JP7612282B2 - Surface condition estimation method and surface condition estimation system - Google Patents

Description

The present invention relates to a surface condition estimation method and a surface condition estimation system.

Generally, when machining a workpiece using a tool of a machine tool, the tool may wear out or deteriorate, resulting in a decrease in machining accuracy. Therefore, it is extremely important to maintain the machining accuracy of the workpiece by managing the deterioration of the tool and correcting the tool as appropriate. For this reason, Patent Documents 1 and 2 disclose techniques for acquiring an image of an evaluation area of a workpiece and estimating the surface roughness of the workpiece machined by the tool.

JP 2006-162558 A JP 2018-40599 A

By the way, when a tool is corrected, frequent corrections can always maintain good machining accuracy for the workpiece. However, when a tool is corrected, time is required to correct the tool, and the cost of correcting the tool is also required. For this reason, appropriately setting the timing of correcting the tool is extremely effective, since it makes it possible to shorten the time required for correction and reduce the cost required for correction. According to the conventional technology described above, it is possible to set the timing of correcting the tool based on the estimated surface roughness, but it is desirable to be able to set the timing of correcting the tool with high accuracy by more accurately estimating the surface condition of the workpiece.

The present invention aims to provide a surface condition estimation method and a surface condition estimation system that can accurately estimate the surface condition of a workpiece.

(Surface condition estimation method)
One aspect of the present invention is a surface condition estimation method that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, comprising :
An evaluation area is set for an image of the entire processing surface;
Dividing the evaluation region into a plurality of areas formed by adjacent pixels among a plurality of pixels forming the image;
Calculating an average luminance value using the luminance values of adjacent pixels in each of the areas ;
calculating a luminance difference value representing a difference between the luminance of each of a plurality of adjacent pixels in each of the areas and the average luminance value;
Calculating the square root of the sum of the squares of the brightness difference values as an index value;
A surface condition estimating method, comprising: using the index value to estimate a surface condition of the machined surface that is correlated with the index value .
Another aspect of the present invention is a surface condition estimation method that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, the method comprising:
Calculating a luminance average value using the luminance of each of a plurality of adjacent pixels among a plurality of pixels forming an image obtained by capturing the processed surface;
calculating a luminance difference value representing a difference between the luminance of each of a plurality of adjacent pixels and the average luminance value;
Calculating the square root of the sum of the squares of the brightness difference values as an index value;
Using the index value, a surface condition of the processed surface that is correlated with the index value is estimated;
The correlation is expressed by a linear approximation equation,
a slope of the linear approximation equation increases as an average luminance of the plurality of pixels forming the image increases, and an intercept of the linear approximation equation decreases as the average luminance increases,
The surface condition of the processed surface is estimated using the index value and the linear approximation equation having the slope and the intercept determined based on the average luminance.

(Surface condition estimation system)
Another aspect of the present invention is a surface condition estimation system that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece , comprising:
an imaging device that images the processing surface and outputs the captured image;
an image evaluation device having: an evaluation area setting unit that sets an evaluation area for the image; an area dividing unit that divides the evaluation area into a plurality of areas formed by a plurality of adjacent pixels among a plurality of pixels that form the image; a brightness average value calculation unit that calculates a brightness average value using the brightness of each of the adjacent pixels in each of the areas ; a brightness difference value calculation unit that calculates a brightness difference value representing a difference between the average brightness value calculated by the brightness average value calculation unit in each of the areas and the brightness of each of the adjacent pixels; and an index value calculation unit that calculates the square root of the sum of the squares of the brightness difference value calculated by the brightness difference value calculation unit as an index value;
an estimation calculation device having an estimation unit that estimates a surface state of the processed surface that is correlated with the index value calculated by the image evaluation device;
The present invention relates to a surface condition estimation system comprising:
Another aspect of the present invention is a surface condition estimation system that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, comprising:
an imaging device that images the processing surface and outputs the captured image;
an image evaluation device having: a luminance average calculation unit that calculates a luminance average value using the luminance of each of a plurality of adjacent pixels among a plurality of pixels forming the image output from the imaging device; a luminance difference value calculation unit that calculates a luminance difference value representing a difference between the luminance average value calculated by the luminance average calculation unit and the luminance of each of the plurality of adjacent pixels; and an index value calculation unit that calculates a square root of the sum of squares of the luminance difference value calculated by the luminance difference value calculation unit as an index value;
an estimation calculation device having an estimation unit that estimates a surface state of the processed surface that is correlated with the index value calculated by the image evaluation device;
Equipped with
The correlation is expressed by a linear approximation equation,
a slope of the linear approximation equation increases as an average luminance of the plurality of pixels forming the image increases, and an intercept of the linear approximation equation decreases as the average luminance increases,
The surface condition of the processed surface is estimated using the index value and the linear approximation equation having the slope and intercept determined based on the average brightness, in a surface condition estimation system.

As a result, the surface condition of the machined surface of the workpiece can be accurately estimated using index values resulting from the brightness of the machined surface of the workpiece. Furthermore, the surface condition estimation system can accurately estimate the condition of the tool of the machine tool that processes the workpiece using the estimated surface condition of the machined surface of the workpiece. This allows the machine tool to maintain good machining accuracy and to modify the tool or change the machining conditions at the appropriate time.

This makes it possible to prevent excessive and delayed correction of the machine tool's tools. This makes it possible to shorten the time required to correct the tools and reduce the cost of correcting the tools.

FIG. 1 is a diagram showing a configuration of a surface state estimation system. FIG. 2 is a diagram showing a configuration of a grinding device which is an example of the machine tool shown in FIG. 1 . Fig. 3 is a diagram showing the configuration of a camera which is an example of the image acquisition device in Fig. 2. Fig. 4 is a front view showing a rack bar which is an example of an article (product). FIG. 2 is a diagram showing a functional block configuration of the image evaluation device shown in FIG. 1 . 5 is a diagram for explaining an entire image captured by the camera of FIG. 3 and an evaluation area set by the image evaluation device of FIG. 4 . FIG. 5 is a diagram for explaining an area set by the image evaluation device of FIG. 4. FIG. 13 is a diagram for explaining the numbers of pixels that make up an area. FIG. 13 is a diagram for explaining a change in luminance in the circumferential direction of a workpiece. FIG. 11 is a diagram for explaining a luminance difference value. FIG. 2 is a diagram showing a functional block configuration of the surface state estimation system. FIG. 13 is a diagram for explaining the correlation between an index value and surface roughness. 13A and 13B are diagrams illustrating the influence of average luminance on a linear approximation equation according to the first modified example. 13A and 13B are diagrams illustrating changes in the slope and intercept of a linear approximation equation with respect to changes in average luminance according to a first modified example. 13A and 13B are diagrams illustrating acquisition of luminance in a first direction according to a second modified example. 13A and 13B are diagrams illustrating how the center position of luminance in the first direction is determined in the second modified example. 13 is a diagram for explaining acquisition of luminance in a second direction according to a second modified example. FIG. 13A and 13B are diagrams illustrating how the center position of luminance in the second direction is determined in the second modification example. 13A and 13B are diagrams illustrating how to determine the center position of an evaluation area according to a second modified example. 13A and 13B are diagrams illustrating setting of an evaluation area when the center position is shifted in the second modified example. 13A and 13B are diagrams illustrating setting of an evaluation area when the center position is shifted according to a second modified example. FIG. 11 is a diagram for explaining setting of an evaluation area according to a second modified example. FIG. 11 is a diagram for explaining setting of an evaluation area according to a second modified example.

(1. Machine tools to which the surface condition estimation system can be applied)
The surface condition estimation system estimates the surface condition of a machined surface of a workpiece machined by a machine tool. Examples of machine tools that machine the workpiece include various cutting machines that perform cutting processing (e.g., gear processing machines, machining centers, etc.) and various grinding machines that perform grinding processing (e.g., cylindrical grinding machines, cam grinding machines, surface grinding machines, etc.).

In this example, the machine tool is exemplified as a grinding device that grinds a workpiece through grinding processes such as rough grinding, fine grinding, fine grinding, and spark-out. Examples of the workpiece ground by the grinding device include a simple shaft-shaped member, a crankshaft, a camshaft, and a flat plate. Note that the workpiece in this example is a simple shaft-shaped member with a circular cross section that is ground by a cylindrical grinding machine, as described below.

(2. Overview of the Configuration of Surface State Estimation System 1)
Next, an overview of the configuration of the surface condition estimating system 1 will be described with reference to Fig. 1. The surface condition estimating system 1 includes at least one grinding device 10, an image evaluation device 20, and at least one computing device (30, 40). The surface condition estimating system 1 may target one grinding device 10, or may target multiple grinding devices as shown in Fig. 1.

The surface condition estimation system 1 in this example is an example that includes multiple grinding devices 10. The surface condition estimation system 1 in this example is an example that includes an image evaluation device 20 provided on each grinding device 10, one calculation device 30, and multiple calculation devices 40 provided on each of the grinding devices 10.

The grinding device 10 grinds the workpiece W through grinding processes such as rough grinding, fine grinding, fine grinding, and spark-out. The grinding device 10 is provided with at least a camera 13 as an imaging device that images the grinding surface W1 that can be observed during the grinding process of the workpiece W.

The camera 13 captures an image of the ground surface W1 of the workpiece W that is ground by the grinding device 10. The camera 13 outputs overall image data that represents an overall image of the entire ground surface W1 of the workpiece W for each grinding process. In this example, the camera 13 is used as an image acquisition device to capture an image of the ground surface W1 of the workpiece W, for example, the outer peripheral surface. However, the image acquisition device is not limited to capturing an image of the ground surface W1 of the workpiece W to obtain an overall image, as with the camera 13, and it is also possible to use, for example, a 3D scanner device that can capture the unevenness of the ground surface W1 as an overall image.

The image evaluation device 20 acquires the entire image (entire image data) output from the camera 13. The image evaluation device 20 then outputs an index value representing the brightness that changes in response to the surface condition of the evaluation area in the entire image (entire image data) for which the surface condition is estimated.

The image evaluation device 20 of this example calculates an area index value for each of a plurality of areas, each of which includes adjacent pixels, for an imaged evaluation area, and calculates an index value for the evaluation area by taking the square root of the sum of the squares of all the area index values. Here, the image evaluation device 20 of this example calculates the area index value as the difference between the luminance of each pixel constituting the area (for example, luminance expressed in 256 levels from 0 to 255) and the average luminance value in the area.

In this example, the arithmetic device (30, 40) applies machine learning using the index value output from the image evaluation device 20 to estimate the surface roughness as the surface condition of the ground surface W1 of the workpiece W, which has a correlation with the index value. The correlation obtained by applying machine learning represents the correlation between the index value obtained from the image evaluation device 20 and the surface roughness (actual roughness value) of the ground surface W1 of the workpiece W actually measured by the roughness meter 2.

In this example, as shown in FIG. 1, the calculation device (30, 40) is composed of a learning processing device 30 and an estimation calculation device 40. Although the learning processing device 30 and the estimation calculation device 40 are shown as independent configurations, they can also be one device. Also, some or all of the calculation devices (30, 40) can be incorporated into a production line that includes multiple grinding devices 10.

In this example, the learning processing device 30 has a so-called server function and is communicatively connected to a plurality of estimation calculation devices 40 provided on each of the plurality of grinding devices 10. In addition, the estimation calculation device 40 is provided one-to-one with each grinding device 10, functions as a so-called edge computer, and enables high-speed calculation processing.

(3. Detailed Configuration of Surface State Estimation System 1)
The configuration of the surface condition estimating system 1 will be described in more detail with reference to Fig. 1. The surface condition estimating system 1 includes a plurality of grinding machines 10 (two in this example), an image evaluation device 20 provided in each of the grinding machines 10, one learning processing device 30 functioning as a part of the calculation device, and an estimation calculation device 40 provided in each of the grinding machines 10 and functioning as another part of the calculation device.

The grinding device 10 mainly comprises a grinding machine 11 that uses a grinding wheel T to grind the grinding surface W1 of the workpiece W, a control device 12 that controls the grinding machine 11, a camera 13 as an image acquisition device, and an interface 14.

In this example, a wheel head traverse type cylindrical grinding machine 50 is used as the grinding machine 11, as shown in FIG. 2. Note that a table traverse type grinding machine can also be used as the grinding machine 11.

The cylindrical grinding machine 50 is a machine tool for grinding the outer peripheral surface, i.e., the grinding surface W1, of the workpiece W, which is a shaft-shaped member. As shown in FIG. 2, the cylindrical grinding machine 50 mainly comprises a bed 51, a headstock 52, a tailstock 53, a traverse base 54, a grinding wheel head 55, a grinding wheel 56 (grinding wheel T), a sizing device 57, and a grinding wheel adjustment device 58.

The bed 51 is fixed on the installation surface. The headstock 52 is provided on the upper surface of the bed 51, at the front side in the X-axis direction (the lower side in FIG. 2) and at one end side in the Z-axis direction (the left side in FIG. 2). The headstock 52 supports the workpiece W so that it can rotate around the Z-axis. The workpiece W is rotated by driving a motor 52a provided on the headstock 52.

The tailstock 53 is provided on the upper surface of the bed 51 at a position facing the headstock 52 in the Z-axis direction, i.e., on the near side in the X-axis direction (the lower side in FIG. 2) and on the other end side in the Z-axis direction (the right side in FIG. 2). As a result, the workpiece W is rotatably supported at both ends by the headstock 52 and the tailstock 53.

The traverse base 54 is provided on the upper surface of the bed 51 so as to be movable in the Z-axis direction. The traverse base 54 is moved by the drive of a motor 54a provided on the bed 51. The grinding wheel head 55 is provided on the upper surface of the traverse base 54 so as to be movable in the X-axis direction. The grinding wheel head 55 is moved by the drive of a motor 55a provided on the traverse base 54. The grinding wheel 56 (tool T) is rotatably supported on the grinding wheel head 55. The grinding wheel 56 is rotated by the drive of a motor 56a provided on the grinding wheel head 55. The grinding wheel 56 is composed of multiple abrasive grains fixed with a bond material.

The sizing device 57 measures the dimensions (diameter) of the workpiece W. The sizing device 57 is provided on the upper surface of the bed 51 so as to be movable in the Z-axis direction. The position of the sizing device 57 in the Z-axis direction is controlled by a feed mechanism 57a provided on the bed 51.

The grinding wheel correction device 58 corrects the shape of the grinding wheel 56. That is, the grinding wheel correction device 58 is a device that performs truing and dressing of the grinding wheel 56. Truing is a reshaping operation, and is an operation of shaping the grinding wheel 56 to match the shape of the workpiece W when the grinding wheel 56 has deteriorated due to grinding, for example, wear or abrasion, the abrasive grains have fallen off, or the abrasive grains have been crushed, and is an operation of removing the runout of the grinding wheel 56 due to uneven wear. Dressing is an operation of dressing (sharpening) the grinding wheel 56, adjusting the amount of protrusion of the abrasive grains, and creating the cutting edge of the abrasive grains. That is, dressing is an operation of correcting dullness, clogging, missing grains, etc., and is usually performed after truing.

The control device 12 includes a CNC device and a PLC device, and is provided in the cylindrical grinding machine 50 (grinding machine 11) in this example, as shown in FIG. 2. The control device 12 controls the motors 52a, 54a, 55a, 56a, etc. in the cylindrical grinding machine 50 according to an NC program generated based on operation command data such as the shape of the workpiece W, the processing conditions, the shape of the grinding wheel 56 (tool T), and coolant supply timing information. That is, the control device 12 generates an NC program based on the operation command data by inputting the operation command data.

As a result, the control device 12 controls the operation of the cylindrical grinding machine 50 (grinding machine 11) to perform grinding of the ground surface W1 of the workpiece W. Here, the control device 12 performs grinding until the ground surface W1 of the workpiece W has a finished shape based on the dimensions (diameter) of the workpiece W measured by the sizing device 57. For this reason, the NC program generated by the control device 12 is generated according to the grinding content, i.e., rough grinding, fine grinding, fine grinding, spark out, etc.

The control device 12 is provided on the bed 51 and is capable of communicating with the estimation calculation device 40 via the interface 14 (see FIG. 1). As a result, the control device 12 of this example controls the motors 52a, 54a, 55a, 56a and the grinding wheel correction device 58, etc., at the timing when the correction signal R output from the estimation calculation device 40 is acquired as the timing for correcting the grinding wheel 56, thereby correcting (truing and/or dressing) the grinding wheel 56. Here, the estimation calculation device 40 of this example outputs the correction signal R to the control device 12 based on the estimated surface condition, i.e., surface roughness, of the ground surface W1, as described later.

As shown in FIG. 2, the camera 13 is mounted on the traverse base 54, and captures an image of the ground surface W1 of the workpiece W while the ground surface W1 is irradiated with imaging light. For this reason, as shown in FIG. 3, the camera 13 is provided with, for example, an illumination device 13a arranged concentrically with the lens of the camera 13 (not shown) and irradiating annular imaging light. The camera 13 then outputs overall image data Z representing the overall image of the ground surface W1 reflecting the imaging light to the image evaluation device 20, for example, via the interface 14 of the control device 12 (see FIG. 1).

As shown in FIG. 4, the image evaluation device 20 includes an evaluation area setting unit 21, an area dividing unit 22, a brightness acquisition unit 23, and an index value calculation unit 24. As shown in FIG. 5, the image evaluation device 20 acquires overall image data Z representing an overall image of the ground surface W1 captured by the camera 13. Here, as described above, the camera 13 irradiates annular imaging light. Therefore, in the overall image data Z capturing the ground surface W1 of the workpiece W, there are a first portion (shown in white in FIG. 5) that strongly reflects light, and a second portion (shown in matte in FIG. 5) that weakly reflects light. Note that in this example, a case is illustrated in which there are two first portions that strongly reflect light in the overall image data Z.

The evaluation area setting unit 21 acquires the overall image data Z output from the camera 13, and sets an evaluation area HR for evaluating the surface condition, i.e., surface roughness, of the ground surface W1 represented by the overall image data Z. In this example, as shown by the thick solid line in Figure 5, the evaluation area setting unit 21 sets the evaluation area HR so that it includes two first portions.

As shown in Fig. 6, the area dividing unit 22 sets four adjacent pixels (indicated by small squares in Fig. 6) constituting the evaluation area HR as area A, and divides the entire evaluation area HR into n areas A. Here, in this example, area A has four pixels, but the number of pixels constituting area A is not limited to this. Also, in this example, as shown in Fig. 6, a case is illustrated in which the evaluation area HR is composed of ^2N x ^2M pixels.

In this case, each pixel is assigned a number for identification purposes, and for example, as shown in Fig. 6, the upper left pixel of the evaluation area HR is assigned "0", the pixel adjacent to the right of pixel "0" is assigned "1", and the last pixel in the first column of the evaluation area HR is assigned " ^2N -1". Similarly, the pixel immediately below and adjacent to pixel "0" in the evaluation area HR is assigned " ^2N ", the pixel immediately below and adjacent to pixel " ^2N " is assigned " ^2N +1", and the last pixel in the second column of the evaluation area HR is assigned " ^2x2N- 1". Similarly, the pixel immediately below and adjacent to pixel " ^2N " in the evaluation area HR is assigned " ^2x2N ", the first pixel in the Mth column of the evaluation area HR is assigned "( ^2M -1) ^x2N ", and the last pixel in the Mth column is assigned "2 ^(M+N) -1".

As a result, the numbers assigned to the four pixels constituting the m-th area Am shown by the matte finish in Fig. 6 are as follows: "2i+2j× ^2N " is assigned to the first pixel P1 located at the upper left of the area Am, and "2i+2j× ^2N+ 1" is assigned to the second pixel P2 adjacent to the first pixel P1 on the right, as shown in Fig. 7. Also, in the area Am, "2i+ ^2N +2j× ^2N " is assigned to the third pixel P3 adjacent to the first pixel P1 directly below, and "2i+ ^2N +2j× ^2N+ 1" is assigned to the fourth pixel P4 adjacent to the third pixel P3 on the right (i.e., adjacent to the second pixel P2 directly below). Here, "i" is a value that varies from "0" to " ^2N-1-1 ", and "j" is a value that varies from "0" to " ^2M-1-1 ".

Then, the area dividing unit 22 outputs to the brightness acquiring unit 23 brightness data SP(n) including the respective brightness φ of the four first pixels P1, second pixels P2, third pixels P3 and fourth pixels P4 constituting each of the n divided areas A, that is, the brightness φ ₁ of the first pixel P1, the brightness φ ₂ of the second pixel P2, the brightness φ ₃ of the third pixel P3 and the brightness φ ₄ of the fourth pixel P4 constituting the area A. Here, in this example, the brightness data SP(n) is data expressing each brightness φ in 256 stages, for example, from 0 to 255.

The closer the brightness φ is to "0", the lower the brightness (the darker it is), and the closer it is to "255", the higher the brightness (the brighter it is). For this reason, if the brightness φ of each pixel is arranged in numerical order for the entire evaluation area HR, as shown in Figure 8, for example, as shown by the arrow in Figure 5, the brightness φ changes along the circumferential direction of the ground surface W1 of the workpiece W. In particular, for pixels corresponding to the first portion of the overall image data Z, the brightness φ is constant at "255". In other words, the pixels corresponding to the first portion will be what is known as "blown out highlights."

Here, the distribution of the luminance φ in the evaluation region HR changes depending on the surface condition of the ground surface W1, specifically, the surface roughness. In other words, when light irradiated from the lighting device 13a hits the ground surface W1, specular reflection light and scattered reflection light are generated according to the microscopic inclination angle of the surface, but as the surface roughness increases, microscopically steep changes occur, resulting in the occurrence of areas with large differences in brightness. As a result, the change in the distribution of the luminance φ tends to increase as the surface roughness (actual roughness value JS) of the ground surface W1 becomes rougher (larger), and to decrease as the surface roughness (actual roughness value JS) becomes smoother (smaller).

The luminance acquisition unit 23 acquires luminance data SP(n) for each of the n areas output from the area division unit 22. Here, the luminance data SP(n) acquired by the luminance acquisition unit 23 is data expressed as SP(n)=(φ1, _φ2 , _φ3 , φ4 ₎ using the luminance _φ1 of the first pixel P1, the luminance _φ2 of the second pixel P2, the luminance _φ3 of the third pixel P3, and the luminance _φ4 of the fourth pixel _P4 . As shown in FIG. 4, the luminance acquisition unit 23 includes a luminance average value calculation unit 231 and a luminance difference value calculation unit 232.

The brightness average value calculation unit 231 calculates the brightness average value φ a (n) using the brightnesses (φ ₁ , φ ₂ , φ ₃ , φ ₄ ) of the first pixel P1, the second pixel P2, the third pixel P3, and the fourth pixel P4 that constitute the area A. That is, the brightness average value calculation unit 231 calculates _{the brightness average value φ a (} n) according to the following formula 1.

The luminance difference value calculation unit 232 calculates a luminance difference value Δφ(n) between the luminance (φ ₁ , φ ₂ , φ ₃ , φ ₄ ) of each of the first pixel P1, second pixel P2, third pixel P3, and fourth pixel P4 constituting the area A and the average luminance value φ _a (n) calculated by the average luminance calculation unit 231. That is, the luminance difference value calculation unit 232 calculates the luminance difference value Δφ(n) in accordance with the following formula 2-5. Here, the brightness difference value Δφ(n) is expressed as _Δφ (n) = ( _Δφ1 , Δφ2, Δφ3, Δφ4) using the brightness difference value _Δφ1 between _φa (n) and the luminance _φ1 of the first pixel P1, the brightness difference value _Δφ2 between _φa (n) and the luminance _φ2 of _the second pixel _P2 , the brightness difference value _Δφ3 between _φa (n) and the _{luminance φ3} of the third pixel P3, and the brightness difference value Δφ4 between _φa (n) and the luminance φ4 of the fourth pixel _P4 .

Here, as for the brightness difference value Δφ(n) calculated by the brightness difference value calculation unit 232, as shown in FIG. 9 corresponding to FIG. 8, the brightness difference value Δφ(n) of the blown-out pixel corresponding to the first portion of the ground surface W1 is "0". In other words, the state in which a pixel is blown-out is when it has a brightness that exceeds the upper brightness limit (e.g., "255"), and does not accurately reflect the surface condition (surface roughness) of the ground surface W1 of the workpiece W. Therefore, by calculating the brightness difference value Δφ(n), it is possible to eliminate the influence of the blown-out pixel on the estimation of the surface condition (surface roughness), and as a result, it becomes possible to accurately estimate the surface condition (surface roughness) of the ground surface W1 described later.

そして、輝度差分値算出部２３２は、ｎ個のエリアＡの各々について輝度合算値Δφｇ（ｎ）を算出すると、算出した輝度差分二乗総和値Δφｇ（ｎ）（又は、輝度差分値Δφ（ｎ））を指標値算出部２４に出力する。 In this example, when the brightness difference value calculation unit 232 calculates the brightness difference values Δφ(n)=( _Δφ1 , _Δφ2 , _Δφ3 , _Δφ4 ), it can calculate a brightness difference squared sum value Δφg(n) by squaring and adding up each of the brightness difference values ( _Δφ1 , _Δφ2 , _Δφ3 , _Δφ4 ) for area A. That is, the brightness difference value calculation unit 232 can calculate the brightness difference squared sum value Δφg(n) for each area A according to the following formula 6.

Then, the brightness difference value calculation unit 232 calculates the brightness sum value Δφg(n) for each of the n areas A, and outputs the calculated brightness difference squared sum value Δφg(n) (or the brightness difference value Δφ(n)) to the index value calculation unit 24.

尚、前記式７における右辺の分母「４ｎ」は、本例のエリアＡが２×２のピクセルを有するためである。例えば、エリアＡが４×４のピクセルを有する場合には、前記式７における右辺の分母は「１６ｎ」になる。そして、指標値算出部２４は、指標値ＫＩを算出すると、学習処理装置３０及び推定演算装置４０に指標値ＫＩを出力する。 The index value calculation unit 24 calculates the index value KI by calculating the square sum average value using the luminance difference square sum total value Δφg(n) (or the luminance difference value Δφ(n)) calculated by the luminance difference value calculation unit 232 for all of the n areas A. That is, the index value calculation unit 24 calculates the square sum average value of the luminance difference square sum total value Δφg(n) (or the luminance difference value Δφ(n)) of all areas A as the index value KI in accordance with the following formula 7.

The denominator "4n" on the right side of the formula 7 is because the area A in this example has 2 × 2 pixels. For example, if the area A has 4 × 4 pixels, the denominator on the right side of the formula 7 becomes "16n." After calculating the index value KI, the index value calculation unit 24 outputs the index value KI to the learning processing device 30 and the estimation calculation device 40.

As shown in FIG. 1, the learning processing device 30 is configured with a processor 31, a storage device 32, an interface 33, etc. The learning processing device 30 also has a server function and is communicatively connected to a plurality of grinding devices 10, i.e. grinding machines 11, each of which is provided with an image evaluation device 20 on a cylindrical grinding machine 50 (or includes an image evaluation device 20 provided on a cylindrical grinding machine 50 operating in another remote factory).

The learning processing device 30 performs machine learning based on the index value KI output from the image evaluation device 20 and the actual roughness value JS corresponding to the surface condition of the ground surface W1 of the workpiece W output from the roughness meter 2 connected via the interface 33. The learning processing device 30 then generates a learned model for estimating the surface condition, i.e., the surface roughness, of the ground surface W1 of the workpiece W.

As shown in FIG. 1, each estimation calculation device 40 is configured to include a processor 41, a storage device 42, an interface 43, etc. In addition, the estimation calculation device 40 is connected to the learning processing device 30 as a server and the corresponding image evaluation device 20 so as to be able to communicate with each other.

The estimation calculation device 40 is installed so as to be able to communicate with the control device 12 of each grinding device 10, and functions as an edge computer. Using the learned model generated by the learning processing device 30, the estimation calculation device 40 estimates the surface condition, i.e., surface roughness, of the ground surface W1 of the workpiece W based on the index value KI acquired as a feature value from the image evaluation device 20 during (or after) grinding of the workpiece W.

(4. Functional Block Configuration of Surface State Estimation System 1)
The functional blocks of the surface condition estimating system 1 will be described with reference to Fig. 10. As described above, the surface condition estimating system 1 includes the camera provided in the grinding device 10, the image evaluation device 20, the learning processing device 30, and the estimation calculation device 40.

The learning processing device 30 generates a learned model for estimating the surface roughness of the ground surface W1 based on the index value KI output from the image evaluation device 20 and the actual roughness value JS representing the surface roughness of the ground surface W1 measured by the roughness meter 2 (see FIG. 1). The learning processing device 30 includes a training data set acquisition unit 61, a training data set storage unit 62, and a model generation unit 63.

The training dataset acquisition unit 61 acquires a training dataset for performing machine learning. The training dataset acquisition unit 61 includes an index value acquisition unit 61a and an actual roughness value acquisition unit 61b. The index value acquisition unit 61a acquires the index value KI from the image evaluation device 20 as a training dataset. The actual roughness value acquisition unit 61b acquires the actual roughness value JS from the roughness meter 2 as a training dataset.

The training dataset storage unit 62 associates the index value KI and the actual roughness value JS acquired by the training dataset acquisition unit 61 and stores them as a training dataset. That is, the training dataset storage unit 62 associates and stores the index value KI, which varies depending on the surface condition, i.e., the surface roughness, of the ground surface W1 of the workpiece W, with the actually measured surface roughness (actual roughness value) of the ground surface W1.

The model generation unit 63 performs machine learning using the training dataset stored in the training dataset storage unit 62. Specifically, the model generation unit 63 performs machine learning using the index value KI stored in the training dataset storage unit 62 in association with each other as explanatory variables and the surface roughness (actual roughness value) as the objective variable. The model generation unit 63 then generates a trained model that represents the correlation between the index value KI and the surface roughness JS (actual roughness value).

Here, an example of a trained model generated by the model generation unit 63 is shown. When performing machine learning, the model generation unit 63 uses the index value KI and surface roughness JS for the ground surface W1 of multiple workpieces W. The surface roughness JS is measured at multiple grinding points (e.g., seven points) on the ground surface W1 of one workpiece W. In this case, as shown in FIG. 11, for example, the index value KI and surface roughness JS corresponding to multiple grinding points can be plotted for each of eight workpieces W, i.e., workpieces Wa, Wb, Wc, Wd, We, Wf, Wg, and Wh. Regarding the grinding timing of the workpieces W, the workpiece Wa is ground first, and the workpiece Wh is ground last.

According to this, the workpiece Wa (shown by a white circle in FIG. 11) that is ground first has a small index value KI and surface roughness JS, while the workpiece Wh (shown by a black diamond in FIG. 11) that is ground last has a large index value KI and surface roughness JS. That is, the index value KI of workpiece Wb (shown by a black circle in FIG. 11) tends to be larger than that of workpiece Wa, and the index value KI of workpiece Wc (shown by a white triangle in FIG. 11) tends to be larger than that of workpiece Wb. Similarly, with regard to the index value KI, workpiece Wd (shown by a black triangle in FIG. 11) tends to be larger than workpiece Wc, workpiece We (shown by a white square in FIG. 11) tends to be larger than workpiece Wd, workpiece Wf (shown by a black square in FIG. 11) tends to be larger than workpiece We, and workpiece Wg (shown by a white diamond in FIG. 11) tends to be larger than workpiece Wf and smaller than workpiece Wh.

Furthermore, with respect to surface roughness JS, workpiece Wb tends to be greater than workpiece Wa, and workpiece Wc tends to be greater than workpiece Wb. Similarly, with respect to surface roughness JS, workpiece Wd tends to be greater than workpiece Wc, workpiece We tends to be greater than workpiece Wd, workpiece Wf tends to be greater than workpiece We, and workpiece Wg tends to be greater than workpiece Wf and smaller than workpiece Wh.

When the average values (shown by double circles in FIG. 11) of each workpiece W are plotted to show the tendency of change in these index values KI and surface roughness JS, the average values can be approximated by a linear approximation formula, as shown by the thick solid line in FIG. 11. That is, in this example, the correlation between the index value KI and surface roughness JS can be expressed by a linear approximation formula. Therefore, the model generation unit 63 in this example performs machine learning to generate a linear approximation formula that shows the correlation between the index value KI and surface roughness JS as a learned model.

The estimation calculation device 40 estimates the surface roughness SS of the ground surface W1 of the workpiece W using the trained model generated by the learning processing device 30 and the index value KI acquired from the image evaluation device 20, and determines whether or not the grinding wheel 56 (tool T) of the cylindrical grinding machine 50 needs to be modified based on the estimated surface roughness SS. If the grinding wheel 56 (tool T) needs to be modified, the estimation calculation device 40 outputs a modification signal R to the control device 12. For this reason, the estimation calculation device 40 includes a model storage unit 71, an index value acquisition unit 72, a surface roughness estimation unit 73, a determination unit 74, and an output unit 75, as shown in FIG. 10.

The model storage unit 71 stores the trained model (e.g., a linear approximation equation representing the correlation between the index value KI and the surface roughness JS) generated by the model generation unit 63. The index value acquisition unit 72 acquires the index value KI from the image evaluation device 20. Here, the index value acquisition unit 72 performs the same processing as the index value acquisition unit 61a of the training dataset acquisition unit 61.

In this example, the index value acquisition unit 72 is described as a separate element from the index value acquisition unit 61a of the training data set acquisition unit 61. However, it is also possible for the index value acquisition unit 61a of the training data set acquisition unit 61 to serve as the index value acquisition unit 72. In other words, the function of the element 61a in the learning processing device 30 is shared with part of the function of the estimation calculation device 40.

The surface roughness estimation unit 73 acquires the trained model stored in the model storage unit 71. The surface roughness estimation unit 73 also acquires the index value KI acquired by the index value acquisition unit 72. As a result, the surface roughness estimation unit 73 estimates the surface roughness SS of the ground surface W1 of the workpiece W based on the trained model and the index value KI.

The determination unit 74 determines whether the surface roughness SS of the ground surface W1 estimated by the surface roughness estimation unit 73 is equal to or greater than the reference surface roughness SR preset for the ground surface W1 for each grinding process (i.e., rough grinding, fine grinding, fine grinding, and spark out). If the surface roughness SS is equal to or greater than the reference surface roughness SR, the determination unit 74 determines that the grinding wheel 56 (tool T) of the cylindrical grinding machine 50 (grinding machine 11) has deteriorated and the ground surface W1 cannot be properly ground, i.e., correction is required. On the other hand, if the surface roughness SS is less than the reference surface roughness SR, the determination unit 74 determines that the grinding wheel 56 (tool T) of the cylindrical grinding machine 50 (grinding machine 11) has not deteriorated and the ground surface W1 can be properly ground, i.e., correction is not required.

The output unit 75 outputs a correction signal R to the control device 12 according to the result of the judgment by the judgment unit 74. That is, when the judgment unit 74 judges that the grinding wheel 56 (tool T) needs to be corrected, the output unit 75 outputs a correction signal R to the control device 12. Also, when the judgment unit 74 judges that the grinding wheel 56 (tool T) does not need to be corrected, the output unit 75 does not output a correction signal R to the control device 12.

In this way, in the control device 12 of the cylindrical grinding machine 50 (grinding machine 11), when the grinding wheel 56 (tool T) deteriorates, i.e., when the grinding wheel 56 (tool T) needs to be corrected, the control device 12 acquires a correction signal R from the output section 75 of the estimation calculation device 40 via the interface 14. Then, the control device 12 controls the grinding wheel correction device 58 to correct the grinding wheel 56. That is, the grinding wheel correction device 58 performs truing and/or dressing on the grinding wheel 56. This allows the grinding wheel 56 to be corrected at an appropriate time and an appropriate number of times. Therefore, the time required for correction can be shortened, and the cost required for correction can also be reduced.

As can be understood from the above description, the surface condition estimation system 1 of this example can accurately estimate the surface roughness SS, which is the surface condition of the ground surface W1 of the workpiece W, using the index value KI resulting from the brightness of the ground surface W1 of the workpiece W. The surface condition estimation system 1 can then use the estimated surface roughness SS to determine the degree of deterioration of the grinding wheel 56 (tool T) of the cylindrical grinding machine 50 (grinding machine 11), which is the grinding device 10 that grinds the ground surface W1, and output a correction signal R. As a result, the control device 12 of the grinding device 10 can control the grinding wheel correction device 58 at the timing when the correction signal R is acquired, and correct the grinding wheel 56 by truing and/or dressing.

This makes it possible to prevent excessive correction or delays in correction of the grinding wheel 56 (tool T) of the cylindrical grinding machine 50 (grinding machine 11). This makes it possible to shorten the correction time required to correct the grinding wheel 56 and reduce the cost of correcting the grinding wheel 56.

(5. First Alternative Example)
In the above-mentioned embodiment, the estimation calculation device 40 estimates the surface condition, i.e., the surface roughness SS, of the ground surface W1 of the workpiece W using the trained model generated by the learning processing device 30 and the index value KI acquired from the image evaluation device 20. As described above, the index value KI is a value related to the brightness φ that changes due to the surface roughness of the ground surface W1. For this reason, when estimating the surface roughness SS using the index value KI, it is necessary to make the average brightness uniform over the entire ground surface W1, i.e., the entire evaluation region HR. In other words, if the average brightness is the same or the effect of changes in the average brightness can be eliminated, it is possible to convert the index value KI to the surface roughness SS.

As shown in FIG. 12, for example, the learned model used by the estimation calculation device 40 in the above-described example, i.e., the linear approximation equation expressing the correlation (correlation function) between the index value KI and the surface roughness SS, has a tendency that as the average luminance increases (becomes brighter), the slope value increases while the intercept value decreases. Conversely, the linear approximation equation has a tendency that as the average luminance decreases (becomes darker), the slope value decreases while the intercept value increases.

That is, the linear approximation relationship of the slope k(φi) of the linear approximation equation to the change in the average luminance φi in the ground surface W1 (evaluation area HR) and the linear approximation relationship of the intercept b(φi) of the linear approximation equation to the change in the average luminance φi are as shown in Fig. 13. In Fig. 13, the open circles represent the slopes of the linear approximation equation (i.e., the square sum average value line), and the open triangles represent the intercepts of the linear approximation equation (i.e., the square sum average value line). According to these linear approximation relationships, for example, by acquiring the average luminance φi (= the sum of the luminances φ of 2 ^(N+M) pixels divided by the number of pixels ^2N x ^2M ) in the evaluation area HR from the image evaluation device 20, the slope k(φi) and intercept b(φi) corresponding to the average luminance φi can be determined.

Therefore, when the estimation calculation device 40 calculates and estimates the surface roughness SS using the index value KI and a linear approximation formula (learned model), KI = k(φi) × SS + b(φi) holds, so the surface roughness SS can be calculated and estimated according to SS = (KI - B(φi)) / k(φi) using the slope k(φi) and intercept b(φi) according to the average luminance φi. That is, in this case, normalization is performed by using the slope k(φi) and intercept b(φi) according to the average luminance φi, and the index value KI can be directly converted to the surface roughness SS.

Therefore, in the first variant, it is possible to more easily estimate the surface roughness SS by calculation from the index value KI. Furthermore, in the first variant, normalization can be used to eliminate the effect of differences in the average brightness φi on the ground surface W1 (evaluation region HR), allowing the surface roughness SS to be calculated and estimated more accurately.

(6. Second Alternative Example)
In the above-mentioned present embodiment and the first modified embodiment, the image evaluation device 20 sets the evaluation area HR so as to include, for example, the first and second portions present on the ground surface W1. Meanwhile, the grinding device 10 performs grinding on the ground surfaces W1 of a plurality of workpieces W that are sequentially supplied. In this case, the state of the ground surface W1 of each workpiece W changes due to, for example, a machining position error or a machining time difference between the first half and the second half of a grinding lot. In this case, if the evaluation area HR is set based on the same position set in advance, a deviation occurs in the evaluation area HR set for each workpiece W, and as a result, there is a possibility that the estimation accuracy of the surface roughness SS will decrease.

Therefore, in the second modified example, the evaluation area setting unit 21 of the image evaluation device 20 is able to set the evaluation area HR between different workpieces W without causing misalignment. Specifically, the evaluation area setting unit 21 acquires the brightness φ of each pixel in the overall image data Z output from the camera 13, for example, in a first direction corresponding to the circumferential direction of the workpiece W and a second direction corresponding to the axial direction of the workpiece W, i.e., perpendicular to the first direction. Then, the center position of the overall image data Z in the first direction is identified based on the change in brightness φ acquired along the first direction, and the center position of the overall image data Z in the second direction is identified based on the change in brightness φ acquired along the second direction.

Then, the intersection of a line passing through the center position in the first direction and a line passing through the center position in the second direction is set as the center of the evaluation area HR, and an evaluation area HR of a preset size is set. This makes it possible to appropriately set the evaluation area HR even for different workpieces W. This second example is described in detail below.

The evaluation area setting unit 21 of the second example acquires the overall image data Z output from the camera 13. As shown by the arrow in FIG. 14, the evaluation area setting unit 21 acquires the brightness φ of multiple pixels along a first direction corresponding to the circumferential direction of the workpiece W (ground surface W1) in the overall image data Z, and averages the multiple acquired brightness φ to calculate the brightness sum average value φsa1. The evaluation area setting unit 21 then calculates the brightness sum average value φsa1 in the first direction for only multiple columns along the axial direction (i.e., the second direction) of the workpiece W (ground surface W1). This makes it possible to acquire the change in the brightness sum average value φsa1 in the first direction, as shown in FIG. 15.

In the change in the acquired average luminance sum value φsa1, the evaluation area setting unit 21 sets the point at which the average luminance sum value φsa1 exceeds a preset threshold value L1 (shown by a dashed line in FIG. 15) from a small value to a large value as a lower limit point Q11, and sets the point at which the average luminance sum value φsa1 exceeds the threshold value L1 from a large value to a small value as an upper limit point Q12. Then, the evaluation area setting unit 21 sets the median value between the lower limit point Q11 and the upper limit point Q12 as the central position Q13 in the first direction, as shown by a dashed line in FIG. 15.

The evaluation area setting unit 21 also acquires the brightness φ of multiple pixels in the overall image data Z along a second direction corresponding to the axial direction of the workpiece W (ground surface W1) as shown by the arrow in FIG. 16, and calculates the average brightness summation value φsa2 by averaging the multiple acquired brightnesses φ. The evaluation area setting unit 21 then calculates the average brightness summation value φsa2 in the second direction for only multiple rows along the circumferential direction (i.e., the first direction) of the workpiece W (ground surface W1). This makes it possible to acquire the change in the average brightness summation value φsa2 in the second direction, as shown in FIG. 17.

In the change in the acquired average luminance sum value φsa2, the evaluation area setting unit 21 sets the point at which the average luminance sum value φsa2 exceeds a preset threshold value L2 (shown by a dashed line in FIG. 17) from a small value to a large value as a lower limit point Q21, and sets the point at which the average luminance sum value φsa2 exceeds the threshold value L2 from a large value to a small value as an upper limit point Q22. Then, the evaluation area setting unit 21 sets the median value between the lower limit point Q21 and the upper limit point Q22 as the median position Q23 in the second direction, as shown by a dashed line in FIG. 17.

Next, as shown in Figure 18, the evaluation area setting unit 21 sets the intersection point of a line passing through the central position Q13 in the first direction and a line passing through the central position Q23 in the second direction as the center RC of the evaluation area HR. Then, the evaluation area setting unit 21 finally determines the evaluation area HR using points RP1 and RP2 (in Figure 18, point RP1 corresponding to the upper left of the evaluation area HR and point RP2 corresponding to the lower right of the evaluation area HR) that have been set in advance to determine the evaluation area HR.

As a result, even if the grinding surface W1 is displaced from the center of the image in the overall image data Z as shown in FIG. 19 (FIG. 19 shows an example in which the grinding surface W1 is displaced to the left), the evaluation area setting unit 21 can appropriately determine the evaluation area HR in response to this displacement. Similarly, even if the apex of the grinding surface W1 in the circumferential direction is displaced from the center of the image in the overall image data Z as shown in FIG. 20 (FIG. 20 shows an example in which the apex of the grinding surface W1 is displaced upward), the evaluation area setting unit 21 can appropriately determine the evaluation area HR in response to this displacement.

21, even if the entire image data Z is the ground surface W1, the evaluation area setting unit 21 can determine the evaluation area HR by determining the lower limit point Q11 and the upper limit point Q12 as in the above-mentioned case, and setting the intersection point of a straight line passing through the center position Q13 in the first direction and a straight line passing through the center position Q13 in the second direction as the center RC of the evaluation area HR. Also, as shown in FIG. 22, even if there are two first portions as in the above-mentioned example, the evaluation area setting unit 21 can determine the evaluation area HR by determining the lower limit point Q11 and the upper limit point Q12 as in the above-mentioned case, and setting the intersection point of a straight line passing through the center position Q13 in the first direction and a straight line passing through the center position Q13 in the second direction as the center RC of the evaluation area HR.

(7. Other Examples)
In the present example, the first modified example, and the second modified example described above, the learning processing device 30 performs machine learning to generate a learned model, i.e., a correlation between the index value KI and the surface roughness SS. As a result, the estimation calculation device 40 uses the learned model generated by the learning processing device 30 to estimate the surface roughness SS of the ground surface W1 of the workpiece W.

However, the surface condition estimation system 1 does not necessarily have to include a learning processing device 30 that performs machine learning. In this case, the estimation calculation device 40 can perform, for example, a multiple regression analysis using the index value KI obtained from the image evaluation device 20 and the surface roughness JS of the ground surface W1 obtained in advance from the roughness meter 2, and determine the correlation used to estimate the surface roughness SS. In this case, the estimation calculation device 40 can accurately estimate the surface roughness SS using the index value KI obtained from the image evaluation device 20 based on the correlation obtained by the multiple regression analysis. Therefore, even in this case, the same effects as those of the present example, the first variant, and the second variant described above can be obtained.

In the present example, the first variant example, and the second variant example described above, the surface condition estimation system 1 is configured so that the grinding device 10 is equipped with a cylindrical grinding machine 50 as the grinding machine 11. The surface condition estimation system 1 is configured to estimate the surface condition, i.e., the surface roughness SS, of the ground surface W1 of the workpiece W, which is a shaft-shaped member ground by the cylindrical grinding machine 50.

However, the surface condition estimation system 1 is not limited to estimating the surface roughness SS of the ground surface W1 ground by the cylindrical grinding machine 50. For example, the grinding device 10 may be equipped with a cam lathe or a surface grinding machine as the grinding machine 11, and the surface condition estimation system 1 may estimate the surface condition, i.e., the surface roughness SS, of the ground cam surface or flat surface.

Furthermore, the surface condition estimation system 1 is not limited to estimating the surface condition of the ground surface W1, but can also estimate, for example, the surface condition of the cut surface of a workpiece (e.g., a gear) that has been cut by a cutting device. In this case, as in the present example, the first variant, and the second variant described above, the surface condition of the cut surface, for example, the surface roughness, can be estimated using the index value KI obtained from the image evaluation device 20. Therefore, in this case, too, the same effects as in the present example, the first variant, and the second variant described above can be obtained.

In addition, in the present example, the first modified example, and the second modified example described above, the estimation calculation device 40 outputs a correction signal R for correcting the tool T based on the estimated surface roughness SS. However, the estimation calculation device 40 can also output a signal for appropriately changing the machining conditions under which the machine tool uses the tool T to machine the workpiece W based on the estimated surface roughness SS. In this way, the machine tool can acquire the output signal, appropriately change the machining conditions, and accurately machine the workpiece W.

Furthermore, in the present example, the first variant example, and the second variant example described above, the surface roughness SS is estimated as the surface condition of the ground surface W1. However, the surface condition that can be estimated is not limited to the surface roughness SS. The surface condition estimation system 1 can accurately estimate the shape of the machined surface as the surface condition, for example, the shape (outer line, inner line, steps, etc.) of the machined surface formed by the tool T coming into contact with the workpiece W, using the index value KI obtained from the image evaluation device 20.

1...Surface condition estimation system, 2...Roughness meter, 10...Grinding device, 11...Grinding machine, 12...Control device, 13...Camera, 13a...Illumination device, 14...Interface, 20...Image evaluation device, 21...Evaluation area setting unit, 22...Area division unit, 23...Brightness acquisition unit, 231...Brightness average value calculation unit, 232...Brightness difference value calculation unit, 24...Index value calculation unit, 30...Learning processing device (arithmetic device), 31...Processor, 32...Storage device, 33...Interface, 40... Estimation calculation device (calculation device), 41... Processor, 42... Storage device, 43... Interface, 50... Cylindrical grinding machine, 51... Bed, 52... Headstock, 52a... Motor, 53... Tailstock, 54... Traverse base, 54a... Motor, 55... Grinding wheel head, 55a... Motor, 56... Grinding wheel, 56a... Motor, 57... Sizing device, 57a... Feed mechanism, 58... Grinding wheel correction device, 61... Training data set acquisition unit, 61a... Index value acquisition unit, 61b... Actual roughness value acquisition unit, 62...training data set storage unit, 63...model generation unit, 71...model storage unit, 72...index value acquisition unit, 73...estimation unit, 74...determination unit, 75...output unit, A...area, Am...area, HR...evaluation area, JS...actual roughness value, KI...index value, P1...first pixel, P2...second pixel, P3...third pixel, P4...fourth pixel, L1...threshold, L2...threshold, Q11...lower limit, Q12...upper limit, Q13...center position, Q21 ...Lower limit, Q22...Upper limit, Q23...Center position, R...Correction signal, RC...Center, RP1, RP2...Point, SP...Brightness data, SS...(Estimated) surface roughness (surface condition), T...Tool, W...Workpiece, W1...Grinding surface, Z...Overall image data, b...Intercept, k...Slope, Δφ(n)...Brightness difference value, Δφg(n)...Brightness difference squared sum value, φ...Brightness, φa...Brightness average value, φi...Average brightness, φsa1...Brightness sum average value, φsa2...Brightness sum average value

Claims (14)

1. A surface condition estimation method that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, comprising:
An evaluation area is set for an image of the entire processing surface;
Dividing the evaluation region into a plurality of areas formed by adjacent pixels among a plurality of pixels forming the image;
Calculating an average luminance value using the luminance values of adjacent pixels in each of the areas ;
calculating a luminance difference value representing a difference between the luminance of each of a plurality of adjacent pixels in each of the areas and the average luminance value;
Calculating the square root of the sum of the squares of the brightness difference values as an index value;
A surface condition estimating method, comprising: using the index value to estimate a surface condition of the machined surface that is correlated with the index value. determining a first central position in a first direction of the image based on a variation in luminance of a plurality of pixels forming the image in the first direction of the image;
determining a second center position in a second direction different from the first direction based on the change in luminance of the pixel in the second direction;
determining a center of the evaluation area based on the first center position and the second center position;
The surface condition estimating method according to claim 1 , further comprising: setting the evaluation area based on the center. When the first direction and the second direction are perpendicular to each other,
The surface condition estimating method according to claim 2 , wherein the center is an intersection point between a line perpendicular to the first direction and passing through the first central position, and a line perpendicular to the second direction and passing through the second central position. The first central position and the second central position are
4. The surface condition estimating method according to claim 2, wherein the change in luminance is determined based on a median between a lower limit point where the luminance changes from a lower luminance to a higher luminance relative to a preset luminance threshold value and an upper limit point where the luminance changes from a higher luminance to a lower luminance relative to the threshold value. the correlation is a trained model generated by performing machine learning using a training data set including the index value as an explanatory variable, the surface state of the processed surface as an objective variable, and the explanatory variable and the objective variable;
The surface condition estimating method according to any one of claims 1 to 4 , wherein the surface condition of the processed surface is estimated using the index value and the trained model. The correlation is expressed by a linear approximation equation,
a slope of the linear approximation equation increases as an average luminance of the plurality of pixels forming the image increases , and an intercept of the linear approximation equation decreases as the average luminance increases,
The surface condition estimating method according to any one of claims 1 to 4 , wherein the surface condition of the processed surface is estimated using the index value and the linear approximation equation having the slope and the intercept determined based on the average luminance. 1. A surface condition estimation method that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, comprising:
Calculating a luminance average value using the luminance of each of a plurality of adjacent pixels among a plurality of pixels forming an image obtained by capturing the processed surface;
calculating a luminance difference value representing a difference between the luminance of each of a plurality of adjacent pixels and the average luminance value;
Calculating the square root of the sum of the squares of the brightness difference values as an index value;
Using the index value, a surface condition of the processed surface that is correlated with the index value is estimated;
The correlation is expressed by a linear approximation equation,
a slope of the linear approximation equation increases as an average luminance of the plurality of pixels forming the image increases, and an intercept of the linear approximation equation decreases as the average luminance increases,
a surface condition estimating method, wherein the surface condition of the processed surface is estimated using the index value and the linear approximation equation having the slope and the intercept determined based on the average luminance. determining a linear approximation of the slope with respect to the average luminance and a linear approximation of the intercept with respect to the average luminance based on the slope and the intercept of the linear approximation equation for each of the average luminances;
8. The surface condition estimating method according to claim 6, wherein the surface condition of the processed surface is estimated using the index value, the slope obtained from a linear approximation of the slope corresponding to the average luminance in the image, and the intercept obtained from a linear approximation of the intercept corresponding to the average luminance in the image. The surface condition estimation method according to any one of claims 1 to 8, further comprising: setting a timing for correcting the tool based on the estimated surface condition of the machined surface; or changing the machining conditions under which the machine tool uses the tool to machine the workpiece. The surface condition estimation method according to any one of claims 1 to 9, wherein the adjacent pixels are four or more adjacent pixels. The surface condition estimation method according to any one of claims 1 to 10, wherein the machine tool is a grinding device having a grinding wheel as the tool and grinding the workpiece with the grinding wheel. The surface condition estimation method according to any one of claims 1 to 11, wherein the surface condition of the processed surface is surface roughness. A surface condition estimation system that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, comprising:
an imaging device that images the processing surface and outputs the captured image;
an image evaluation device having: an evaluation area setting unit that sets an evaluation area for the image; an area dividing unit that divides the evaluation area into a plurality of areas formed by a plurality of adjacent pixels among a plurality of pixels that form the image; a brightness average value calculation unit that calculates a brightness average value using the brightness of each of the adjacent pixels in each of the areas ; a brightness difference value calculation unit that calculates a brightness difference value representing a difference between the average brightness value calculated by the brightness average value calculation unit in each of the areas and the brightness of each of the adjacent pixels; and an index value calculation unit that calculates the square root of the sum of the squares of the brightness difference value calculated by the brightness difference value calculation unit as an index value;
an estimation calculation device having an estimation unit that estimates a surface state of the processed surface that is correlated with the index value calculated by the image evaluation device;
A surface condition estimation system comprising: A surface condition estimation system that is applied to a machine tool that machines a workpiece with a tool and estimates a surface condition of a machined surface of the tool and the workpiece, comprising:
an imaging device that images the processing surface and outputs the captured image;
an image evaluation device having: a luminance average calculation unit that calculates a luminance average value using the luminance of each of a plurality of adjacent pixels among a plurality of pixels forming the image output from the imaging device; a luminance difference value calculation unit that calculates a luminance difference value representing a difference between the luminance average value calculated by the luminance average calculation unit and the luminance of each of the plurality of adjacent pixels; and an index value calculation unit that calculates, as an index value, a square root of the sum of squares of the luminance difference value calculated by the luminance difference value calculation unit;
an estimation calculation device having an estimation unit that estimates a surface state of the processed surface that is correlated with the index value calculated by the image evaluation device;
Equipped with
The correlation is expressed by a linear approximation equation,
a slope of the linear approximation equation increases as an average luminance of the plurality of pixels forming the image increases, and an intercept of the linear approximation equation decreases as the average luminance increases,
a surface condition estimating system, wherein the surface condition of the processed surface is estimated using the index value and the linear approximation equation having the slope and the intercept determined based on the average luminance.

Images