Apparatus for estimating processing state and method for estimating processing state

Abstract

To provide a machining state estimation device that accurately estimates the machining state. [Solution] The storage device of the machining state estimation device stores a plurality of machining state values ​​that define the machining state of the machining machine, and a plurality of reference data, each corresponding to a combination of the plurality of machining state values. The processor acquires first measurement data measured during machining and second measurement data corresponding to the first measurement data, and divides at least a portion of the first measurement data into a first range and a second range. The processor determines a first determination value corresponding to the first range and a second determination value corresponding to the second range in the second measurement data, determines a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range in the plurality of reference data, and determines the similarity between each of the plurality of reference data and the second measurement data based on the first determination value, second determination value, first reference parameter, and second reference parameter, and estimates the machining state based on the similarity.

Description

Technical Field

[0001] The present disclosure relates to a processing state estimation device and a processing state estimation method.

Background Art

[0002] Patent Documents 1 and 2 disclose technologies for acquiring measurement data indicating measurement results of a processing load installed in a processing machine and estimating a processing state. The processing state estimation device of Patent Document 1 determines the similarity between reference data corresponding to a combination of a plurality of parameters defining the processing state of the processing machine and measurement data indicating the measurement result of the processing load, and estimates the processing state. The processing state estimation device of Patent Document 2 generates reference data regarding the processing load based on the definition of the lengths of a plurality of regions obtained by dividing a punching contour, determines the similarity with measurement data indicating the measurement result of the processing load, and estimates the processing state, compared with the processing state estimation device of Patent Document 1.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] An object of the present disclosure is to provide a processing state estimation device and a processing state estimation method capable of estimating the processing state of a processing machine with higher accuracy than the prior art.

Means for Solving the Problems

[0005] A processing state estimation device according to an aspect of the present disclosure includes a storage device and a processor, wherein the storage device a plurality of processing state values defining the processing state of the processing machine, and Each stores multiple reference data corresponding to the combination of the multiple processing state values, The aforementioned processor, During the machining process of the aforementioned machining machine, first measurement data and second measurement data corresponding to the first measurement data are acquired. At least a portion of the above-mentioned first measurement data is divided into a first range and a second range, In the second measurement data, a first determined value corresponding to the first range and a second determined value corresponding to the second range are determined. In the aforementioned plurality of reference data, a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range are determined. Based on the first determined value, the second determined value, the first reference parameter, and the second reference parameter, the similarity between each of the plurality of reference data and the second measurement data is determined. Based on the determined similarity, the processing state is estimated.

[0006] A method for estimating the processing state according to one aspect of this disclosure is: A step of acquiring first measurement data and second measurement data corresponding to the first measurement data, which are measured during machining by a machining machine. The steps include dividing at least a portion of the first measurement data into a first range and a second range, The steps include determining a first determined value corresponding to the first range and a second determined value corresponding to the second range in the second measurement data, A step of determining a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range in a plurality of reference data, each corresponding to a plurality of combinations of machining state values ​​that define the machining state of the machining machine, A step of determining the similarity between each of the plurality of reference data and the second measurement data based on the first determined value, the second determined value, the first reference parameter, and the second reference parameter, A step of estimating the processing state based on the determined similarity, including.

Advantages of the Invention

[0007] According to the processing state estimation device and the processing state estimation method according to the present disclosure, the processing state of a processing machine can be estimated with higher accuracy and at higher speed than in the prior art.

Brief Description of the Drawings

[0008] [Figure 1] It is a block diagram showing a configuration example of a processing state estimation device according to an embodiment. [Figure 2] It is a schematic cross-sectional view showing a press machine to which the load sensor and the distance sensor shown in FIG. 1 are attached. [Figure 3] It is a schematic graph showing an example of a measurement waveform by the load sensor and the distance sensor shown in FIG. 1. [Figure 4] It is a flowchart exemplifying the procedure of the processing state estimation process executed by the CPU of the processing state estimation device in FIG. 1. [Figure 5] It is a diagram for explaining a processing waveform extracted from a measurement waveform. [Figure 6] It is a diagram for explaining an example of the division of a processing waveform. [Figure 7] It is a diagram showing the state of a cross section obtained by the simulation of the punching process shown in FIG. 2. [Figure 8] It is a diagram for explaining another example of the division of a processing waveform. [Figure 9] It is a schematic graph showing an example of a stress-strain curve of a workpiece. [Figure 10] It is a flowchart exemplifying the procedure of the reference data determination process. [Figure 11] It is a table showing an example of the reference data shown in FIG. 1. [Figure 12] It is a graph for explaining that the actual processing state value is reflected in the measurement waveform.

Modes for Carrying Out the Invention

[0009] (Findings on which the present disclosure is based) As a result of repeated research to estimate the machining state in machining by a machine tool, particularly in cycle machining in which the same machining is repeated, the inventors have obtained the following findings. Here, the "machining state" indicates parameters that define the state of machining in the present disclosure, and includes, for example, at least one of the wear amount of a tool, the clearance, the thickness of a workpiece, or the hardness of the workpiece. The "machining state value" is a value indicating the machining state.

[0010] The wear amount of the tool includes the punch wear amount and the die wear amount. The punch wear amount and the die wear amount are examples of a punch wear parameter, which is an index indicating the degree of wear of the punch, and a die wear parameter, which is an index indicating the degree of wear of the die, respectively. The wear amount of the tool such as the punch wear amount and the die wear amount is represented, for example, by a dimensional change from the dimension of the tool at the time of newly manufactured or re-polished. Alternatively, the wear amount of the tool may be represented by a dimensional change from the design value of the tool. The wear amount of the tool may be represented by a change amount such as a shape change, a volume change, or a mass change. Further, the wear amount of the tool may be represented by the radius of the arc when the wear is approximated as an arc.

[0011] The clearance is the gap between the die and the punch. For example, the clearance is the gap between the die and the punch when a punching hole is opened in the workpiece. The clearance may be represented by the ratio of the gap between the die and the punch to the thickness of the workpiece.

[0012] The workpiece thickness is the thickness of the workpiece to be machined, and is measured as a length by, for example, a micrometer, a laser displacement meter, etc., and its unit is, for example, μm.

[0013] The workpiece hardness is a value indicating the hardness of the workpiece to be machined, and is measured as Vickers hardness by, for example, a Vickers hardness test, etc., and its unit is, for example, HV.

[0014] In machining using processing machines, the machining state is reflected in the specific operations performed using the tool. For example, it is known that the machining state is reflected in measurement data obtained by sensing the machining point.

[0015] For example, in press working, the load applied to the tool corresponds to the processing point. Therefore, it is conceivable to estimate the processing state value from the load waveform obtained by sensing the load applied to the tool.

[0016] The load applied to the punch or workpiece during machining depends on parameters such as punch wear, die wear, clearance, workpiece thickness, and workpiece hardness. Therefore, it is conceivable to estimate these parameters from the load waveform obtained during machining. For example, if the tool condition can be estimated from the change in the slope of the waveform, the optimal timing for tool maintenance can be determined in machining equipment that performs cyclic machining. Performing maintenance at the optimal timing can prevent situations such as machining workpieces with worn tools and producing a large number of defective products, thereby increasing productivity.

[0017] In machining centers that perform cyclical processing, there is an advantage to using the estimated processing state from the previous punching operation to estimate the processing state. One reason for this is that values ​​such as clearance, punch wear, and die wear usually do not change significantly from the values ​​in the previous punching operation. By performing estimations under the condition that the results do not change significantly from the previous estimation, the accuracy of the estimation can be improved.

[0018] Regarding the condition of the workpiece, if the processing state can be estimated in-line, for example, for each processing cycle by the processing machine, there is no need to measure the condition of the workpiece using a separate measuring instrument. The workpiece condition estimated in this way can not only be used for traceability, but by observing its trends, it can also be used for process control.

[0019] Furthermore, if the condition of the workpiece can be estimated for each processing cycle, fluctuations in the workpiece condition within and between lots can be managed or displayed as a trend. This can then be used for managing the condition of workpieces fed into the press machine, process control, and process improvement through analysis of these factors.

[0020] The load waveform used in estimating the machining state is obtained for each machining cycle. Here, a machining cycle is the period from the start of the punch's descent to the end of punching, when the punch rises and stops. The waveform obtained in a machining cycle includes the period during which machining is actually being performed on the workpiece, from the start of punching (for example, when the punch makes contact with the workpiece) to the end of punching (for example, when the punch is no longer in contact with the workpiece).

[0021] Conventional techniques estimate the processing state by calculating the similarity within this range all at once. Here, similarity is an index that indicates the degree of similarity between two waveforms.

[0022] In conventional techniques, for example, the similarity between a load waveform obtained through measurement and a reference waveform stored in a pre-set processing machine is calculated. However, within the range of the load waveform from the start to the end of punching, the final stage of punching near the end of punching includes a range where the load decreases rapidly from its peak to near zero. Since the load waveform in this range can vary significantly from one punching cycle to the next, if the similarity is calculated collectively without considering this range, the similarity may be calculated poorly or may vary significantly from cycle to cycle.

[0023] As a result of diligent research in this regard, the inventors conceived a method for calculating the similarity between load waveforms obtained by sensing the processing point and reference waveforms. This method involves dividing each waveform into, for example, the end of punching and the remaining range, and comparing the waveforms within each range. The inventors found that calculating the similarity in this way and estimating the processing state based on the calculation results improves the accuracy of the estimation. Furthermore, they found that dividing the waveform and calculating the similarity for each divided range reduces the amount of data required for calculation, thereby shortening the calculation time while maintaining the accuracy of the estimation. Based on these findings, the inventors discovered that by considering the physical processing state when calculating the similarity between processing waveforms obtained by sensing the processing point and reference waveforms, the processing state can be estimated with greater accuracy and the estimation time can be shortened, leading to the present invention.

[0024] Embodiments of this disclosure will be described in detail below, with reference to the drawings as appropriate. However, unnecessary details may be omitted. For example, detailed explanations of already well-known matters or redundant explanations of substantially identical configurations may be omitted. This is to avoid the following explanation becoming unnecessarily verbose and to facilitate understanding by those skilled in the art. The inventors provide the accompanying drawings and the following explanation so that those skilled in the art can fully understand this disclosure, and do not intend to limit the subject matter described in the claims by means of these.

[0025] (Embodiment) [1. Structure] Figure 1 is a block diagram showing an example configuration of a machining state estimation device 100 according to the present disclosure. The machining state estimation device 100 comprises a CPU 1, a storage device 2, an input interface (I / F) 3, and an output interface (I / F) 4.

[0026] CPU1 performs information processing to realize the functions of the processing state estimation device 100, which will be described later. Such information processing is realized, for example, by CPU1 operating according to the instructions of program 21 stored in memory device 2. CPU1 is an example of a processor of this disclosure. The processor does not need to include an arithmetic circuit that performs calculations for information processing, and is not limited to a CPU. For example, the processor may be composed of circuits such as an MPU or FPGA.

[0027] The storage device 2 is a recording medium that stores various information, including data such as reference data 22, which includes the reference waveform described later, and a program 21 necessary to realize the functions of the processing state estimation device 100. The storage device 2 can be implemented, for example, as a semiconductor storage device such as flash memory or a solid-state drive (SSD), a magnetic storage device such as a hard disk drive (HDD), or other recording media, either alone or in combination thereof. The storage device 2 may also include volatile memory such as SRAM or DRAM.

[0028] The input interface 3 is an interface circuit that connects the machining state estimation device 100 to an external device in order to input information such as detection results from the load sensor 11 and the distance sensor 12 to the machining state estimation device 100. Such external devices are, for example, the load sensor 11 and other information processing terminals. The input interface 3 may also be a communication circuit that performs data communication according to an existing wired communication standard or wireless communication standard.

[0029] The output interface 4 is an interface circuit that connects the machining state estimation device 100 to an external output device in order to output information from the machining state estimation device 100. Such an output device may be, for example, a display or another information processing terminal. The output interface 4 may also be a communication circuit that performs data communication according to an existing wired communication standard or wireless communication standard. The input interface 3 and the output interface 4 may be implemented by similar hardware.

[0030] Figure 2 is a schematic cross-sectional view showing a press machine 50 to which the load sensor 11 and distance sensor 12 shown in Figure 1 are attached. For ease of explanation, Figure 2 shows the X, Y, and Z axes perpendicular to each other. The Z axis represents the vertical direction.

[0031] The press machine 50 is an example of a processing machine that performs cyclic processing, which involves repeating a predetermined processing step. The press machine 50 comprises a bolster 51 and a slide 52 that repeatedly performs an up-and-down cyclic motion from top dead center to bottom dead center relative to the bolster 51. A die backing plate 61 is mounted on top of the bolster 51, and a die plate 62 is mounted on top of the die backing plate 61. The die plate 62 grips the die 63.

[0032] A punch backing plate 71 is attached to the lower part of the slide 52, and a punch plate 72 is attached to the lower part of the punch backing plate 71. The punch plate 72 grips the punch 73. The press machine 50 further includes a stripper plate 74. The stripper plate 74 is attached to the punch plate 72 or punch backing plate 71 and fasteners such as bolts via positioning guides such as posts (not shown). The stripper plate 74 is biased downward by, for example, a compression spring and has the function of guiding the punch 73 to maintain a constant position, removing material adhering to the punch 73 after punching the workpiece 80, and / or fixing the workpiece 80 when punching it.

[0033] The load sensor 11 is installed, for example, between the punch 73 and the punch backing plate 71. The load sensor 11 is an electrical force sensor such as a piezoelectric force sensor, a semiconductor strain sensor, or a strain gauge type, and measures the load applied to the punch 73 when the punch 73 punches out the workpiece 80. Its unit is, for example, Newtons (N).

[0034] The distance sensor 12 is installed, for example, on the die backing plate 61. The distance sensor 12 is, for example, an eddy current gap sensor or a laser displacement meter, and measures the distance to, for example, the punch plate 72, which is opposite in the Z-axis direction. The distance measured by the distance sensor 12 indicates the relative position of the punch 73 with respect to a predetermined reference position of the press machine 50.

[0035] Figure 3 is a schematic graph showing an example of a measurement waveform 200 from the load sensor 11 and the distance sensor 12. The horizontal axis of the graph in Figure 3 represents the position of the punch 73 relative to its initial position. The position of the punch 73 is measured by the distance sensor 12. The horizontal axis of the graph in Figure 3 can also be considered the distance the punch 73 has traveled in the negative direction of the Z axis. The vertical axis of the graph in Figure 3 represents the load measured by the load sensor 11.

[0036] The measurement waveform 200 in Figure 3 represents the entire processing cycle for punching, including the start of the punch 73's descent, punching, the rise of the punch 73, and the stopping of the punch 73. The measurement waveform 200 shows a bell-shaped waveform that rises from the moment the punch 73 descends and contacts the workpiece 80, reaches the peak load value HM, and then rapidly decreases to near 0 after the workpiece 80 is punched out.

[0037] The punching period for a punching process can be determined, for example, as the period from the point in the measurement waveform when the load first exceeds the threshold Th until the point in the measurement waveform when it first falls below the threshold Th. Such a threshold Th may be defined as an absolute value or as a percentage of the peak value of the load. Furthermore, there is not limited to one threshold, but there may be two or more. For example, a rising threshold and a falling threshold may be stored in the storage device 2, and the CPU 1 may define the punching period as the period from the point in the measurement waveform when the load first exceeds the rising threshold until the point in the measurement waveform when it first falls below the falling threshold.

[0038] [2. Operation] The overview of the machining state estimation process will be explained with reference to Figures 4-12. Figure 4 is a flowchart illustrating the procedure of the machining state estimation process performed by the CPU 1 of the machining state estimation device 100 shown in Figure 1.

[0039] First, the CPU 1 acquires measurement waveforms from the load sensor 11 and the distance sensor 12, showing the measurement results of a series of position data measured by the distance sensor 12 and the corresponding load data applied to the load sensor 11 during press processing by the press machine 50 (S1). Thereafter, when it is necessary to combine the position data and load data, they are collectively called measurement data, and any single point in the measurement data is called a measurement point. In this embodiment, the measurement data is a collection of measurement points.

[0040] Next, CPU1 extracts the machining waveform to be processed by CPU1 from the measurement waveform obtained in step S1 (S2). Figure 5 is a diagram illustrating the machining waveform W extracted from the measurement waveform. For example, CPU1 obtains the starting load HS and ending load HE from the storage device 2, determines the starting point WS corresponding to the starting load HS and the ending point WE corresponding to the ending load HE in the measurement waveform, and defines the portion of the measurement waveform between the starting point WS and the ending point WE as the machining waveform W.

[0041] The starting point WS of the processed waveform W is determined, for example, as the position where the starting load HS is first exceeded. The starting load HS is obtained by multiplying the peak load value HM by a predetermined ratio. The ending point WE of the processed waveform W is determined, for example, as the position where the ending load HE is last fallen below the peak load value HM. The ending load HE is obtained by multiplying the peak load value HM by a predetermined ratio.

[0042] In this embodiment, the start point WS and end point WE of the processed waveform W were determined based on the start point load HS and end point load HE, respectively, which are obtained by multiplying the peak load value HM by a predetermined ratio. However, the method for determining the start and end points is not limited to this. For example, the CPU1 may determine the start and end points by using predetermined measurement points as the start and end point loads.

[0043] Returning to Figure 4, the CPU1 then divides the machining waveform W determined in step S2 into several ranges that exhibit characteristics based on the machining process (S3). The CPU1 can then determine the load value and / or width (machining length) of the waveform in each of the divided ranges.

[0044] In this embodiment, the division step S3 includes a step S3A for vertical division of the processed waveform and a step S3B for horizontal division.

[0045] Figure 6 is a diagram illustrating an example of the vertical division process S3A of the machining waveform W. In the vertical division process S3A, the CPU1 vertically divides the range from the start point to the end point of the machining waveform W into multiple ranges. The vertical division is performed such that the width of each divided range relative to the width of the machining waveform W (WE-WS) is a predetermined ratio. For example, position A1 is the 0% position when the start point of the machining waveform W is 0% and the end point is 100%, and position A8 is the 100% position. For example, position A2 is the 10% position, and position A3 is the 30% position. The same applies to positions A4 to A8, only the ratio values ​​differ.

[0046] Positions A1 to A8 do not need to be equally spaced from the start to the end of the machining waveform W. For example, positions A1 to A8 may be determined by considering characteristics based on the machining process. For instance, if a portion of the waveform that changes significantly when the machining state value changes is known, CPU1 may divide the machining waveform W such that at least one of the resulting waveforms includes that portion. This results in divided waveforms that reflect the actual changes in the machining state.

[0047] Figure 7 shows the cross-sectional state obtained from the punching simulation shown in Figure 2. Various phases exist in the changes of the workpiece during processing. For example, Figure 7(a) shows the phase in which the tool and workpiece come into contact, and Figure 7(b) shows the phase in which the workpiece undergoes elastic deformation. Figure 7(c) shows the phase in which the workpiece is mainly pulled by the tool and undergoes plastic deformation due to stress. Figure 7(d) shows the phase in which the workpiece undergoes plastic deformation due to shear stress. Shear stress is mainly applied to the workpiece by the tool. Figure 7(e) shows the phase in which the workpiece fractures. It can be seen that the stress distribution changes in response to the punching load from the phase shown in Figure 7(a) to the phase shown in Figure 7(e).

[0048] Thus, the stress distribution in the workpiece changes with changes in the machining state values ​​of the tool and the workpiece. For example, the width of the machining waveform W changes with changes in the machining state values. Characteristic points in the change of stress distribution roughly correspond to the ratio value to the width of the machining waveform W. Therefore, in this embodiment, by determining positions A1 to A8 within the range of the machining waveform W using predetermined ratio values ​​corresponding to the actual changes in the machining state, it is possible to determine positions that more accurately characterize the machining state.

[0049] Returning to Figure 6, in the vertical division process S3A, the CPU1 determines the load values ​​H1 to H8 corresponding to positions A1 to A8. Here, the measurement data input to the CPU1 via the load sensor 11, distance sensor 12, and input interface 3 may be discrete data digitized by A / D conversion. In this case, the measurement waveform shown in Figure 3 is obtained by connecting discrete measurement points. The CPU1 determines the load values ​​H1 to H8 by finding the measurement points in the vicinity of each of positions A1 to A8 from the measurement data. For example, the load value H1 corresponding to position A1 may be determined by linearly interpolating the load values ​​corresponding to the measurement points immediately before and after position A1. The same applies to positions A2 to A8.

[0050] The number of load values ​​determined in the vertical partitioning process S3A corresponds to the number of dimensions of the load values ​​in the calculation of the load distance described later. In the example in Figure 6, the load values ​​are 8-dimensional, and the values ​​of each dimension correspond to load values ​​H1 to H8.

[0051] In this embodiment, the load values ​​H1 to H8 were determined by linear interpolation from measurement points immediately before and after each of the corresponding positions A1 to A8. However, the method for determining the load values ​​is not limited to this. For example, CPU1 may determine the load values ​​H1 to H8 from the nearest measurement point to each of the positions A1 to A8, or it may determine them by interpolation curve processing from multiple measurement points located before and after each of the positions A1 to A8. Furthermore, the dimensions of the load values ​​are not limited to eight dimensions, but may be one or more.

[0052] Figure 8 is a diagram illustrating the horizontal division process S3B in Figure 4. In the horizontal division process S3B, CPU1 divides the height (load) of the machining waveform W into multiple ranges. The height of the machining waveform W is represented, for example, as the peak value HM of the machining waveform with zero load (horizontal axis in Figure 8) as the reference. The division is performed so that the height of each range after division is a predetermined ratio with respect to the height of the machining waveform W. Load B1 in Figure 8 is 90% of the peak value HM, and load B2 is 80% of the peak value HM. Loads B3 and B4 are similar, only their respective ratio values ​​differ.

[0053] The loads B1 to B4 do not need to be spaced equally and may be determined by considering characteristics based on the machining process. For example, if a portion of the waveform that changes significantly when the machining state value changes is known, CPU1 may divide the machining waveform W such that at least one of the resulting waveforms includes that portion. This results in divided waveforms that reflect the actual changes in the machining state.

[0054] Figure 9 is a schematic graph showing an example of a stress-strain curve for workpiece 80. The graph in Figure 9 shows the stress-strain curve for SUS301, a type of stainless steel. In the graph in Figure 9, the horizontal axis represents true strain, and the vertical axis represents true stress. The stress-strain curve in Figure 9 consists of the elastic region D1, the plastic deformation region D2, and the plastic deformation region D3. In the elastic region D1, the stress-strain curve is close to a straight line. In the plastic deformation region D2, the increase in true stress is greater in response to the increase in true strain compared to the plastic deformation region D3. In each phase shown in Figures 7(a) to 7(e), the stress acting inside workpiece 80 is affected by material properties as represented by the elastic region D1, the plastic deformation region D2, and the plastic deformation region D3.

[0055] Thus, the rate of change of true stress in response to a change in true strain differs in the elastic region D1, the plastic deformation region D2, and the plastic deformation region D3. Therefore, loads B1 to B4 in Figure 8 can be determined in response to this change. For example, load B4 is set to the same value as the stress corresponding to the boundary between the elastic region D1 and the plastic deformation region D2. Since the stress corresponding to the boundary between the elastic region D1 and the plastic deformation region D2 corresponds to the load value applied during punching, changes in the waveform are likely to appear around load B4, which is set as described above, due to changes in the processing state value. Loads B1 to B3 may also be determined in response to changes influenced by material properties, such as those represented by the plastic deformation region D2 and the plastic deformation region D3. In this way, by determining loads B1 to B4 within the range of the processing waveform W using characteristic values ​​that correspond to changes in the actual processing state, it is possible to determine loads that more accurately characterize the processing state.

[0056] Returning to Figure 8, in the horizontal division process S3B, CPU1 determines widths W1 to W4 corresponding to loads B1 to B4. In the graph of Figure 8, width W1 is the width of the processed waveform W for load B1. Similarly, widths W2 to W4 are the widths of the processed waveform W for loads B2 to B4, respectively. CPU1 determines widths W1 to W4 by finding the nearest measurement points for each of the loads B1 to B4 from the measurement data. For example, width W1 corresponding to load B1 may be determined by linear interpolation of the widths corresponding to the measurement points immediately before and after load B1. The same applies to loads B2 to B4.

[0057] The number of widths determined by the horizontal division process S3B corresponds to the number of dimensions of the width in the calculation of the width distance described later. In the example in Figure 8, the width is 4-dimensional, and the values ​​of each dimension correspond to widths W1 to W4.

[0058] In this embodiment, widths W1 to W4 were determined by linear interpolation from the measurement points immediately before and after each of the corresponding loads B1 to B4, but the method of determining the width is not limited to this. For example, CPU1 may determine the load values ​​H1 to H8 from the nearest measurement point for each of the loads B1 to B4, or it may determine them by interpolation curve processing from multiple measurement points located before and after each of the loads B1 to B4. Furthermore, the dimension of the width is not limited to four dimensions, but may be one or more.

[0059] Returning to Figure 4, CPU1 performs a reference data determination process (S4) to determine the reference data 22. Figure 10 is a flowchart illustrating the steps of the reference data determination process S4.

[0060] In the reference data determination process S4, the CPU1 determines the machining state value (S41).

[0061] Figure 11 is a table showing an example of the reference data 22 shown in Figure 1. The reference data 22 is a database that includes a dataset consisting of a combination of multiple machining state values ​​and a reference waveform, reference load, and reference width corresponding to this combination. According to this embodiment, the machining state values ​​include values ​​representing clearance, punch wear amount, die wear amount, workpiece thickness, and workpiece hardness.

[0062] In step S41, CPU1 determines one combination of machining state values ​​from the reference data 22 (for example, the i-th combination in reference data 22, where i is an integer greater than or equal to 1). In reference data 22, clearance, punch wear amount, die wear amount, workpiece thickness, and workpiece hardness are discrete state values, and reference data 22 is composed of combinations thereof. For example, clearance is in a range of 1 μm to 10 μm in 1 μm steps, punch wear amount is in a range of 0 μm to 20 μm in 1 μm steps, die wear amount is in a range of 0 μm to 20 μm in 1 μm steps, workpiece thickness is in a range of 25 μm to 35 μm in 1 μm steps, and workpiece hardness is in a range of 300 HV to 500 HV in 20 HV steps.

[0063] Next, CPU1 acquires a reference waveform corresponding to the combination of machining state values ​​determined in step S41 (S42). The reference waveform acquired in step S42 is predetermined by simulation using a combination of clearance, punch wear amount, die wear amount, workpiece thickness, and workpiece hardness, and is stored, for example, in storage device 2.

[0064] Next, CPU1 obtains the reference load and reference width corresponding to the reference waveform acquired in step S42 from the reference data 22 (S43).

[0065] CPU1 determines whether it has acquired the reference loads and reference widths corresponding to the reference waveform (S44). In step S43, CPU1 succeeds in acquiring the reference loads Y1 to Y8 and reference widths X1 to X4 only if the reference data 22 contains the corresponding reference loads Y1 to Y8 and reference widths X1 to X4. If CPU1 was able to acquire the reference loads Y1 to Y8 and reference widths X1 to X4 from the reference data 22 in step S43 (Yes in S44), CPU1 outputs the reference loads Y1 to Y8 and reference widths X1 to X4.

[0066] On the other hand, if the reference loads Y1 to Y8 and reference widths X1 to X4 could not be obtained from the reference data 22 in step S43 (No in S44), the CPU1 performs a calculation to determine the reference loads and reference widths from the reference waveform (S45). In step S45, the same processing as in step S3 is performed. The processing waveform in step S3 corresponds to the reference waveform, the load values ​​H1 to H8 correspond to the reference loads Y1 to Y8, and the widths W1 to W4 correspond to the reference widths X1 to X4.

[0067] Next, after completing step S45, CPU1 adds the reference loads Y1-Y8 and reference widths X1-X4 obtained in step S45 to the reference data 22 as the reference loads and reference widths corresponding to the reference waveform acquired in step S42. CPU1 saves the updated reference data 22 to the storage device 2 (S46). In addition, CPU1 determines the reference loads Y1-Y8 and reference widths X1-X4 saved in step S46 as the output of step S4, and completes the reference data determination process S4.

[0068] In this embodiment, an example has been described in which the reference loads Y1 to Y8 and reference widths X1 to X4 are determined by calculation in step S4 after acquiring the measurement waveform (S1) and saved as reference data 22 (S46). However, this disclosure is not limited to this. For example, the reference loads Y1 to Y8 and reference widths X1 to X4 for all combinations of machining state values ​​may be calculated in advance and stored in the storage device 2 as reference data 22. Using this, in step S43, the CPU 1 may acquire the reference loads Y1 to Y8 and reference widths X1 to X4 corresponding to the determined machining state values ​​and reference waveforms from the reference data 22.

[0069] Returning to Figure 4, CPU1 performs a calculation to determine the similarity between the processed waveform and the reference waveform (S5). Here, similarity is an index that indicates the degree of similarity between the two waveforms. Similarity can be calculated, for example, using the cosine similarity, Euclidean distance, or Manhattan distance between the two waveforms during the punching period. Another example of similarity is the degree of agreement and the degree of mismatch. The degree of agreement is an index that indicates the degree of agreement between the two waveforms. The degree of mismatch is an index that indicates the degree of mismatch between the two waveforms. CPU1 may use either the degree of agreement or the degree of mismatch as the similarity.

[0070] In step S5, CPU1 calculates the similarity by associating the load values ​​H1~H8 and width W1~W4 in the machining waveform with the reference loads Y1~Y8 and reference widths X1~X4 in the reference waveform. Here, the load values ​​in the machining waveform are Hn (n=1~8) and the width is Wm (m=1~4). If the reference load in the reference waveform is Yn and the reference width is Xm, the Euclidean distance between the machining waveform and the reference waveform with respect to the load value (hereinafter referred to as "load distance") HY and the Euclidean distance with respect to the width (hereinafter referred to as "width distance") WX are calculated using the following formulas.

[0071]

number

[0072]

number

[0073] Since the unit of the load value is N and the unit of the width is mm, the units can be standardized by converting from mm to N, for example, by multiplying the width by a coefficient E. Therefore, the waveform distance F, which indicates the distance between the processed waveform and the reference waveform, can be calculated using the following formula.

[0074]

number

[0075] The waveform similarity indicating the similarity between the processed waveform and the reference waveform is, for example, the reciprocal of the waveform distance F, and the larger it is, the more the processed waveform and the reference waveform match. Also, the waveform similarity is obtained through an n-dimensional load value and an m-dimensional width determined in consideration of the characteristics of the actual processing state.

[0076] In this embodiment, an example in which the load distance HY, the width distance WX, and the waveform distance F are calculated as Euclidean distances has been described, but part or all of the distances between these waveforms may be calculated as Manhattan distances.

[0077] The reference data determination process S4 and the similarity calculation process S5 in FIG. 4 are repeatedly executed a predetermined number of times while changing the combination of the processing state values determined in step S41. For example, the above integer i is incremented each time the above process is repeated. By repeatedly executing the above process, the CPU 1 searches for the reference waveform having the largest similarity with the processed waveform (within the search range).

[0078] To perform such a search, the CPU 1 determines whether it has finished comparing the processed waveform with a predetermined number of reference waveforms (S6). The predetermined number is an integer of 2 or more. The predetermined number may, for example, match the number of rows N of the reference data 22. In this case, the CPU 1 will compare the processed waveform with all the reference waveforms in the reference data 22.

[0079] Alternatively, the predetermined number may be less than the number of rows N of the reference data 22. For example, the CPU 1 compares the processed waveform with each of the reference waveforms from the i-th to the (i + k)-th. Here, k is a predetermined integer of 1 or more, and (i + k) < N. Thereby, in step S6, it is possible to determine whether the search for the processing state value close to the previous processing state value has been completed.

[0080] If it is determined that the comparison with a predetermined number of reference waveforms has not been completed (No in S6), CPU1 returns to step S4 and restarts the process of determining another reference data (for example, the (i+1)th reference data).

[0081] On the other hand, if it is determined that comparison with a predetermined number of reference waveforms has been completed (Yes in S6), CPU1 determines the reference waveform with the greatest similarity to the processed waveform (S7).

[0082] Next, CPU1 outputs machining state values ​​corresponding to the reference waveform determined in step S7 (S8). As an output of step S8, for example, the machining state estimation device 100 displays a trend graph showing the changes in machining state values ​​for the user. This allows the user to refer to the machining state values ​​and, if necessary, perform maintenance on tools or adjust machining conditions.

[0083] Figure 12 is a graph illustrating how the actual machining state values ​​are reflected in the measured waveform. The graph in Figure 12 shows the measured waveform 200, reference waveform 201, and reference waveform 202. The measured waveform 200 is the waveform measured when the workpiece thickness is 32.5 μm. Reference waveform 201 shows the reference waveform of reference data 22 when the workpiece thickness is 33 μm and the machining state values ​​other than the workpiece thickness are equal to the actual values. Reference waveform 202 shows the reference waveform of reference data 22 when the workpiece thickness is 32 μm and the machining state values ​​other than the workpiece thickness are equal to the actual values. Under these conditions, as shown in Figure 12, the measured waveform 200 constitutes a waveform intermediate between reference waveform 201 and reference waveform 202. The combinations of machining state values ​​are generally close for the three waveforms, measured waveform 200, reference waveform 201, and reference waveform 202, and the machining states related to the three waveforms can be said to be similar.

[0084] Even though the processing conditions are similar, a large difference can occur between the measured waveform 200 and the reference waveform 201 in the falling edge range G1 where the waveform changes sharply. When comparing waveforms (calculating similarity), if the Euclidean distance between waveforms is calculated without dividing the entire range G2 of the punching period, the difference between waveforms in the falling edge range G1 will affect the Euclidean distance of the entire range G2, resulting in a larger and poorly calculated similarity.

[0085] Therefore, in methods for calculating similarity without dividing the waveform, even if the processing state is very similar, the similarity may be calculated poorly, making it difficult to accurately estimate the processing state. On the other hand, according to this embodiment, by dividing the waveform and calculating the similarity, the influence of the falling edge range G1 is reduced, and the processing state can be estimated with greater accuracy. Furthermore, compared to the case where similarity is calculated without dividing the waveform over the entire range G2 of the punching period, the number of dimensions required for the calculation decreases according to the number of divisions. As a result, the CPU1 can perform calculations for estimation at a faster speed and estimate the processing state. For example, if there are 100 sampling points in the entire range G2, it becomes necessary to calculate the Euclidean distance in 100 dimensions. However, according to this embodiment, the Euclidean distance only needs to be calculated in a lower dimension than the entire range G2, which is the dimensions of the load and distance acting on the load distance HY and width distance WX. This makes it possible to estimate the processing state in real time, even for high-speed press processing exceeding 2000 times per minute.

[0086] [3. Effects, etc.] As described above, the processing state estimation device 100 according to this embodiment comprises a storage device 2 and a CPU 1, which is an example of a processor. The storage device 2 stores a plurality of processing state values ​​that define the processing state of a press machine 50, which is an example of a processing machine, and a plurality of reference data, each corresponding to a combination of the plurality of processing state values. The CPU 1 acquires distance data, which is an example of first measurement data, and load data, which is an example of second measurement data corresponding to the distance data, measured during processing of the press machine 50 (S1). The CPU 1 divides at least a portion of the distance data into a first range and a second range, and determines a load value H1, which is an example of a first determined value corresponding to the first range, and a load value H2, which is an example of a second determined value corresponding to the second range, in the load data (S3). The CPU 1 determines a reference load Y1, which is an example of a first reference parameter corresponding to the first range, and a reference load Y2, which is an example of a second reference parameter corresponding to the second range, in the plurality of reference data (S4). CPU1 determines the similarity between each of the multiple reference data and the load data based on load value H1, load value H2, reference load Y1, and reference load Y2 (S5). Based on the determined similarity, CPU1 estimates the processing state (S7).

[0087] Unlike conventional technologies, this configuration allows for more accurate estimation of the machining state by considering the characteristics of the actual machining state when calculating the similarity between the machining waveform obtained by sensing the machining point and the reference waveform. For example, CPU1 can more accurately estimate machining state values ​​such as clearance, punch wear amount, and die wear amount. Therefore, maintenance can be performed at the appropriate time by having the operator refer to the machining state values ​​or by automatically stopping the equipment when a set threshold is exceeded. In addition, by dividing the range of the machining waveform and reducing the amount of data required for calculations within that range, the calculation time can be shortened while maintaining accuracy. For example, CPU1 can estimate machining state values ​​such as workpiece thickness and workpiece hardness in real time, even with high-speed presses.

[0088] With this configuration, the division of the machining waveform is determined by a predetermined ratio corresponding to the actual changes in the machining state, allowing it to be divided into ranges that better characterize the machining state, and as a result, the machining state can be estimated with greater accuracy.

[0089] CPU1 may determine values ​​corresponding to at least some of the starting points and values ​​corresponding to the ending points of the distance data based on the peak values ​​of the load data.

[0090] With this configuration, the range of the machining waveform can be determined based on values ​​that indicate characteristics based on the machining process, allowing for a more accurate estimation of the machining state.

[0091] CPU1 may determine at least one of the load values ​​H1 and H2 by interpolating the load data.

[0092] This configuration allows for improved accuracy in calculating similarity by interpolating discrete measurement data, thereby enabling more accurate estimation of the processing state.

[0093] The CPU1 may determine the degree of similarity between each of the multiple reference data and the load data by comparing the load value H1 with the reference load Y1 and the load value H2 with the reference load Y2.

[0094] This configuration allows for more accurate estimation of processing state values ​​compared to conventional technologies.

[0095] The similarity may also be the load distance HY, which is an example of the Euclidean distance between each of the multiple reference data and the load data.

[0096] This configuration allows for more accurate estimation of processing state values ​​compared to conventional technologies.

[0097] CPU1 may divide at least a portion of the load data into a third range and a fourth range. CPU1 determines a width W1, which is an example of a third determined value corresponding to the third range, and a width W2, which is an example of a fourth determined value corresponding to the fourth range, in the distance data. CPU1 determines a reference width X1, which is an example of a third reference parameter corresponding to the third range, and a reference width X2, which is an example of a fourth reference parameter corresponding to the fourth range, in the multiple reference data. Based on the widths W1, W2, X1, and X2, CPU1 further determines a width distance WX, which is an example of a second similarity between each of the multiple reference data and the distance data. CPU1 may estimate the processing state based on a load distance HY, which is an example of similarity, and a width distance WX, which is an example of a second similarity.

[0098] This configuration allows for a more accurate estimation of the machining state by considering the characteristics of the actual machining state in relation to the two measurement data points.

[0099] (Other embodiments) As described above, the above embodiments have been explained as examples of the technology disclosed in this application. However, the technology in this disclosure is not limited thereto and can be applied to embodiments that have been modified, replaced, added, or omitted as appropriate. Therefore, the following examples illustrate other embodiments.

[0100] (First variation) In the above embodiment, the unit conversion was performed on the width when calculating the waveform distance F. However, the conversion is not limited to distance, and the unit conversion may also be performed on the load value. In this modified example, the units can be standardized by multiplying the load value by a coefficient J, thereby converting from N to mm. In this case, the waveform distance F between the processed waveform and the reference waveform can be obtained by the following formula.

[0101]

number

[0102] The waveform distance F in the first deformation also indicates the distance between the machined waveform and the reference waveform; a smaller value indicates a better match between the machined waveform and the reference waveform. The waveform distance F can be determined by the characteristics of the actual machining state.

[0103] (Second variation) In the above embodiment, the waveform distance F was calculated by first obtaining the load distance HY and the width distance WX. However, the waveform distance F may be calculated by mixing the load value and the width without first obtaining the load distance HY and the width distance WX. In this modified example, the units can be standardized by multiplying the widths W1 to W4 and the reference widths X1 to X4 by a coefficient E, thereby converting from mm to N. At this time, the waveform distance F between the processed waveform and the reference waveform can be obtained by the following formula.

[0104]

number

[0105] In the second modified example, the waveform distance F also indicates the distance between the processed waveform and the reference waveform; a smaller value indicates a better match between the processed waveform and the reference waveform. The waveform distance F can be determined by the characteristics of the actual processing state.

[0106] (Third variation) In the above embodiment, an example was described in which both vertical division processing S3A and horizontal division processing S3B are performed (see Figure 4) to calculate the waveform distance F, but the present disclosure is not limited thereto. For example, the CPU 1 can achieve the objective of the present invention by performing at least one of vertical division processing S3A and horizontal division processing S3B. For example, depending on the type of machining, the machining waveform W may not change significantly even if the machining state value changes. If the machining waveform W does not change significantly, the machining state value can be estimated with good accuracy even if the waveform distance is calculated using only the load distance.

[0107] (Fourth variation) The above embodiment shows an example of determining the machining waveform from a measurement waveform consisting of a load corresponding to a position. However, the machining state estimation device according to this disclosure is not limited to a measurement waveform consisting of a load corresponding to a position. Since the machining device performs the same cycle machining, for example, the machining waveform may be determined from a measurement waveform consisting of a load corresponding to time. With this configuration, there is no need to measure distance, and the machining state value can be estimated with a simpler configuration.

[0108] (Fifth variation) The above embodiment shows an example of estimating machining state values ​​from the measured waveform of the load applied to the tool. However, the machining state estimation device according to this disclosure may estimate machining state values ​​based on, for example, pressure, strain, flow velocity, current, voltage, or temperature. With this configuration, for example, machining state values ​​can be estimated even when it is not possible to install a sensor to measure the load. Furthermore, in injection molding that performs cyclic machining, machining state values ​​such as shape in injection molding can be estimated by determining the machining waveform from the measured temperature waveform.

[0109] (Sixth variation) In the above embodiment, a press machine 50 was described as an example of a processing machine, but the processing machine is not limited to this and may be a molding device. This disclosure is applicable to these processing machines, and the processing state can be estimated based on parameters that affect abnormal processing in these processing machines.

[0110] (Example of a particular form) The following are examples of the aspects of this disclosure.

[0111] <Aspect 1> Equipped with a memory device and a processor, The aforementioned storage device is Multiple machining state values ​​that define the machining state of a machining machine, Each stores multiple reference data corresponding to the combination of the multiple processing state values, The aforementioned processor, During the machining process of the aforementioned machining machine, first measurement data and second measurement data corresponding to the first measurement data are acquired. At least a portion of the above-mentioned first measurement data is divided into a first range and a second range, In the second measurement data, a first determined value corresponding to the first range and a second determined value corresponding to the second range are determined. In the aforementioned plurality of reference data, a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range are determined. Based on the first determined value, the second determined value, the first reference parameter, and the second reference parameter, the similarity between each of the plurality of reference data and the second measurement data is determined. Based on the determined similarity, the processing state is estimated. Processing state estimation device.

[0112] <Aspect 2> The processing state estimation apparatus according to embodiment 1, wherein the processor determines a value corresponding to at least a portion of the starting point and a value corresponding to the ending point of the first measurement data based on the peak value of the second measurement data.

[0113] <Aspect 3> The processing state estimation apparatus according to embodiment 1 or 2, wherein the processor determines at least one of the first determined value and the second determined value by interpolating the second measurement data.

[0114] <Aspect 4> The processing state estimation apparatus according to any one of embodiments 1 to 3, wherein the processor determines the similarity between each of the plurality of reference data and the second measurement data by comparing the first determined value with the first reference parameter and comparing the second determined value with the second reference parameter.

[0115] <Aspect 5> The processing state estimation apparatus according to any one of embodiments 1 to 4, wherein the similarity is the Euclidean distance between each of the plurality of reference data and the second measurement data.

[0116] <Aspect 6> The aforementioned processor, At least a portion of the second measurement data is divided into a third range and a fourth range, In the first measurement data, a third determined value corresponding to the third range and a fourth determined value corresponding to the fourth range are determined. In the aforementioned plurality of reference data, a third reference parameter corresponding to the third range and a fourth reference parameter corresponding to the fourth range are determined. Based on the third determination value, the fourth determination value, the third reference parameter, and the fourth reference parameter, a second similarity is further determined between each of the plurality of reference data and the first measurement data. Based on the determined similarity and the second similarity, the processing state is estimated. A processing state estimation device according to any one of embodiments 1 to 5.

[0117] <Aspect 7> The processing machine is a processing machine that performs cycle processing, which involves repeating a predetermined processing method, as described in any of embodiments 1 to 6, for the processing state estimation device.

[0118] <Aspect 8> The processing machine is a press machine, according to any one of embodiments 1 to 7, for the processing state estimation device.

[0119] <Pattern 9> The aforementioned processor, By acquiring the first measurement data and the second measurement data, waveform data showing the measurement results of the change in processing load during each processing cycle by the press machine is obtained. The first measurement data indicates the relative position of the punch of the press machine with respect to a predetermined reference position of the press machine. The second measurement data indicates the processing load, The processing state estimation device described in embodiment 8.

[0120] <Aspect 10> A step of acquiring first measurement data and second measurement data corresponding to the first measurement data, which are measured during machining by a machining machine. The steps include dividing at least a portion of the first measurement data into a first range and a second range, The steps include determining a first determined value corresponding to the first range and a second determined value corresponding to the second range in the second measurement data, A step of determining a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range in a plurality of reference data, each corresponding to a plurality of combinations of machining state values ​​that define the machining state of the machining machine, A step of determining the similarity between each of the plurality of reference data and the second measurement data based on the first determined value, the second determined value, the first reference parameter, and the second reference parameter, A step of estimating the processing state based on the determined similarity, including, Method for estimating the state of processing. [Industrial applicability]

[0121] This disclosure is applicable to processing machines. [Explanation of Symbols]

[0122] 1 CPU1 2 Storage device 3. Input Interface 4 Output Interfaces 11. Load sensor 12 Distance Sensors 21 Programs 22 Reference Data 50 Press Machines 51 Bolster 52 slides 61 Die Backing Plate 62 Die Plates 63 Die 71 Punch backing plate 72 Punch Plates 73 punches 74 Stripper Plate 80 Work 100 Processing state estimation device

Claims

1. Equipped with a memory device and a processor, The aforementioned storage device is Multiple machining state values ​​that define the machining state of a machining machine, Each stores multiple reference data corresponding to the combination of the multiple processing state values, The aforementioned processor, During the machining process of the aforementioned machining machine, first measurement data and second measurement data corresponding to the first measurement data are acquired. At least a portion of the first measurement data is divided into a first range and a second range, In the second measurement data, a first determined value corresponding to the first range and a second determined value corresponding to the second range are determined. In the plurality of reference data, a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range are determined. Based on the first determined value, the second determined value, the first reference parameter, and the second reference parameter, the similarity between each of the plurality of reference data and the second measurement data is determined. Based on the determined similarity, the processing state is estimated. Processing state estimation device.

2. The processing state estimation apparatus according to claim 1, wherein the processor determines a value corresponding to the start point and a value corresponding to the end point of at least a portion of the first measurement data, based on the peak value of the second measurement data.

3. The processing state estimation apparatus according to claim 1, wherein the processor determines at least one of the first determined value and the second determined value by interpolating the second measurement data.

4. The processing state estimation apparatus according to claim 1, wherein the processor determines the similarity between each of the plurality of reference data and the second measurement data by comparing the first determined value with the first reference parameter and the second determined value with the second reference parameter.

5. The processing state estimation device according to claim 1, wherein the similarity is the Euclidean distance between each of the plurality of reference data and the second measurement data.

6. The aforementioned processor, At least a portion of the second measurement data is divided into a third range and a fourth range, In the first measurement data, a third determined value corresponding to the third range and a fourth determined value corresponding to the fourth range are determined. In the aforementioned plurality of reference data, a third reference parameter corresponding to the third range and a fourth reference parameter corresponding to the fourth range are determined. Based on the third determination value, the fourth determination value, the third reference parameter, and the fourth reference parameter, a second similarity is further determined between each of the plurality of reference data and the first measurement data. Based on the determined similarity and the second similarity, the processing state is estimated. The processing state estimation device according to claim 1.

7. The processing machine is a processing machine that performs cyclic processing, which involves repeating a predetermined processing method, as described in claim 1, for the processing state estimation device.

8. The processing machine is a press machine, according to claim 1, for the processing state estimation device.

9. The aforementioned processor, By acquiring the first measurement data and the second measurement data, waveform data showing the measurement results of the change in processing load during each processing cycle by the press machine is obtained. The first measurement data indicates the relative position of the punch of the press machine with respect to a predetermined reference position of the press machine. The second measurement data indicates the processing load, The processing state estimation device according to claim 8.

10. A step of acquiring first measurement data and second measurement data corresponding to the first measurement data, measured during machining by a machining machine, The steps include dividing at least a portion of the first measurement data into a first range and a second range, The steps include determining a first determined value corresponding to the first range and a second determined value corresponding to the second range in the second measurement data, A step of determining a first reference parameter corresponding to the first range and a second reference parameter corresponding to the second range in a plurality of reference data, each corresponding to a plurality of combinations of machining state values ​​that define the machining state of the machining machine, A step of determining the similarity between each of the plurality of reference data and the second measurement data based on the first determined value, the second determined value, the first reference parameter, and the second reference parameter, A step of estimating the processing state based on the determined similarity, including, Method for estimating the state of processing.

Images