JP7344625B2 - Measurement start condition determination device - Google Patents

Description

The present invention relates to a shape measuring device.

Coordinate measuring machines (coordinate measuring machines) are widely used to measure the shape and dimensions of objects to be measured.

When measuring an object with a coordinate measuring machine, first, a measurement procedure corresponding to the object to be measured is prepared in advance as a measurement part program. In actual measurement work, after placing the object to be measured on a surface plate, moving stage, or rotary table, the operator issues a command to "start measurement" to the three-dimensional measuring machine. After that, the three-dimensional measuring machine automatically measures the object one by one according to the measurement part program according to a predetermined procedure. Furthermore, by using a system that automatically replaces the objects to be measured, measurements can be carried out automatically and continuously one after another. In recent years, in-line measurement systems, in which a three-dimensional measuring machine is installed beside a production line and sequentially measures workpieces, have come into use.

JP2008-241420A Japanese Patent Application Publication No. 2013-238573 Japanese Patent Application Publication No. 2014-21004 JP2017-062194 Patent 6482244 Special Public Interest 02-062006

When measuring the shape of an object to be measured, it is important that the temperature of the object be adjusted to a predetermined temperature. For example, if the temperature of the workpiece immediately after processing is high, it is necessary to wait until the temperature reaches an appropriate temperature. Alternatively, in order to increase measurement efficiency, shape measurement of the object to be measured may be started at a slightly higher temperature, but even in this case, it is necessary to adjust to a temperature at which temperature correction is known to be effective. .

It is necessary to determine through trial and error how long a workpiece must be left in a room immediately after processing to cool down to a temperature at which shape measurement can begin before mass production of the product begins. Here, the workpiece cools down from the surface, so even if the temperature of a certain part of the surface has fallen to a predetermined temperature, it does not necessarily mean that the temperature of the entire workpiece or the center of the workpiece has adjusted to the predetermined temperature. .
If the shape measurement is started before the workpiece has fully become familiar with the workpiece, the workpiece will still be judged as a failure even after temperature correction for linear expansion is taken into account. In this case, the work that should have passed the test will be discarded.
Alternatively, if the shape is measured again after some time, the temperature of the workpiece has adjusted to the predetermined temperature this time, so it is often determined that the workpiece has passed. In this case, the time and effort of re-measuring is wasted.
In general, even if it is known that it is somewhat wasteful, in order to avoid wasteful disposal or re-measurement, a longer acclimatization time than necessary is often set (for example, waiting two or more hours extra).
However, it is natural to want to eliminate such waste if possible and achieve further improvement in measurement efficiency in terms of time, cost, effort, etc.

An object of the present invention is to provide a shape measuring device that can further improve measurement efficiency when measuring the shape of an object to be measured.

The shape measuring device of the present invention includes:
A shape measuring device that includes a detector that detects an object to be measured, and a movement mechanism that relatively moves the detector and the object to be measured, and that measures the shape of the object to be measured with the detector. ,
moreover,
A thermo camera that obtains the temperature distribution of the measurement target,
a measurement start condition determining unit that determines whether or not it is appropriate to start measuring the shape of the object to be measured based on the surface temperature distribution of the object to be measured acquired by the thermo camera;
The method is characterized in that, when the measurement start condition determination section permits the start of shape measurement of the measurement object, the measurement of the shape of the measurement object is started.

In one embodiment of the invention,
The measurement start condition determination section includes:
a measurement start condition storage unit storing one or more temperature evaluation points set on the surface of the measurement object and a temperature condition range set for each temperature evaluation point;
a measurement start permission unit that permits the start of measurement of the measurement object when the temperature of the temperature evaluation point in the surface temperature distribution of the measurement object acquired by the thermo camera is within the temperature condition range; It is preferable to prepare.

The measurement start condition determination device of the present invention includes:
A measurement start condition determining device for determining whether or not to start measurement of the measurement object when measuring the shape of the measurement object with a shape measuring machine,
The measurement start condition determination device has an input layer and an output layer, and is configured with a neural network that performs machine learning.
The input data input to the input layer includes a surface temperature distribution of the measurement target obtained by a thermo camera,
The output data outputted from the output layer is characterized in that it indicates whether or not it is appropriate to start measuring the object to be measured.

In one embodiment of the invention,
In preparing training data that is a combination of input data and output data for machine learning the measurement start condition determination device,
Prepare multiple samples of the object to be measured that have been confirmed to pass the final pass/fail judgment result for processing accuracy,
heating or cooling the sample to different temperatures;
Obtain the surface temperature distribution at each temperature of the sample using a thermo camera, use this surface temperature distribution as input data,
Measure the shape of the sample at each temperature to obtain shape data, and obtain a pass/fail judgment result for the processing degree of each sample at each temperature from comparison with design value data, and use it as a pass/fail judgment result by temperature;
If the temperature-specific pass/fail judgment result matches the final pass/fail judgment result, a positive reward is given to the output data that permits the start of measurement, and a negative reward is given to the output data that does not permit the start of the measurement;
If the temperature-based pass/fail judgment result does not match the final pass/fail judgment result, the machine It is preferable to let them learn.

It is a figure showing an example of a shape measuring device. It is a figure showing an example of composition of a measurement start condition judgment part. FIG. 3 is a diagram illustrating temperature evaluation points. FIG. 3 is a functional block diagram of a motion controller and a host computer. It is a flowchart for explaining the operation of the shape measuring device. FIG. 2 is a diagram showing an example of a neural network. FIG. 2 is a diagram illustrating a table of sample data for training a neural network.

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described with reference to the drawings and reference numerals attached to each element in the drawings.
(First embodiment)
A first embodiment of a shape measuring device 100 of the present invention will be described.
FIG. 1 is a diagram showing an example of a shape measuring device 100.
The shape measuring device 100 may be installed in a measurement room that is provided exclusively for measurement and whose temperature is controlled. Alternatively, the shape measuring device 100 may be installed beside a manufacturing line (not shown), for example, along with a processing machine (not shown). In this case, the workpieces may be automatically sequentially loaded into the three-dimensional measuring machine using a loading/unloading device having a conveyor belt or a robot arm.

The shape measuring device 100 includes a coordinate measuring machine 200, a thermo camera 610, a measurement start condition determining section 620, a motion controller 300, and a host computer 500.

A coordinate measuring machine (CMM) includes a rotary table 210, a probe 220, and a drive mechanism 230.
The rotary table 210 is a workpiece table on which a workpiece is placed, and rotates about the Z-axis as a rotation center.
The probe 220 detects the surface of the workpiece in a contact or non-contact manner.
The drive mechanism 230 has a Z drive shaft 231 along the Z direction, a Y drive shaft 232 along the Y direction, and an X drive shaft 233 along the X direction, and drives the probe 220 three-dimensionally. move it.
A moving mechanism is configured by the cooperation of the rotary table 210 and the drive mechanism 230, and the workpiece and the probe 220 can be moved relative to each other in three dimensions.
The drive mechanism 230 and the rotary table 210 are provided with a drive motor (not shown) and an encoder (not shown).

A drive motor is driven by a drive control signal from the motion controller 300, and coordinate values (and rotation angles) sampled by the encoder are sent to the motion controller 300.
A measured value of the surface shape of the workpiece is obtained from the X, Y, and Z coordinate values of the drive mechanism 230 and the rotation angle of the rotary table 210.

Of course, the moving mechanism may be a mechanism that moves the probe 220 three-dimensionally in the X, Y, and Z directions, or a mechanism that moves the workpiece three-dimensionally in the X, Y, and Z directions. It may be a mechanism, or there may be a rotation mechanism such as rotation around the X axis, rotation around the Y axis, or rotation around the Z axis.

The thermo camera 610 is attached to the coordinate measuring machine 200, for example, and is installed so as to capture the object to be measured within the imaging range when the object is placed on the rotary table 210 or the surface plate. The thermo camera 610 is a so-called infrared thermography camera, and measures the surface temperature of the object in a non-contact manner.

FIG. 2 is a diagram showing an example of the configuration of the measurement start condition determining section 620.
The measurement start condition determination unit 620 determines whether or not it is appropriate to start measuring the shape of the measurement target, based on the thermography of the measurement target obtained by the thermo camera 610.
The measurement start condition determining unit 620 may be configured with a so-called computer including an image data input unit (image data input interface) that inputs thermography image data, a CPU, a ROM, a RAM, and the like.
The measurement start condition determination section 620 includes a measurement start condition storage section 630 and a measurement start permission section 640.

The measurement start condition storage unit 630 stores the positions (coordinates) of temperature evaluation points set on the surface of the object to be measured and the temperature condition range set for each temperature evaluation point.
The temperature evaluation point is a point on the surface of the measurement object that is selected as appropriate for determining whether the temperature of the measurement object is suitable for starting shape measurement (for example, see Figure 3). . A temperature range is set for each temperature evaluation point. For example, for the first temperature evaluation point, the temperature condition range is from 28°C to 10°C. Regarding the second temperature evaluation point, the temperature condition range is from 25°C to 15°C. Regarding the third temperature evaluation point, the temperature condition range is from 23°C to 18°C.

Where to set the temperature evaluation points and how to set the temperature range for each temperature evaluation point is set by the operator based on, for example, appropriately determined predetermined experiments and verifications.
For example, first, a sample of the object to be measured that has been processed according to the set value data is prepared. After processing, this sample should be cooled sufficiently until it reaches a predetermined temperature suitable for shape measurement (for example, standard temperature), and then measured with a shape measuring machine to obtain accurate shape data. . Confirm that the machining error from the design value data is within the tolerance and that the machining accuracy of the sample is acceptable as a product.

This sample is then heated to a temperature corresponding to the post-processing temperature, followed by cooling the sample for different waiting times.
Then, a plurality of contact temperature sensors (thermocouples and thermistors) are attached to the surface of the sample, and temperature data is acquired at a plurality of points (multiple points, such as 20 points or 30 points) on the sample surface. Alternatively, a sample may be imaged with a thermo camera, and several temperature evaluation point candidates (for example, 20 points, 30 points, etc.) may be specified from the imaged data.
Once the temperature has been measured, the shape of the sample is subsequently measured using a shape measuring device while maintaining that temperature to obtain shape data. Verify whether the sample can be determined to pass by comparing the shape data obtained through shape measurement and the design value data.
(Since the temperature of the surface of the sample has been obtained, temperature correction may be taken into consideration.)

At this time, by repeating the above experiment while changing the waiting time from heating to the start of measurement, it is possible to understand which point on the sample surface must be at a certain number of degrees or less before pass/fail determination of the sample can be made appropriately.
The object to be measured cools down from the surface, so if the object to be measured is large or has a complex shape, even if several points on the object's surface have cooled down to a certain temperature, the object to be measured will cool down from the surface. There is still heat in the core of things. It is not possible to say unequivocally which point on the surface reflects the temperature of the center based on the size and shape of the object to be measured, and if the temperature at any point on the surface falls below this temperature, the entire object to be measured is suitable for starting measurement. I can't say for sure whether it has cooled down to that temperature.
Therefore, through appropriately determined experiments and verification, conditions are found under which it is possible to judge whether or not it is appropriate to start measurement at several points on the surface of the object to be measured.
The conditions thus found are set in the measurement start condition storage section 630 as measurement start conditions.

When actually measuring the shape of the object to be measured, the measurement start permission unit 640 verifies whether the temperature of the object to be measured satisfies the measurement start conditions, and determines whether the temperature of the object to be measured satisfies the measurement start conditions. If so, give a command to the motion controller to start measurement. The measurement start condition determination unit 620 receives a thermography of the measurement object acquired by the thermo camera 610 in real time from time to time. The measurement start permission unit 640 acquires the temperature at each temperature evaluation point from the thermography of the measurement object, and determines whether each temperature satisfies the temperature condition range.

Note that when the object to be measured is placed on the surface plate or rotary table 210, if the relative position and orientation of the thermo camera 610 and the object to be measured are always determined, the temperature evaluation points can be specified in two dimensions. It would be sufficient to specify the coordinates in the image data.
If the relative position and orientation of the measurement object and the thermo camera 610 cannot be fixed uniquely, the temperature evaluation point may be recognized as a three-dimensional position on the surface of the measurement object using image recognition. good. For example, the work coordinate system of the object to be measured may be recognized based on the characteristic geometrical shape of the object.

FIG. 4 shows a functional block diagram of the motion controller 300 and the host computer 500.
The motion controller 300 includes a measurement command acquisition section 310, a counter section 320, a movement command generation section 330, a drive control section 340, and a measurement start command section 350.

The measurement command acquisition unit 310 acquires a measurement command from the host computer 500.
The measurement command is generated by arithmetic processing or the like by the host computer 500 based on the design data of the object to be measured and the location to be measured. The measurement command includes the target point on the probe trajectory and the target movement speed.

The counter unit 320 measures the displacement amount of the drive mechanism 230 and the rotary table 210 by counting the detection signal output from the encoder, and also measures the displacement of the probe 220 by counting the detection signal output from the probe sensor. . As a result, the current position of the probe 220, that is, the coordinate values of the workpiece surface is obtained.

The movement command generation unit 330 calculates a movement path of the probe 220 for measuring the surface of the object to be measured with the probe 220, and calculates a velocity vector along the movement path. For example, a vector command is generated to move the probe 220 from the current position to the next target point while autonomously adjusting the push amount.

The drive control section 340 drives and controls the drive mechanism 230 and the rotary table 210 based on the vector command calculated by the movement command generation section 330.

The measurement start command unit 350 receives the measurement permission command sent from the measurement start condition determination unit 620 and gives permission to the movement command generation unit 330 to start measurement.

The host computer 500 can be a so-called personal computer. The host computer 500 includes a CPU (Central Processing Unit), a memory, and the like, and controls the coordinate measuring machine 200 via the motion controller 300. Host computer 500 further includes a storage section 520 and a shape analysis section 530. The storage unit 520 stores design data (CAD data, NURBS data, etc.) regarding the shape of the object to be measured (workpiece), measurement data obtained by measurement, and a measurement control program (measurement part program) that controls the entire measurement operation. Store.

The operation of the shape measuring device 100 will be briefly described with reference to FIG.
An object to be measured is loaded into the three-dimensional measuring machine 200 by a loading/unloading device or the like (ST110).
Then, the measurement start condition determining unit 620 obtains a thermography of the measurement object from the thermo camera 610 (ST120).
Then, the measurement start condition determining unit 620 recognizes each temperature evaluation point in the thermography and acquires the temperature at each temperature evaluation point (ST130).

The measurement start permission unit 640 compares the temperature at each temperature evaluation point with the temperature condition range set as the measurement start condition, and if the temperature at all temperature evaluation points is within the temperature condition range, it allows the measurement to start. Permission is given to motion controller 300.
On the other hand, if the temperature at the temperature evaluation point is not within the temperature condition range, the measurement start permission unit 640 returns to ST120 and updates the thermography of the measurement object, while updating the motion controller 300 until the measurement start condition is met. Instruct to wait.
While the workpieces to be measured are sequentially introduced into the three-dimensional measuring machine 200, ST110 to ST160 are repeated until the termination condition is satisfied.

Note that the measurement start permission unit 640 has several methods for evaluating the temperature at a plurality of temperature evaluation points to determine that the measurement start condition is satisfied (whether or not it is determined that the temperature is within the temperature condition range). Variations are possible. For example, an appropriate evaluation function may be created, and based on the score of the evaluation function, it may be determined that the temperature of the object to be measured is within the temperature condition range.

According to the shape measuring device 100 of the embodiment described above, there is no need for an operator to attach a contact temperature sensor to each object to be measured, so that an unmanned operation is possible in which the objects to be measured are automatically measured one after another in, for example, in-line measurement.
Since the operator does not have to manually attach the contact temperature sensor to the object to be measured, the number of temperature evaluation points can be increased to a desired number. This can be expected to improve the accuracy of temperature evaluation of the object to be measured.

The temperature of the object to be measured can be accurately and efficiently measured using the thermography obtained by the thermo camera 610, and can be correctly compared with the measurement start conditions. Thereby, there is no need to wait excessively until the temperature of the measurement target reaches the appropriate temperature (until it cools down), and the shape measurement of the measurement target can be started as soon as the measurement start condition is satisfied.

(Second embodiment)
In the second embodiment of the present invention, the measurement start condition determination unit 700 is configured with a neural network that performs machine learning.
The neural network itself for machine learning is well known, and as illustrated in FIG. 6, it consists of an input layer, a middle layer, and an output layer.

Teacher data, which is a combination of input data and output data for causing the measurement start condition determining section 700 to perform machine learning, is prepared in the following manner, for example.
First, a sample of the object to be measured that has been processed according to the set value data is prepared.
After processing, the sample is sufficiently cooled down to a predetermined temperature (for example, standard temperature) suitable for shape measurement, and then the shape is measured using a shape measuring machine to obtain accurate shape data.
Confirm that the processing error from the design value data is within the tolerance and that the sample is acceptable as a product. It is a good idea to prepare multiple such samples.

After heating the sample to a temperature corresponding to the post-processing temperature, the sample is cooled down with different waiting times.
Image the sample with a thermo camera and record the image data.
This thermography image is the data input to the input layer. If the resolution of the input data that can be input to the input layer is too fine, it should be filtered as appropriate.

After acquiring the thermographic image, the shape of the sample is measured using a shape measuring device to obtain shape data.
Verify whether the sample can be determined to pass by comparing the shape data obtained through shape measurement and the design value data. (Since the temperature of the surface of the sample has been obtained, temperature correction may be taken into consideration.)
The pass/fail determination results obtained for each waiting time (that is, for each temperature) will be referred to as temperature-specific pass/fail determination results.

If the above experiment is repeated while changing the waiting time from heating to the start of measurement, a table as shown in FIG. 7, for example, is obtained.
We now know that all samples ultimately pass.
Therefore, if the temperature-based pass/fail determination result is pass, a positive reward is given to the output data that allows the start of measurement, and a negative reward is given to the output data that does not allow the start of the measurement.
Conversely, if the temperature-specific pass/fail determination result is FAIL, a negative reward is given to the output data that allows the start of measurement, and a positive reward is given to the output data that does not allow the start of the measurement.
Using such teacher data, a neural network is obtained which is machine-learned to determine the relationship between the thermography image as an input and the permission/disapproval of measurement start as an output, and this is referred to as the measurement start condition determining section 700.
When the thermography image of the object to be measured is given as input to the learned measurement start condition determination unit 700 that has been learned in this way, permission or disapproval (standby command) for starting the measurement is obtained as output data.

In the second embodiment, there is no need for the operator to specifically designate temperature evaluation points, and anyone can obtain a highly accurate measurement start condition determining section 700 as long as they have a combination of a thermography image and a pass/fail determination result for each temperature.

Note that the present invention is not limited to the above embodiments, and can be modified as appropriate without departing from the spirit.
In both the first and second embodiments, only the surface temperature of the object to be measured is used as a criterion for starting measurement (measurement start condition), but in addition to this, for example, indoor temperature (environmental temperature) data).

100...shape measuring device,
200...3D measuring machine,
210... rotary table, 220... probe,
230... Drive mechanism,
231...Z drive axis, 232...Y drive axis, 233...X-axis drive axis,
300...Motion controller,
310...Measurement command acquisition section, 320...Counter section, 330...Movement command generation section, 340...Drive control section, 350...Measurement start command section,
500... host computer,
520...Storage unit, 530...Shape analysis unit,
610...Thermo camera,
620, 700...Measurement start condition determination unit,
630...Measurement start condition storage unit,
640...Measurement start permission section.

Claims (2)

A measurement start condition determining device for determining whether or not to start measurement of the measurement object when measuring the shape of the measurement object with a shape measuring machine,
The measurement start condition determination device has an input layer and an output layer, and is configured with a neural network that performs machine learning.
The input data input to the input layer includes a surface temperature distribution of the measurement target obtained by a thermo camera,
A measurement start condition determination device, wherein the output data output from the output layer indicates whether or not it is appropriate to start measurement of the measurement target. The measurement start condition determination device according to claim 1 ,
In preparing training data that is a combination of input data and output data for machine learning the measurement start condition determination device,
Prepare multiple samples of the object to be measured that have been confirmed to pass the final pass/fail judgment result for processing accuracy,
heating or cooling the sample to different temperatures;
Obtain the surface temperature distribution at each temperature of the sample using a thermo camera, use this surface temperature distribution as input data,
Measure the shape of the sample at each temperature to obtain shape data, and obtain a pass/fail judgment result for the processing degree of each sample at each temperature from comparison with design value data, and use it as a pass/fail judgment result by temperature;
If the temperature-specific pass/fail judgment result matches the final pass/fail judgment result, a positive reward is given to the output data that permits the start of measurement, and a negative reward is given to the output data that does not permit the start of the measurement;
If the temperature-based pass/fail judgment result does not match the final pass/fail judgment result, the machine A measurement start condition determination device characterized by being trained.

Images