JPH01147608A - Correction system for positioning error - Google Patents

Abstract

PURPOSE:To perform the accurate position control in the title system by estimating the positioning error of the mobile part of a numerically controlled machine to hold it as an error table and obtaining a positioning error with a desired shift target distance of the mobile part by the interpolation arithmetic based on said error table and by means of the Gregory Newton method to perform the position correction of the mobile part. CONSTITUTION:An error table contains the correspondence between the distance x0, x1...xi-1 from the initiations of the points P1...Pi set on a distribution curve of the pitch errors corresponding to the shift distances and the error values y0, y1...yi-1 of the pitch errors m1...mi of the point P1...pi respectively. Then the error table is stored in a ROM. The Gregory Newton method is applied to calculate a pitch error (y) with a desired shift target distance (x) at an arithmetic processing part 12 by the curve approximation.

Description

Detailed Description of the Invention [Industrial Application Field] The present invention relates to a method for correcting positioning errors of movable parts in numerically controlled machines. An error table is created that associates the positioning error of the movable part at each division point divided by The present invention relates to a positioning error correction method that performs interpolation calculations using curve approximation.

[Background of the Invention] In recent years, automation has been promoted in manufacturing factories with the development of numerically controlled machine tools and various industrial robots. For example, numerically controlled machine tools use mechanical drive systems that move movable parts such as tables and tools to desired positions by rotationally driving a ball screw connected to a motor based on command signals.

By the way, it is well known that in a mechanical drive system using a ball screw as described above, there is an error in the processing of the ball screw, an error called a so-called pitch error.

Eliminating this pitch error is virtually impossible due to limitations in manufacturing and processing accuracy of ball screws, and usually the pitch error is corrected using various methods to improve the positioning accuracy of the movable part. things are being done.

For example, the pitch error is an amount that mainly depends on the processing accuracy of the ball screw, so by moving the movable part from the starting end of the ball screw to the terminal end, dividing the entire moving distance at regular intervals, and determining the pitch error at each division point. , a pitch error distribution curve specific to the mechanical drive system is obtained. Therefore, (1)
Limit switches and the like are provided along the moving path of the movable part at predetermined intervals according to the pitch error distribution, and the position of the movable part is detected by the limit switches and the correction value corresponding to the pitch error distribution at the detection point is applied to the motor. (2) Set a correction point using software instead of a limit switch, etc., and perform correction according to the pitch error distribution when the moving part reaches that correction point. (3) Determine two points before and after the movement target position of the movable part, approximate a straight line between the two points from the pitch error distribution of each point, and calculate the pitch error at the target position. A correction method is used.

In this case, methods (1) and (2) are so-called point correction methods, and if the pitch error is irregularly distributed over a wide range, a large number of correction points are required, making it difficult to calculate the structure and correction value. In addition to the disadvantage of being complicated, since it is a point correction, there is a limit to the ability to perform accurate correction, and there is an essential problem of poor positioning accuracy.

In addition, method (3) requires two steps before and after the movement target position of the movable part.
Since the pitch error amount at the moving target position is calculated by linear approximation based on the pitch error at the point,
Although positioning accuracy is improved compared to methods (1) and (2), there is a limit to positioning accuracy because pitch errors with a curved distribution are approximated by a straight line, making it difficult to achieve high-precision positioning. This is causing an inconvenience.

Furthermore, in order to increase the positioning accuracy using this correction method, it is necessary to infinitely subdivide the segment to obtain the pitch error distribution over the entire travel distance of the ball screw, which in effect makes it impossible to increase the accuracy beyond a certain limit. is impossible.

[Purpose of the Invention] The present invention has been made to overcome the above-mentioned disadvantages, and the movable range of a movable part in a numerically controlled machine is divided at regular intervals, and positioning errors of the movable part at division points are eliminated. The purpose is to provide a positioning error correction method that enables highly accurate positioning by approximating the positioning error at the moving target position to a curve using the Gregory-Newton method based on an error table that correlates the be.

[Means for Achieving the Object] In order to achieve the above object, the present invention provides a positioning error correction system for a movable part in a numerically controlled machine, in which the range of motion of the movable part is divided into predetermined intervals. Each point (xo
An error table is created in which the positioning error (yoSy3, y2...yi) of the movable part at each point is associated with the positioning error (yoSy3, y2...yi) of the movable part in the vicinity of the target movement distance according to the target movement distance X of the movable part. multiple predetermined points) (X, 1 , XJ, l , X, +2
...) and calculate the corresponding positioning error (5'J s 'lJ+l % yj+i...) from the error table.
), and the positioning error y at the movement target distance X is calculated as (yj.2 2yi.1+Yj+ (n 2)・(
3'a+3-3(yt+2-ya++)-yi)))
+...However, the position of the movable part is corrected based on the positioning error y at the calculated movement target distance do.

[Embodiments] Next, preferred embodiments of the positioning error correction method according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a schematic diagram of a numerically controlled machine tool to which a positioning error correction method according to the present invention is applied.

In the figure, reference numeral 10 is a control device that controls the machine tool main body 30, and includes an arithmetic processing section 12 composed of a microprocessor or the like.

The arithmetic processing unit 12 has an output port 1 via an internal bus 14.
6, human power port 18, keyboard 20, display 2
2. RAM 24 and ROM 26 are connected respectively. In this case, the output capo 16 has the function of sending various operation command signals to the machine tool main body 30. The human capo 18 has a function of taking in various sensor information such as a position detection signal from the machine tool main body 30 into the control device 10. Further, the keyboard 20 and the display 22 are connected to the control device 1, respectively.
0 or input operation instructions to the machine tool main body 30 from the keyboard 20, and display and monitor the status of each part of the control device 10 or the machine tool main body 30.

A random access memory (RAM) 24 stores an operation command program for the machine tool body 30, a positioning error calculation program to be described later, and the like. Furthermore, a read-only memory (ROM) 26 stores an error table (details will be described later), a basic operation program for the control device 100, and the like.

On the other hand, the machine tool main body 30 is connected to the ball screw 32 via a ball screw 32, a motor 34 such as a pulse motor which is a driving means for rotationally driving the ball screw 32, and an engagement member 36. A table 38, which is a movable part that moves in the axial direction (direction of arrow X), and a position detector 4, such as an inductosin, that detects the position of the table 38.
It is composed of 0.

In FIG. 1, only the mechanical drive system in the X direction is shown for the machine tool main body 30 to simplify the explanation, but there are also mechanical drive systems in the other two directions orthogonal to the X direction, and each has a movable part. Configured to be movement controlled. These mechanical drive systems including the motor 34 and the position detector 40 are respectively coupled to the output port 16 and the manual port 18 of the control device 10, and are sequentially outputted from the arithmetic processing unit 12 according to the operation command program read into the RAM 24. A predetermined operation is performed in response to an operation command signal, and various machining operations are performed.

The machine tool to which the positioning error correction method according to the present embodiment is applied is basically constructed as described above, and its operation and effects will be explained next.

In FIG. 1, a ball screw 32 that constitutes a mechanical drive system is shown.
It is inevitable that there will be machining errors in screw manufacturing.

Therefore, even if a command signal corresponding to a movement command value for moving the table 38 to a desired position is given to the motor 34, a slight deviation will occur in the position at which the table 38 stops. This deviation is called a pitch error. In this case, since the pitch error mainly depends on the machining accuracy of the ball screw 32, once the ball screw 32 constituting the mechanical drive system is determined, the pitch at each position can be known in advance.

That is, the table 38, which is a movable part, is connected to the ball screw 32.
If the ball screw 32 is moved from the starting end to the terminal end, the entire moving distance is divided at regular intervals, and the pitch error is determined for each division point, a pitch error distribution curve specific to the ball screw 32 is obtained.

Figure 2 shows this pitch error distribution curve.
The horizontal axis is the moving distance (mm) of the table 38, and the point P+
, P2...PL are points divided at predetermined intervals on the ball screw 32. Also, the vertical axis is the pitch error (
μm). In this case, each point PI, P2...
When the pitch error m, Sm2...mi at Pt is measured and plotted, the pitch error ITLt, m2...
- mi is not uniform in each section and exhibits an irregular curved distribution unique to the ball screw 32.

The positioning error correction method according to this embodiment is for highly accurately correcting pitch errors having a distribution curve as shown in FIG. Therefore, each point shown in Figure 2) P+, P2...
・Distance xo,
Vt-+, an error table shown in FIG. 3 is created and stored in the ROM 26. Then, the pitch error y at the desired movement target distance X is calculated by curve approximation in the arithmetic processing unit 12 using the Gregory-Newton method described below.

First, two points before and after the movement target distance X are selected. In this case, points P2, Pl and P in FIG.
, ,P are selected and the error table (
It is assumed that pitch errors (yl, yl, y3, y4) corresponding to each point P2, P-, Pl, Ps are read out from FIG. 3). Here, according to the Gregory-Newton method, the pitch error y with respect to the moving target distance
1)・(y3-2 y2+y+ +-(n-2)・(y
4-3(V3-yl-yl))...(1) However, it can be calculated as x=x, +nh. By adding this pitch error y as a correction value to the movement command signal to the motor 34, extremely accurate correction of the pitch error becomes possible.

In the above embodiment, 1<n<2, and the two points closest to the movement target distance X are selected. On the other hand, the movement target distance X is Pl near the starting end of the ball screw 32, PI-1 between 22 or near the end
If it is between Pl, the four points closest to the movement target distance X are selected. That is, in FIG. 2, when the movement target distance X is P+ <
−, when <x<Pt, set 2<n × 3, and point P
Select I-3, Pl-2, Pi-11PI and
) to find the correction value.

Next, the above positioning error correction method will be further explained using a specific example.

FIG. 4 shows a specific example of the pitch error distribution curve of the ball screw 32, in which the moving distance is divided into 100 mm intervals along the ball screw 32 at each point P to P6.
The pitch error amounts 1 to yl+ are measured and preset. In this case, h=100 mm. Further, the movement distances xl to x6 of each point P1 to P6 and the pitch error y at each point P to P6.

The error table that associates y6 to y6 has the values shown in parentheses in FIG. 3.

Now, if the target moving distance
, P2, Pl, P are selected and the error table shown in FIG. 3 stored in the ROM 26 is referred to. and the corresponding pitch error y.

(1,0μm), yl (4,0μm), 3'!
(8,0 μm) and ya (9,0 μm) are read out. In this case, from equation (1) above, the moving target distance @X = 2
The pitch error y at 50mm is (9,0-3(8,0-4,0) -1,0)))=6
．． It becomes 125 μm.

On the other hand, when the pitch error y is determined by linear approximation using gradient correction according to the prior art, the pitch error y when performing linear correction between the two points P2 and P3 before and after the moving target distance X = 250 mm is 'l = 'lz+'l'
...It can be found by (3), and (
From equation 2), and from equation (3), y = 4.0 + 2.0 = 6.0 μm, and the pitch error is corrected by only 6.0 μm at the moving target distance X. On the other hand, according to the positioning error correction method according to the present embodiment, the pitch error at the movement target distance X is corrected by 6.125 μm, making it possible to correct the pitch error more accurately.

In addition, in the above-mentioned embodiment, an example was explained in which the pitch error at each of the four points before and after the moving target distance X is used to calculate the pitch error, but it is also possible to select four or more points for calculation. It is possible.

[Effects of the Invention] As explained above, according to the present invention, the positioning error of the movable part in a numerically controlled machine is measured in advance and maintained as an error table, and the desired movement of the movable part is determined based on the error table. The positioning error at the target distance is calculated by interpolation using the Gregory-Niniton method, and the position of the movable part is corrected. in this case,
Since the positioning error in the movement target distance can be calculated by curve approximation using a high-order function, it is possible to obtain extremely high correction accuracy, and there is an advantage that accurate position control can be performed.

Although the present invention has been described above with reference to preferred embodiments, the present invention is not limited to these embodiments.
Of course, various improvements and changes in design are possible without departing from the gist of the present invention.

[Brief explanation of the drawing]

Fig. 1 is a schematic configuration diagram of a numerically controlled machine tool to which the positioning error correction method according to the present invention is applied, Fig. 2 is an explanatory diagram of a positioning error distribution curve, and Fig. 3 is an illustration of each point and positioning error in Fig. 2. FIG. 4 is an explanatory diagram of a pitch error distribution curve in a ball screw that is an embodiment of the positioning error correction method according to the present invention. 10... Control device 12... Arithmetic processing unit 2
4...RAM 26...ROM30
...Machine tool body 32...Ball screw 34...
・Motor 38...Table 40...Position detector Patent applicant: Toshiba Machine Co., Ltd.

Claims (1)

[Claims] (1) A positioning error correction method for a movable part in a numerically controlled machine, in which the movable range of the movable part is divided at predetermined intervals h at each point (x_0, x_1, x_2...x_i) and the position of the movable part at each point. Positioning error (y_0
, y_1, y_2...y_i), and a plurality of predetermined points (x_
j, x_j_+_1, x_j_+_2...) and calculate the corresponding positioning error (y_j, y
_j_+_1, y_j_+_2...), and the positioning error y at the movement target distance x is calculated as y=y_j+n(y_j_+_1-y_j+1/2(n
-1)・(y_j_+_2-2y_j_+_1+y_j
+1/3(n-2)・(y_j_+_3-3(y_j_
+_2-y_j_+_1)-y_j)))+...However,
A positioning error correction method characterized in that the position of the movable part is corrected based on the positioning error y at the calculated moving target distance x, which is calculated using the Gregory-Newton method: x=x_j+nh.