first, the position number PN is counted up (+1) as shown in step 5P24. For example, between the %26 point and the P7 point, the position number PN is loop LP ft from 0 to 20.
Updated each time it is repeated.

(9) Then, in step 5P25, the current correction amount NCP is calculated. For example, in the case between point k P6 and P7, the correction pitch amount CPU = 6Cμ) and the number of interval points IPN = 20, so the = A correction amount NCP for the position number PN will be as shown in Table 2. . However, the points below the current correction amount NCP are rounded down and discarded.

aC Furthermore, the correction amount CP shown in step 5P26
Calculate. Here, the memory correction amount MCP is the previous digit! The current correction amount NCP at the time of number PN is stored (see step 5P28). Set % 26 points and P7
In the case between points, the correction amount CP becomes L shown in Table 2.

al) Next, in step 1C, the correction amount CP and the minimum correction unit (the minimum unit defined in the numerical control device 2, in this embodiment, 1) are compared in step 5P27.

Table 2 cz ・i Then, as a result of the comparison, if the correction 1icp is greater than or equal to the minimum correction unit (1), the current correction - 11NCP
is stored in the memory correction amount MCP (Step 5P28
reference). Also, the correction amount CP is the minimum correction unit fllj! If the difference is smaller, the correction amount CP is set to 0 (see step 5P29).

1st floor Next, the correction amount CP is set to the correction number N that matches the value of the point register 4 of the correction amount table T in the storage device 5.
is stored in the correction amount area of point register 4.
Update (+1). In the case between point 26 and point P7, the value of point register 4 is 121 every time loop LP is repeated.
Corresponding to ~140, correction number N = 1 as shown in Figure @5
Correction amount areas 21 to 140 (correction amount CP for each
is memorized. In addition, in FIG. 5, the correction amount CP at the correction number N=120 is obtained by the correction amount calculation process between the P5 point and the P6 point, and is in the second place, and in this example, *P=0. .

After the above steps (1) to a are completed, the correction number N=1 to 250 is stored in the correction amount area of the correction amount table T in the storage device 5.
After storing the correction amount CP corresponding to , the correction amount table Te is used to correct the pitch error accompanying the movement of the moving body 12. The procedure will be explained below.

First, after returning the table (moving body) 12 to its origin and clearing the point register 4 (to 0), the CPU
1 outputs movement amount data to the interpolator 7 in accordance with the movement command inputted from the input section 3 on the tape.
correction interval (2.5 rnx in this example)
Rotation signal (pulse signal) of encoder PC output to
Six counter circuits are set to output a counting completion signal to the CPU 1 after counting the number. The interpolator 7 that has received the movement amount data outputs a command pulse PA corresponding to the movement amount data to the subtracter 8. The rotation command continues to be output to the servo motor M via the D/A converter 10 and amplifier 11 until the feedback pulse FB inputted via the synchronization circuit 9 matches. As a result, the servo motor M starts rotating, so the table 12 starts moving via the feed screw 14 and nut l3.

When the servo motor M rotates, the encoder PG connected to the servo motor M rotates accordingly, inputting the position detection pulse PD to the synchronization circuit 9, which converts it into a feedback pulse PB and outputs it as a trigger. Since it is output to the calculator 8, feedback control is performed using the command pulse PA and feedback pulse PB.

On the other hand, when the encoder PG rotates once from the stop position (the position where it resonates with the origin of the table 12), a rotation signal PC is output from the encoder PG, and this rotation signal PC is
Input to the interrupt terminal of U1. This is it, 9, CP
U1 increments the counter circuit 6 by +1. Then, each time the encoder PG rotates once, the % CPU 1 increments the counter circuit 6 by 1, and when the counter circuit 6 outputs a counting completion signal to the CPU 1 (the table 12 moves beyond the correction interval and ), CPU
1 resets the counter circuit 6 and starts pitch error correction processing. In other words, CPU i updates the contents of point register 4 ( + IL ) % based on the contents fllK of point register 4, refers to the correction amount table T in storage device 5, and updates the contents of point register 4 (matches 11). Correction amount CP of correction number N=l
(CP=0 in figure f'g5) is added to the command pulse PA output by the interpolator 7. Then, when the mobile object 12 continues to move and the counter circuit 6 completes counting,
The CPU 1 performs pitch error correction processing, and when the corrected command pulse PA and feedback pulse PB match, the servo motor M stops rotating and the table 12 stops moving.

Among this pitch error correction processing, the updating processing of the point register 4 has a different sign in the moving direction of the table 12, and if the table 1z is moving away from the origin, it is +1, and the point register 4 is incremented by +1 in the direction closer to the origin. In this case, add -1. Further, after updating the point register 4, when finding the correction amount table TO correction number N that matches the contents of the point register 4 and adding the correction amount CP corresponding to the correction number N to the command pulse PA, the table 12 is When the direction of movement is toward the starting point, the sign of the correction amount CP is inverted and added to the command pulse PA output from the interpolator 7. For example, when the table 12 approaches the origin and the content of the point register 4 is 125, the counter circuit 6
When the count is completed, CPU 1 registers point register 4.
The contents of are updated (-1). Therefore, the content of the point register 4 becomes 124, and the %CPU 1 inverts the sign of the correction amount cp=+i corresponding to the correction number N=124 that matches this value (in other words, makes it "-1"), and the interpolator 7
is added to the command pulse PA.

Note that in this embodiment, the pitch error correction process is performed by combining the rotation signal PC of the encoder PG and the counter circuit 6, but if the correction interval matches the rotation signal PC of the encoder PG, , there is no need to use the counter circuit 6, and the CPU 1 only needs to perform pitch error correction processing every time the rotation signal pc of the encoder PG is input. In addition, in place of the rotation signal PC of the encoder PG, a position detection pulse PD is output from the encoder PG.
fc may be used in combination with the counter circuit 6 to perform pitch error correction processing.

As explained above, the present invention linearly approximates the pitch error data of a nine-feed P screw measured in advance, and calculates the position of the bending point of the straight line, the amount of error at the position, and the total movement stroke length of the moving body. Since it is input into the numerical control device, there is no need to input all pitch error data, the number of input operations can be kept to a minimum, input errors are less likely to occur, and the amount of error (measured value) can be input as is. Therefore, calculation errors do not occur when calculating the correction amount from the error amount as in the conventional method, and the work when inputting error data can be performed extremely easily. In addition, according to the above position data, the amount of error in the position data, and the total movement stroke length of the moving body, the size of the storage device [j is used to set a correction point, calculate the correction amount at that correction point, and store it in the self-memory. Therefore, it is possible to maximize and effectively utilize the storage area prepared for the correction amount in the storage device.

Therefore, since it is possible to set as many correction points as possible, compared to the conventional method where there are fewer correction points, the amount of correction that is fixed at the correction point increases and the accuracy decreases significantly. It has the inherited effect of being able to improve the shaft loading capacity.

[Brief explanation of the drawing]

Fig. 1 is a schematic configuration diagram showing an example of an apparatus for implementing the feed screw pitch error correction method according to the present invention, and Fig. 2 is a characteristic diagram showing the pitch error of the feed screw. ) respectively represent states obtained by linear approximation of (a), Figure 3 is a flowchart explaining the process of calculating the correction interval, Figure 4 (2) is a flowchart explaining the process of setting the correction amount, and Figure 5 is a flowchart explaining the process of setting the correction amount. FIG. 2 is a diagram showing the contents of a file table. 2...Numeric firsJra device, death...
Storage device, 12... Table (mobile object), 14
...Feed screw, PG...Encoder%
l”14... Servo motor, CP...
·Correction amount.

Claims (1)

[Claims] A numerically controlled machine comprising a numerical control device that controls the movement and stopping of a movable body such as a table of a machine tool body, and a servo motor that moves the movable body via a feed screw in response to a command signal from the numerical control device. In the feed screw pitch error correction method in a machine, the previously measured pitch error data of the feed screw is linearly approximated, and the position of the bending point of the straight line, the amount of error at the position, and the total movement stroke length are numerically controlled as described above. input into the device, set correction points for each of the above positions according to the above position, error amount, and total movement stroke length, calculate the correction amount at the correction point, and store it in the storage device in the numerical control device. After memorizing the
A method for correcting a feed screw pitch error in a numerically controlled machine tool, characterized in that each time the movable body reaches the correction point, the correction amount at the correction point is added to the command signal to control the servo motor.