JP7455488B2 - Method and device for diagnosing calibration values of reference instruments - Google Patents

Description

The present invention relates to a method and apparatus for diagnosing the calibration value of a reference device used for correction control of motion errors of machines such as machine tools.

FIG. 1 is a schematic diagram of a machining center 1 having three translational axes.
The spindle head 2 is a translational axis and is capable of translational movement with two degrees of freedom by Y and Z axes that are orthogonal to each other. The table 3 is capable of translational movement with one degree of freedom by the X axis, which is a translation axis and is perpendicular to the Y and Z axes. Therefore, the spindle head 2 has three translational degrees of freedom with respect to the table 3. Each axis is driven by a servo motor controlled by a numerical control device, and the workpiece is fixed on a table 3, and a tool is attached to the spindle head 2 and rotated to machine the workpiece into an arbitrary shape.

Motion errors of such machine tools include positioning errors and straightness. These motion errors are transferred to the shape of the workpiece and become a cause of shape and dimensional errors of the workpiece. The motion error can be measured using a reference device in which the relative position of the target is calibrated in advance. For example, in Patent Document 1, the linear part of a master block serving as a reference device is measured with a probe by linear interpolation feed, and the linear interpolation feed is determined based on the deviation between the calibration value and the measurement result regarding the shape of the straight part of the master block. A method is disclosed that calculates a correction parameter for an error and performs correction control based on the calculated correction parameter.
However, if there is an error in the calibration value of the standard, this becomes a measurement error.
On the other hand, in Patent Document 2, after performing linear motion measurement on one side of the check gauge to obtain first data, the check gauge is rotated 180 degrees around the axis to perform linear motion measurement on the same side. An invention is disclosed in which second data is obtained by performing measurement, and the difference between both data is used as a straightness error correction amount. By using such an inversion method, the calibration value of the reference device and the motion error of the machine can be measured separately.

Japanese Patent Application Publication No. 6-138921 Japanese Patent Application Publication No. 5-187868

In the inversion method described in Patent Document 2, it is necessary to set the reference device at the same position on the table and perform the measurement multiple times, so it takes time for the measurement. Further, if the measurement is performed only once per installation position, even if there is an error in the calibration value, it will not be noticed. Furthermore, it is not possible to diagnose the calibration value regarding the spacing of the targets of the reference device.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a method and apparatus for diagnosing the calibration value of a reference device, which can diagnose the presence or absence of an error in the calibration value of the reference device.

In order to achieve the above object, a first aspect of the present invention provides a table on which a reference device having two or more translational axes and a plurality of measurement points can be installed, and an advanced device on which a sensor can be attached. The calibration value of the reference device is determined using a machine in which the sensor held in the advanced device is capable of relative movement of two or more degrees of freedom in translation with respect to the reference device using the translation axis. A method for diagnosing errors,
a reference device installation step of installing the reference device at a predetermined installation position on the table;
a measuring step of detecting each of the measurement points of the reference device using the sensor to obtain a measurement value regarding the relative position of each of the measurement points;
an error value calculation step of calculating an error value based on the calibration value regarding the relative position of each of the measurement points obtained in advance and the measurement value;
After changing the installation position and executing the above multiple times,
a similarity calculation step of calculating the similarity between the respective error values at each of the installation positions;
a diagnosis step of diagnosing the presence or absence of an error in the calibration value based on the similarity;
It is characterized by carrying out.
Another aspect of the first invention is that in the above configuration, in the diagnosis step, it is determined whether all the similarities calculated in the similarity calculation step exceed a preset threshold, and at least The method is characterized in that it is diagnosed that an error has occurred in the calibration value when one of the degrees of similarity exceeds the threshold value.
Another aspect of the first invention is characterized in that, in the above configuration, when it is diagnosed in the diagnostic step that an error has occurred in the calibration value, a notification step is further executed to notify the occurrence of the error. do.
Another aspect of the first invention is that in the above configuration, in the similarity calculation step, the error value is obtained at one of the plurality of installation positions and the error value is obtained at another of the installation positions. The method is characterized in that the degree of similarity is calculated based on a Euclidean distance with the error value.
Another aspect of the first invention is that in the above configuration, in the similarity calculation step, the error value is obtained at one of the plurality of installation positions and the error value is obtained at another of the installation positions. The method is characterized in that the degree of similarity is calculated after performing sharpening filter processing on each of the error values.
Another aspect of the first invention is that in the above configuration, before executing the similarity calculation step, a difference between a maximum value and a minimum value of each of the error values calculated at each of the installation positions is calculated, The method is characterized in that the similarity calculation step is executed when the difference is greater than or equal to a preset difference threshold.
In order to achieve the above object, a second invention of the present invention provides a table capable of installing a reference device having two or more translational axes and a plurality of measurement points, and an advanced device capable of holding a sensor. The calibration value of the reference device is determined using a machine in which the sensor held in the advanced device is capable of relative movement of two or more degrees of freedom in translation with respect to the reference device using the translation axis. A device for diagnosing errors,
Measuring means that uses the sensor to detect each of the measurement points of the reference device installed at a predetermined installation position of the table, thereby obtaining a measurement value regarding the relative position of each of the measurement points;
an error value calculation means for calculating an error value based on a calibration value regarding the relative position of each of the measurement points obtained in advance and the measurement value;
can be executed multiple times by changing the installation position, and
Similarity calculation means for calculating the similarity between the respective error values at each of the installation positions;
Diagnostic means for diagnosing the presence or absence of an error in the calibration value based on the similarity;
It is characterized by having the following.
Another aspect of the second invention is that in the above configuration, the diagnostic means determines whether all the similarities calculated by the similarity calculation means exceed a preset threshold, and at least The method is characterized in that it is diagnosed that an error has occurred in the calibration value when one of the degrees of similarity exceeds the threshold value.
Another aspect of the second invention is characterized in that, in the above configuration, when the diagnosis means diagnoses that an error has occurred in the calibration value, the device further includes a notification means for notifying the occurrence of the error. .
Another aspect of the second invention is that in the above configuration, the similarity calculation means calculates the error value obtained at one of the plurality of installation positions and the error value obtained at another of the installation positions. The method is characterized in that the degree of similarity is calculated based on a Euclidean distance with the error value.
Another aspect of the second invention is that in the above configuration, the similarity calculation means calculates the error value obtained at one of the plurality of installation positions and the error value obtained at another of the installation positions. The method is characterized in that the degree of similarity is calculated after performing sharpening filter processing on each of the error values.
Another aspect of the second invention is, in the above configuration, before executing the similarity calculation means, calculates a difference between a maximum value and a minimum value of each of the error values calculated at each of the installation positions, The method is characterized in that when the difference is greater than or equal to a preset difference threshold, the similarity calculation means calculates the similarity.

According to the present invention, it is possible to diagnose whether there is an error in the calibration value of the standard by performing measurements while changing the position or direction of the standard on the table and calculating the similarity of multiple measurement results. . This eliminates the need to perform measurements at the same position multiple times as in the inversion method, thereby reducing the time required for measurements. Further, even when measurement is performed only once per installation position, the reliability of measurement can be improved because measurement is not continued with a large error in the calibration value. Furthermore, it is also possible to diagnose errors in calibration values regarding the spacing of the targets of the reference device.

1 is a schematic diagram of a machining center having translational axes of an X-axis, a Y-axis, and a Z-axis. 3 is a flowchart of a method for diagnosing a calibration value of a reference device in Form 1; 3 is a flowchart of a method for calculating similarity in Form 1. FIG. 2 is a block diagram showing a control mechanism of a machining center in a first embodiment. FIG. 3 is a schematic diagram of a touch probe and a reference device installed on a table in Form 1. FIG. 3 is a schematic diagram of the touch probe and the reference device after movement in Form 1; This is an example of an error value when the entire reference device in Form 1 is deformed. This is an example of an error value in which the seventh target in Form 1 has a calibration value error. This is an example of random error values due to variations in measurement reproduction in Form 1. This is an example in which an error value having a calibration value error in the seventh target in Form 1 is filtered. This is an example in which random error values in Form 1 are filtered. FIG. 3 is an explanatory diagram of calculated similarity in Form 1. FIG. 7 is a schematic diagram of a touch probe and a reference device installed on a table in a second embodiment. 12 is a flowchart of a method for diagnosing a calibration value of a reference device in a second embodiment.

Embodiments of the present invention will be described below based on the drawings.
[Form 1]
One embodiment of the present invention will be described based on the flowcharts of FIGS. 2 and 3. As an example of a machine to which the present invention is applied, the machining center 1 shown in FIG. 1 will be explained.
The machining center 1 is controlled by a control device (NC device) 21 shown in FIG. In addition to controlling the translational servo motors for the X, Y, and Z axes, the control device 21 executes a method for diagnosing a calibration value, which will be described later as a diagnosing device for a calibration value of a reference device according to the present invention. That is, the measuring means 26 measures the position of the target, the storage means 23 stores the measurement results, and the calculating means 22 performs various calculation processes such as calculating an error value and similarity degree, which will be described later. Therefore, the calculation means 22 functions as an error value calculation means, a similarity calculation means, and a diagnosis means of the present invention.
The control device 21 also includes an input means 24 for inputting calibration values of the reference device, and an output means 25 as a notification means for transmitting information to the operator.

In the present invention, as shown in FIG. 5, a touch probe 11 as a sensor is attached to the spindle head 2 as an advanced device, and a reference device 12 to be measured is installed at the first installation position of the table 3 (S1 : reference device installation step), and the Z-direction positions of a plurality of target balls (10 in this case) on the reference device 12 are measured with the touch probe 11 (S2: measurement step). The relative positional relationship of each target sphere is measured in advance using a highly accurate measuring device, and the measurement results are recorded as calibration values.
If the distance between the target sphere _P1 and Pi (i=1 to 10) determined from the calibration value is c(i), and the measurement result of the distance is m1(i), the first error value δ1(i) is calculated by the following formula. (S3: error value calculation step).

Next, as shown in FIG. 6, the reference device 12 is moved to the second installation position (S4: reference device movement (installation) step), and the same measurement as before the movement is performed (S5: measurement step). Note that the second installation position may not be in the same direction as the first installation position. If the distance between target sphere _P1 and Pj (j = 1 to 10) determined from the calibration value is c(j), and the measurement result of the distance is m2(j), the second error value δ2(j) is calculated by the following formula. (S6: error value calculation step).

Subsequently, the similarity between the obtained first error value and second error value is calculated (S7: similarity calculation step). In this example, there is an error value example 1 with a concave shape due to deformation of the entire reference device in Figure 7, an error value example 2 where there is an error in the calibration value at the seventh point in Figure 8, and an error due to measurement reproducibility in Figure 9. Error value example 3 in which the error value occurs will be explained as an example.
Regarding the obtained first error value and second error value, if the error in the calibration value occurring in a specific target as shown in FIG. 8 is to be diagnosed, filter processing is performed using a sharpening filter (S7- 1). Thereby, highly accurate diagnosis can be performed.
For example, by extracting the gradual change component of the error value using the Savitzky-Golay filter and subtracting it from the error values δ1 and δ2, as shown in the formula below, the steep change component of the error value δ1', δ2' can be calculated.

When filter processing is performed on error value example 2 in FIG. 8 and error value example 3 in FIG. 9, the results are as shown in FIGS. 10 and 11. On the other hand, when diagnosing errors that change slowly as shown in Fig. 7, filter processing is not necessary, and δ1'(i) = δ1(i), δ2'(i) = δ2(i) (i = 1 to 10).
Subsequently, the error value is normalized using the reference length (S7-2). In this example, the longer one of the lengths L1 and L2 of two 10-dimensional vectors, the first vector and the second vector, each having 10 error values as elements, the first error value and the second error value. is the reference length L. The length of the vector is calculated using the following formula.

By normalizing each vector using the reference length L as shown in the following equation, a first normalized vector e1(i) and a second normalized vector e2(i) having a length of 1 or less are obtained.

Next, the Euclidean distance |d| is calculated based on the first normalized vector and the second normalized vector as shown in the following equation (S7-3).

Since the first normalized vector and the second normalized vector are vectors with a length of 1 or less, the possible range of the Euclidean distance |d| is 0≦|d|≦2. In order to obtain a value that is easier to handle, the degree of similarity D is calculated using the Euclidean distance |d| as shown in the following formula (S7-4).

The degree of similarity D satisfies -1≦D≦1, and the closer it is to 1, the higher the degree of similarity between the two error values.
When the similarity is calculated for each error value example, the similarity is as shown in FIG. 12, and the similarity is high between error value example 1 and error value example 2 in which there is an error in the calibration value. On the other hand, in error value example 3, which is an error value due to measurement reproducibility, the degree of similarity is low.

Then, it is determined whether all the similarities calculated in S7 are within a preset threshold (S8: diagnosis step). Here, if at least one degree of similarity exceeds the threshold value, the output means 25 is used to output a warning message to the effect that an error has occurred in the calibration value of the reference device (S9: notification step). Therefore, the operator can take measures such as recalibrating the reference device.

According to the method and apparatus for diagnosing the calibration value of the reference device 12 of the first embodiment, measurement is performed by changing the position or direction of the reference device 12 on the table 3, and the similarity of a plurality of measurement results is calculated. , it is possible to diagnose whether there is an error in the calibration value of the reference device 12. This eliminates the need to perform measurements at the same position multiple times as in the inversion method, thereby reducing the time required for measurements. Further, even when measurement is performed only once per installation position, the reliability of measurement can be improved because measurement is not continued with a large error in the calibration value. Furthermore, it is also possible to diagnose errors in calibration values regarding the distance between the target spheres of the standard device 12.

[Form 2]
Next, another embodiment of the present invention will be explained.
The configurations of the machining center 1 and the control device 21 are the same as in the first embodiment, but here, as shown in FIG. The touch probe 11 measures the positions of a plurality of measurement points set on the top surface. The measurement surface of the reference device 12 is processed to have a high level of flatness, and the relative positional relationship of multiple measurement points on the measurement surface is measured in advance with a high-precision measuring device, and the measurement results are recorded as calibration values. ing.

Next, a method for diagnosing the calibration value of the reference device 12 in the second embodiment will be described based on the flowchart of FIG. 14.
First, from S11, a reference device installation position loop is executed. That is, the installation position s is changed and the processes of S11 to S17 are repeated a predetermined number of times. The installation position may be at another position parallel to the same axis, or may be installed in a direction parallel to another axis. In this example, an example will be described in which the device is installed at a different position in the direction parallel to the X axis.
In S12, the reference device 12 is installed at the installation position s (reference device installation step).
In S13, the position of the j-th measurement point Pj of the reference device 12 in the Z direction is measured with the touch probe 11, and a measured value Zms,j at the X-axis command value Xcs,j is obtained (measurement step).
In S14, an error value dZms,j is calculated by taking the difference between the measured value Zms,j and the calibration value Zcj of the reference device 12 (dZms,j=Zms,j−Zcj). Particularly here, an error value dZs,j is calculated by removing the error in the measured value due to the installation error of the reference device 12 (error value calculation step). This calculation will be described later.

In S15, it is determined whether the installation position loop is the second or subsequent time. Steps S16 to S17 are performed only for the second time or later.
In S16, the difference (error width) between the maximum value and the minimum value of the error value dZs,j at each installation position is greater than or equal to the preset difference threshold between the installation position used as a reference when calculating the similarity and the installation position s. Determine whether If the error width is less than the difference threshold, similarity is not calculated.
If it is determined in S16 that the error width is equal to or greater than the difference threshold, then in S17 the similarity between the measurement result at the installation position s and the measurement result at a certain installation position is calculated (similarity calculation step). Details will be described later.
In S18, it is determined whether all the similarities calculated in S17 are within a preset threshold (diagnosis step). If at least one degree of similarity exceeds the threshold value, a warning message is outputted using the output means 25 in S19 (notification step).

Next, a method of calculating the error value dZs,j in S14 will be explained. Here, explanation will be given based on an example in which the straightness of the Z-direction component of the X-axis is measured.
When measuring the straightness of the X-axis, if the measuring surface of the standard 12 is tilted with respect to the X-axis, the Z-direction error of the following formula due to the installation error of the standard 12 will occur when measuring the measurement point Pj at the installation position s. dZws,j occurs.

Here, Xcs,j is the X-axis command value when measuring the measurement point Pj at the installation position s.
The aws and bws in the formula 8 are obtained by solving the following equation using the least squares method or the like.
In addition, in the case of measuring the positioning accuracy, aws=0.

By subtracting the Z-direction error dZws,j from the error value dZms,j for all measurement points as shown in the following formula, find the error value dZs,j of the Z-direction component that removes the influence of the installation error of the reference device 12. be able to.

The error value dZs,j includes an error dZa(Xcs,j) that depends on the X-axis position and an error dZps of the calibration value of the measurement point Pj. Therefore, the error value dZs,j when the reference device 12 is installed at the installation position s and the measurement point Pj is measured can be expressed as in the following equation.

Next, the method of calculating the similarity in S17 will be explained. Here, an example will be described in which the similarity of the installation position 2 with respect to the measurement result of the installation position 1 is calculated.
The difference between the error value dZ 1 ,j obtained at installation position ₁ and the error value dZ ₂ ,j obtained at installation position 2 is expressed as the Euclidean distance of each error value ||d|| as shown in the formula below. be able to.

As shown in Equation 12, the Euclidean distance ||d|| depends only on the error dZa (Xcs,j) that depends on the X-axis position.
On the other hand, the lengths L ₁ and L 2 of N-dimensional vectors with N error values between installation position 1 and installation position ₂ as elements are calculated as shown below, and the longer one is set to the reference length L. shall be.

Using the Euclidean distance ||d|| and the reference length L, the degree of similarity D is defined as in the following formula.

If the error value dZ ₁ ,j obtained at installation position 1 and the error value dZ ₂ ,j obtained at installation position 2 are the same, D=1, exactly the opposite (dZ ₂ ,j=-dZ ₁ ,j), D=-1, and in the case of half (dZ ₂ ,j=-dZ ₁ ,j/2), D=1/2.
In particular, the larger the error dZps of the calibration value of the measurement point Pj is compared to the error dZa(Xci,j) that depends on the X-axis position, the closer the similarity D becomes to 1.

In the method and apparatus for diagnosing the calibration value of the reference device 12 according to the second embodiment, measurement is performed by changing the position or direction of the reference device 12 on the table 3, and the similarity of a plurality of measurement results is calculated. It is possible to diagnose whether there is an error in the calibration value of the reference device 12. This eliminates the need to perform measurements at the same position multiple times as in the inversion method, thereby reducing the time required for measurements. Further, even when measurement is performed only once per installation position, the reliability of measurement can be improved because measurement is not continued with a large error in the calibration value.

Modifications of the present invention will be described below.
Although the above embodiments 1 and 2 are explained using an example of measuring the straightness of the Z-direction component of the can be carried out.
The shape of the reference device is not limited to the above-mentioned forms 1 and 2, and can be changed as appropriate. When measuring positioning accuracy, a reference device having the distance between measurement surfaces as a calibration value may be used.
In the first and second embodiments described above, a machining center is used as an example, but other machine tools such as a multi-tasking machine, a lathe, and a grinding machine may be used as the applicable machine. Moreover, it is not limited to machine tools, but may also be industrial machines or robots.

1...Machining center, 2...Spindle head, 3...Table, 11...Touch probe, 12...Reference device, 21...Control device, 22...Calculating means, 23...Storage means, 24...Input Means, 25... Output means.

Claims (12)

It has two or more translational axes, a table on which a reference device having a plurality of measurement points can be installed, and an advanced device on which a sensor can be attached, and the sensor held by the advanced device is moved by the translational axes. , a method for diagnosing errors in the calibration values of the reference device using a machine capable of relative movement with two or more degrees of freedom in translation with respect to the reference device,
a reference device installation step of installing the reference device at a predetermined installation position on the table;
a measuring step of detecting each of the measurement points of the reference device using the sensor to obtain a measurement value regarding the relative position of each of the measurement points;
an error value calculation step of calculating an error value based on the calibration value regarding the relative position of each of the measurement points obtained in advance and the measurement value;
After changing the installation position and executing the above multiple times,
a similarity calculation step of calculating the similarity between the respective error values at each of the installation positions;
a diagnosis step of diagnosing the presence or absence of an error in the calibration value based on the similarity;
A method for diagnosing a calibration value of a reference device, the method comprising: performing the following steps. In the diagnosis step, it is determined whether all of the similarities calculated in the similarity calculation step exceed a preset threshold, and if at least one of the similarities exceeds the threshold, 2. The method for diagnosing a calibration value of a reference device according to claim 1, further comprising diagnosing that an error has occurred in the calibration value. The calibration value of the reference device according to claim 1 or 2, further comprising performing a notification step of notifying the occurrence of the error when it is diagnosed in the diagnosis step that an error has occurred in the calibration value. diagnostic method. In the similarity calculation step, the similarity is calculated based on the Euclidean distance between the error value obtained at one of the plurality of installation positions and the error value obtained at another of the installation positions. The method for diagnosing a calibration value of a reference device according to any one of claims 1 to 3, further comprising calculating the calibration value of a reference device. In the similarity calculation step, the error value acquired at one of the plurality of installation positions and the error value acquired at another installation position are each subjected to sharpening filter processing, and then The method for diagnosing a calibration value of a reference device according to any one of claims 1 to 4, characterized in that the degree of similarity is calculated. Before executing the similarity calculation step, calculate the difference between the maximum value and the minimum value of each of the error values calculated at each of the installation positions, and if the difference is greater than or equal to a preset difference threshold, the similarity is calculated. 6. The method for diagnosing a calibration value of a reference device according to claim 1, further comprising the step of calculating a degree. It has two or more translational axes, a table on which a reference device having a plurality of measurement points can be installed, and an advanced device capable of holding a sensor, and the sensor held by the advanced device is moved by the translational axes. , an apparatus for diagnosing errors in the calibration values of the reference device using a machine capable of relative movement with two or more degrees of freedom in translation with respect to the reference device,
Measuring means that uses the sensor to detect each of the measurement points of the reference device installed at a predetermined installation position of the table, thereby obtaining a measurement value regarding the relative position of each of the measurement points;
Error value calculation means for calculating an error value based on the calibration value regarding the relative position of each of the measurement points obtained in advance and the measurement value;
can be executed multiple times by changing the installation position, and
Similarity calculation means for calculating the similarity between the respective error values at each of the installation positions;
Diagnostic means for diagnosing the presence or absence of an error in the calibration value based on the similarity;
A diagnosing device for the calibration value of a reference device, comprising: The diagnostic means determines whether or not all of the similarities calculated by the similarity calculation means exceed a preset threshold, and if at least one of the similarities exceeds the threshold, The diagnosing device for the calibration value of a reference device according to claim 7, characterized in that it is diagnosed that an error has occurred in the calibration value. 9. The calibration value of the reference device according to claim 7 or 8, further comprising notification means for notifying the occurrence of an error when the diagnosis means diagnoses that an error has occurred in the calibration value. Diagnostic equipment. The similarity calculation means calculates the similarity based on the Euclidean distance between the error value obtained at one of the plurality of installation positions and the error value obtained at another of the installation positions. The diagnosing device for the calibration value of a reference device according to any one of claims 7 to 9, characterized in that the diagnostic device calculates the calibration value of a reference device. The similarity calculation means performs sharpening filter processing on the error value acquired at one of the plurality of installation positions and the error value acquired at another installation position, respectively. 11. The diagnostic device for a calibration value of a reference device according to claim 7, wherein the degree of similarity is calculated. Before executing the similarity calculation means, calculate the difference between the maximum value and the minimum value of each of the error values calculated at each of the installation positions, and if the difference is greater than or equal to a preset difference threshold, the similarity is calculated. 12. The diagnostic device for a calibration value of a reference device according to claim 7, wherein the degree of similarity is calculated by a degree calculation means.

Images