JPH0134749B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a pitch error correction method for correcting an error in the movement amount of a machine tool axis in the same direction as the movement direction.

For machine tools that perform machining by holding and moving a cutter or workpiece, manufacturing errors such as guide surface manufacturing errors, gaps between the guide surface and slide table, etc., and manufacturing errors during machining. Machining errors occur due to deformation due to cutting force, wear, etc., and it is impossible to completely eliminate such machining errors, and there is a certain limit to machining accuracy.

In recent years, with the development of electronic control technology such as numerical control (NC) and optical control technology, it has become possible to control the machine side with high precision, but it is necessary to fully utilize the functions of such control technology. In order to perform highly accurate feeding and positioning, it is necessary to improve the accuracy of the machine itself. However, although it is not impossible to manufacture the machine itself with a high precision of 1μ or more, in practice the manufacturing cost is too high and it is uneconomical, and the mechanical system is based on cutting force and thermal deformation. It is difficult to manufacture something that takes into account complex deformations.

Therefore, as a method that does not require the machine itself to be finished with high precision, it is possible to continuously measure the deviation of the slide table, etc. from the theoretical required path and feed back the results to the control unit. A method has been proposed in which control commands are corrected to satisfy accuracy. However, this correction method has the problem of complicating the measurement method and processing of measured values, so the reality is that it has not yet been implemented in a practical manner. Furthermore, a method has been proposed in which a correction table is obtained by discretely measuring errors with respect to the amount of movement of an axis in advance, and when the axis position is reached, the corresponding error amount is read out from the correction table and corrected. For example, the error (deviation from the actual position) X in the same direction as the movement direction of the X-axis movement amount (command value) is as shown in Figure 1, so the X-axis position and the corresponding A correction table is created by sequentially tabulating the error X′. Then, when moving the X-axis, the error data corresponding to the X-axis position is read out from the correction table each time. For example, when the X-axis position is X ₁ , the error data X ₁ ' is read out and subtracted from the command value, and the X-axis position is When it reaches X ₂ , read out the error data X ₂ ′ and subtract it from the command value, and then
When the X-axis position is _X3 , error data _X3 ' is read out and added to the command value for correction. Therefore, this correction method has the disadvantage that the correction table becomes large when the amount of error is large or the stroke of axis movement is long, and conversely, if the size of the correction table is limited and reduced, However, the drawback is that only coarse corrections can be made.

SUMMARY OF THE INVENTION Therefore, it is an object of the present invention to provide a method for correcting pitch errors in machine tools that can be flexibly applied to large machines with long axis movement strokes and that can perform corrections with high accuracy.

This invention will be explained below.

The present invention relates to a method for correcting pitch errors in machine tools, which includes sequentially measuring the error in the same direction as the movement direction of the axis of the machine tool, and calculating a straight-line approximation error characteristic curve from this measurement error data.
The breakpoint position indicates the position of the axis corresponding to the bending point of the error characteristic curve, and the amount of movement of the axis for an error of a predetermined unit length, which is an integer fraction of the interval between the breakpoint positions, is determined on the error characteristic curve. Memorize the pitch of each straight line slope,
For the movement of the axis, each time it passes a correction point corresponding to the pitch mentioned above, it is corrected by a predetermined unit length, and when passing each breakpoint position, the change in the slope of the characteristic curve before and after is observed to check the error. It is determined whether or not to perform correction to prevent the increase. That is, in the present invention, first, a linear approximation error characteristic curve l1-l2-l3-l4 as shown in FIG. 2 is determined from X-axis measurement deviation data as shown in FIG. In this case, the amount of X-axis movement necessary to produce an error of a predetermined unit length, for example, 1 μ, is determined as pitches p1 to p4 of each error characteristic curve l1 to l4. and,
The bending point of the adjacent error characteristic curves l1 and l2 is
CR1, the bending point of error characteristic curves l2 and l3
CR2, the bending point of error characteristic curves l3 and l4 is CR3
In addition, the amount of movement of the X axis for these bending points CR1 to CR3 is the break point position.
BP1 to BP3, and the starting point that is the origin of correction is SP,
Let the stroke limit of the correction range be SL. Therefore, the block between the start point SP and the breakpoint position BP1, and the breakpoint position BP1
and BP2 as a block, between break point positions BP2 and BP3 as a block, and between break point position BP3 and stroke limit SL as a block, and the interval between each block is at the corresponding pitch p1 to p4. Adjust so that it is just divisible. In other words, the slopes and intersection points of the error characteristic curves l1 to l4 are determined so that the intervals between blocks are integral multiples of the corresponding pitches p1 to p4, and the breakpoint positions of the error characteristic curves l1 to l4 determined in this way are determined. BP1～BP
3 and the pitches p1 to p4 of each block ~ to the third
With the correspondence as shown in the figure, the memory M along with the starting point SP
1 to M9 as a correction table.

Based on the block position data stored in the correction table and the pitch data of that section,
In this invention, when moving within the same block, each time the block moves by a predetermined pitch amount,
A 1μ correction is made, and when moving to another block, the change in the slope of the characteristic curve before and after is checked to determine whether or not to make the correction so as not to increase the error. A correction of 1μ is performed each time the pitch is moved. The state of this correction method will be explained using the example of FIG. 4. This example shows movement between the block corresponding to the error characteristic curve l3 and the block corresponding to the error characteristic curve l4. Therefore, the pitch in the block is p3,
Since it is pitch p4 in the block,
In this invention, when moving within a block, correction points CM31, CM32,... are determined for each pitch p3, and when moving within a block, correction points CM41, CM42,... are determined for each pitch p4, and each time the block passes through a correction point, correction points CM41, CM42,... A 1μ correction is applied to the For example,
Considering the case where the current position of the axis is X ₁ and the command value moves it to position X ₂ within the same block,
Since it passes through correction points CM33 and CM32, the correction point
CM33 and CM32 are each corrected by 1μ. In this case, the detection of whether or not there are correction points CM33 and CM32 is based on the difference between the current position and the previous correction point (the correction point with the largest position among the correction points smaller than the current position). This is done by asking whether the pitch is greater than p3. Therefore _, when moving from _position , by performing a 1μ correction at the correction point CM33 and updating the previous correction point for the current position to CM33, similarly when the X axis reaches the correction point CM32, a 1μ correction is performed at the correction point CM32. obtain.

Next, considering the case of moving from position X ₁ to X ₃ of another block, the pitches p3 and p4 are different between the blocks, but the correction points within each block are the same as described above. Perform the same correction. i.e. the breakpoint position
BP1 to BP3, starting point SP and stroke limit
Since SL is stored in memories M1 to M4,
It is possible to monitor which block the axis position is currently in, and within the block, corrections of 1 μ are performed at correction points CM33, CM32, and CM31 as described above. Then, when it passes through the correction point CM31 and reaches the breakpoint position BP3, it is recognized that the current position is the breakpoint position BP3, so it is necessary to judge whether or not to perform correction at this breakpoint position BP3. do. Fourth
In the case shown in the figure, since the signs of the slopes of the characteristic curves before and after are different, no correction is performed. however,
When passing the break point position BP3, the previous correction point that should be remembered is from CM31.
Update to BP3 and change this breakpoint position BP3
When the X-axis moves by pitch p4 from , a correction of 1μ is performed at the correction point CM41. After that, correction will be made within the block. Then, similar correction is performed at the correction point CM42, and the X-axis reaches the command value position _X3 within an error range of 1 μ.

Note that although we have explained the case of moving from the lower block to the upper block, the same applies to moving from the upper side to the lower block, and in this case too, the previous correction point on the lower side is A new correction point is determined based on the reference. In addition, in the example of Fig. 4, the position where the error characteristic curves l3 and l4 become 0, and the break point position
The interval with BP3 is an integer multiple of each pitch p3, p4, but it does not necessarily have to be an integer multiple like this, and the interval between breakpoint positions is an integer multiple of each pitch. That's fine.

Here, for the specific correction method, refer to the flowcharts in Figures 5A to 5C, and start from position _X1 in Figure 4.
The case of moving to X ₂ or X ₃ will be explained as an example.

First, when the correction operation starts (step S1), it is determined whether the correction table as shown in FIG. 3 is correct (step S2). In this case, it is determined whether the breakpoint positions are arranged in descending order, etc., and if it is not correct, returns, and if it is correct, it is checked whether the initial setting of axis synchronization (origin setting) is achieved. It is determined (step S3), and if synchronization is achieved, the block in which the machine is currently stopped is set in register 1 as memory (step S4). In this example, the machine is stopped at the block, so
Register 1 is set to "3". and,
Finally, after confirming whether or not the stroke limit SL is within the stroke limit SL (step S5), the contents of registers 2 and 3 are set to ``0'', and the contents of register 1 to ``3'' are set to register 4 (step S5).
S6), the pitch (in this example) of the block (in this example) stored in register 4 in register 5
p3) and determine whether it is positive or negative (step S7). If the pitch is 0, it is left as is, if it is positive, register 3 is set to "1" (step S8), and if it is negative, the pitch in register 5 is converted to an absolute value and corrected, and the pitch is 3 is set to "-1" indicating a negative value (step S9).
In this example, pitch p3 is positive, so "1" is set in register 3. Thus register 5
When it is determined whether the correction point is positive or negative, the correction point immediately before the current position (CM34 in this example) is set in register 6 (step S10), and the correction point is set from the break point position BPF (BP2 in this example) of the previous block. The command position XN (X ₂ , X ₃ in this example) is subtracted and its sign is determined (step S11). If the difference is negative, the routine moves to the lower block (described later), and if it is less than 0, "1" is added to the contents of register 4 ("3" in this example) and the upper block is moved. Find the block (in this example) (step S12),
Next, the command position XN is subtracted from the break point position BPR, and its sign is further determined (step S13). If the difference _is less than 0 (in this example, in the case of command _position Assuming that the movement is within a block, we proceed to a movement routine within the same block as described below. in this way,
Based on the size of the breakpoint position of the block corresponding to the current position, the breakpoint position of the block one block above the current position, and the current position, it is possible to determine whether the command position is within the block, the upper block, or the lower block. Determine whether it is present.

Therefore, when moving to a movement routine within the same block, such as when command value X ₂ is given for current position X ₁ , first the command position (X ₂ in this example) is set in register 7, and , the previous correction point from register 7 that stored this command position
The value of register 6 that stores CM34 is subtracted and the difference is set as the content of register 7 (step S14), and it is determined whether the difference is positive or not (step S14).
S15). If the difference is negative, register 3 is subtracted from register 2 and the difference is set in register 2, and register 5 is subtracted from register 6 and the difference is set in register 6 (step S16). ), if the difference in the register 7 is positive, the process moves to the next step S17. In this example, the value of register 7 is positive, so the process moves to step S17, and the pitch (p3) of the block is determined from the contents of register 7, which stores the difference between command position _X2 and correction point CM34. The contents of register 5, which stores , are subtracted, the difference is set in register 7, and it is checked whether it is negative. If the difference is positive or 0, the content "0" of register 2 and the content "1" of register 3 are added and set in register 2, and the content "CM34" of register 6 is added.
and the contents "p3" of register 5 are added and the added value is set in register 6 (step S19).
As a result, the contents of the register 2 for accumulating correction amounts become "1", and the previous correction point for the current position is updated from CM34 to CM33. The above operation will be repeated until the contents of register 7 become negative, so that at the point when register 7 becomes negative, register 2 will eventually move from position X ₁ to X ₂
This means that the amount of correction for moving up _to
This means that you will remember CM32. In this way, when the judgment in step S18 becomes negative, register 6 updates the current _position is set in register 2,
This corrected value is used as the axis command value (step
S21). Therefore, in this example, the command value
This is a command to perform subtraction correction of 1μ twice for X ₂ .

On the other hand, if the command position is set to X ₃ , the step
From S13, routine B moves to the upper block, but in this case, the correction in the old block (in this example) is first performed in the same way as described above.
S40). Then, add "1" to register 1 that stores the number of the current block ("3" in this example) to make it "4" (step S41), and set the breakpoint corresponding to this block (in this example). The position BP (BP3 in this example) is set in the register 6, and the contents of the register 4 are set to "4" corresponding to the updated block (step S42).
Next, the pitch (in this example) of the block (in this example) stored in register 4 is stored in register 5.
p4) and determines whether it is positive or negative (step S43). If it is positive, "1" indicating positive is set in register 8 (step S44), and if it is negative, the pitch of register 5 is set. is converted into an absolute value and made positive, and "-1" indicating a negative value is set in register 8 (step S45). "0" if 0
(step S44A). In this example, pitch p4 is negative, so "-1" is set in register 8. And register 3 and register 8
Determine whether the sign of has changed (step
S46), if there has been a change, the process jumps to step S49 to determine whether or not the correction in the new block has been completed without performing correction processing. If the breakpoint position BP3 has not been completed, register 6 that stores the breakpoint position BP3 is added to register 5 that stores the pitch p4, and the added value is set in register 6 (step S48). The contents of and the contents of register 8 are added and set in register 2 (step S47). However, in this example, since "-1" is set in register 8, register 2
The content becomes "-1", and the above process is repeated until the correction in the new block is completed. In addition,
If the signs of register 3 and register 8 are the same and do not change, the process of step S47 is performed, so that the contents of register 8 are added to register 2, thereby performing correction. When the correction in the new block is completed in this way, the processes from step S20 described above are executed. Therefore,
The contents of register 2 in this case are the correction point
Subtraction correction at CM33, subtraction correction at correction point CM32, subtraction correction at correction point CM31, breakpoint position
No correction is performed in BP3, and correction points CM41 and
Since addition correction is performed for each CM42, one subtraction correction is performed in total for positions X ₁ to X ₃ .

Furthermore, as shown in FIG. 5B, the routine for moving to the lower block is simply the reverse of the operation for moving to the upper block, but the principle is exactly the same.

As described above, according to the correction method of the present invention, the error in the movement of the axis is approximated by a straight line, and the intersection point and the straight line gradient are stored as parameters,
Since correction is performed based on this, there are advantages in that the memory capacity is small even for large machines with long axis movement strokes, parameters can be easily changed, and high accuracy can be guaranteed.

In addition, although the above mentioned the correction of the X axis,
Similar corrections can be made for other axes. In addition, the error is required to determine the pitch of each error characteristic curve.
Although it is assumed to be 1μ, the length of the unit can be arbitrarily determined.

[Brief explanation of drawings]

Fig. 1 is an error graph showing an example of measuring the error in the same direction X' as the moving direction of the machine tool against the amount of movement of the X-axis, Fig. 2 is a diagram showing an example of the error characteristic curve according to the present invention, and Fig. 3 is a diagram showing an example of the correction table according to the present invention, FIG. 4 is a diagram for explaining the principle of the correction method according to the present invention, and FIGS.
C is a flowchart showing an example of a specific correction method of the present invention. l1 to l4...error characteristic curve, SP...starting point,
SL……Stroke limit, BP1~BP3……
Breakpoint position, M1 to M9...memory.

Claims (1)

[Claims] 1. Sequentially measure the error in the same direction as the movement direction of the axis of the machine tool with respect to the amount of movement, obtain an error characteristic curve of linear approximation from this measurement error data, and find the error characteristic curve of the axis corresponding to the bending point of the error characteristic curve. A breakpoint position indicating a position and a pitch which is an integer fraction of the interval between the breakpoint positions and whose movement amount of the axis for a predetermined unit length error is each linear slope of the error characteristic curve are stored. Then, the movement of the axis is corrected by the predetermined unit length each time it passes a correction point corresponding to the pitch, and when passing each break point position, the error is not increased. A pitch error correction method for a machine tool, characterized in that whether or not to perform correction is determined based on a change in the slope of a characteristic curve before and after the pitch error.