Support system for press working, and method for supporting press work

Abstract

To provide a press work assistance system for assisting an operation for selecting the amount of press work.SOLUTION: An assistance system 1 for predicting a work size of a product 5 after press-working a material 59 to be worked includes a learning model for calculating a prediction value of a work size when a work amount is changed by a prescribed change amount or a change amount of the work size on the basis the press-worked amount of the material 59 to be worked or the work size after press-working. The learning model has learned learning data including a press-worked amount actually executed on the material 59 to be worked and the work size after the press-work.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a support system for press working for predicting the dimensions after press working when press working is performed.

Background Art

[0002] Conventionally, press working has been widely used as a processing technology for shaping workpieces such as metal materials (see, for example, Patent Document 1 below). Press working is a processing method in which a workpiece is pressed against a mold to deform it into the shape of the mold. Press working is suitable for mass production of products and parts with the same shape specifications, and can efficiently shape workpieces such as metal materials.

[0003] In press working, variations in the shape and material composition of the workpiece, slight differences in processing conditions, dimensional errors of the mold, etc. may cause variations in the processed dimensions after press working. Therefore, in the case of products that require shape accuracy, it is necessary to repeatedly perform trial processing to select the processing amount of press working so that the processed dimensions fall within a predetermined standard.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, the selection of the processing amount of press working is very difficult, and even a skilled technician needs to perform multiple trial processes, and in fact, it is necessary to interrupt the production of products and parts during the trial process.

[0006] The present invention has been made in view of the above circumstances, and aims to provide a support system for press working that supports the operation of selecting the processing amount of press working.

Means for Solving the Problems

[0007] One aspect of the present invention is a support system for predicting the processing dimensions, which are the dimensions of a product after performing press working on a workpiece, including a learning model for obtaining a predicted value of the processing dimensions or the amount of change in the processing dimensions when the processing amount is changed by a predetermined amount based on the processing amount of the press working performed on the workpiece and the processing dimensions after the press working. The learning model is a model for press working support that learns learning data including the processing amount of the press working actually performed on the workpiece and the processing dimensions after the press working as teacher data.

[0008] One aspect of the present invention is a method for supporting press work by predicting the processing dimensions, which are the dimensions of a product or part after performing press working on a workpiece, using a learning model that learns learning data including the processing amount of the press working actually performed on the workpiece and the processing dimensions after the press working as teacher data, and obtaining a predicted value of the processing dimensions or the amount of change in the processing dimensions when the processing amount is changed by a predetermined amount based on the processing amount of the press working performed on the workpiece and the processing dimensions after the press working.

Advantages of the Invention

[0009] The present invention is a system or method including a learning model for obtaining a predicted value related to the processing dimensions when the processing amount of press working is changed. This learning model is a model that learns teacher data including the processing amount of the press working performed on the workpiece and the processing dimensions after the press working. According to this learning model, it is possible to accurately obtain a predicted value related to the processing dimensions when press working is performed on the workpiece.

[0010] The support system and support method of the present invention are a system or support method with excellent characteristics capable of predicting the processed dimensions and the like obtained by pressing a workpiece, and can support the work of selecting the processing amount of the press working.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Mode for Carrying Out the Invention

[0012] The embodiments of the present invention will be specifically described below using the following examples. (Example 1) This example relates to a support system 1 for predicting the processed dimensions (dimensions after processing) when a workpiece made of a metal material is pressed, and a support method for the press work. This content will be described with reference to FIGS. 1 to 9.

[0013] The support system 1 (Fig. 1) in this example is a system that supports the process of performing press working on a metal plate as a workpiece to produce a product. The processes exemplified in this example are processes realized by processing equipment 1S including an uncoiler 22 that unwinds a workpiece 59 from a coil material 21 around which a strip-shaped metal plate is wound, a precision leveler 23 that corrects warping and the like of the workpiece 59, and a press 3 that performs press working.

[0014] The product 5 (Fig. 2) in this example is a substantially strip-shaped plate-like part. In the product 5, an intermediate part 53 that protrudes with respect to both end parts 51 is formed by bending, and rectangular through-holes 50 are punched in one end part 51 and the intermediate part 53, respectively. Further, a comb-tooth shape is provided by cutting at the other end part 51 (the front end part in the figure). This product 5 is produced by press working with a press 3 (Fig. 1) based on the workpiece 59. A product including a complex shape such as a comb-tooth shape can be said to be a processed product with a high degree of difficulty in press working and difficult to manage processing dimensions.

[0015] In the product 5, a plurality of measurement locations A to E (Fig. 3) for measuring processing dimensions are preset. For the measurement locations A to E, specifications that are the range of target dimensions are set. In the support system 1 of this example, the processing amount of the press working when producing the product 5 is selected so that the processing dimensions of the measurement locations A to E fall within the specifications.

[0016] For bending, for example, a mold (not shown) including a lower mold (die) and an upper mold (punch) is used. According to the mold, the workpiece (see Fig. 1) can be sandwiched and a product 5 with a predetermined shape can be processed. In particular, the mold in this example is a high-functional mold having a mechanism capable of adjusting the processing amount and enabling adjustment of the processing amount at a plurality of adjustment locations.

[0017] Here, the workpiece 59 to be press-worked is a metal plate that has been unrolled from the coil material 21 (Fig. 1) and then straightened for warping and the like by the precision leveler 23. There is a significant difference in the degree of curvature of the unrolled metal plate between the outer circumference and the inner circumference of the coil material 21. Therefore, even after correction by the precision leveler 23, there may be a difference in the degree of warping of the workpiece 59. When the processing amount by press working is the same, a difference will occur in the dimensions (processed dimensions) of the product 5 after press working between the outer circumference and the inner circumference of the coil material 21. Also, even for coil materials 21 of the same specification, differences in workability may occur due to variations in material composition and the like. If there are differences in the workability of the workpiece 59, a difference will occur in the dimensions (processed dimensions) of the product 5 after processing even if the processing amount by press working is the same.

[0018] Therefore, when a new coil material 21 is set in the uncoiler 22 (see Fig. 1), it is necessary to adjust the processing amount by press working each time. Also, as described above, there are differences in warping and the like of the workpiece 59 between the outer circumference and the inner circumference of the coil material 21. Therefore, in addition to the start point of unrolling the workpiece 59 from the coil material 21, it is necessary to measure the dimensions of the product 5 and adjust the processing amount at intermediate points in time. Furthermore, in order to avoid the occurrence of defects in the product 5, it is necessary to measure the dimensions of the product 5 and confirm the processed dimensions at the end point of unrolling the workpiece 59 from the coil material 21.

[0019] The support system 1 in this example is an example of a system that supports the adjustment work of the processing amount performed at the start point and intermediate points in time of unrolling the workpiece 59 from the coil material 21. By using the support system 1, it is possible to improve the efficiency of the work of selecting an appropriate processing amount for press working and increase the production efficiency of the product 5.

[0020] As shown in Fig. 4, the support system 1 includes a press machine 3 having a function of automatically measuring the dimensions (machined dimensions) of a product 5 (see Fig. 3), and a computer device 10 capable of communicating with the press machine 3. The press machine 3 is provided with a control unit 38 for setting the machining amount at each adjustment point in the mold, and a measurement unit 39 for automatically measuring the dimensions (machined dimensions) of the product 5. The measurement unit 39 can output measured data representing the machined dimensions at each measurement point of the product 5.

[0021] Note that the adjustment points of the mold and the measurement points of the product 5 (for example, the points A to E in Fig. 3) may be at the same position or different positions. Also, the adjustment of the machining amount in the mold may be performed by an operator instead of automatic control by the control unit 38. Furthermore, the measurement of the dimensions of the product 5 may be manual measurement instead of automatic measurement by the measurement unit 39. Additionally, the measurement unit 39 may be incorporated into the press machine 3. That is, it is also possible to adopt a press machine having a dimension measurement function.

[0022] The computer device 10 (Fig. 4) is provided with a circuit for executing various arithmetic processes based on machining data representing the machining amounts, etc. at a plurality of adjustment points of the mold, measured data by the press machine 3 (measurement unit 39), etc. The result of the arithmetic process by the computer device 10 is output using a display device such as a liquid crystal display, and can also be printed out by a printer (not shown).

[0023] The general flow of the process in which the support system 1 configured as described above supports the selection operation of the processing amount will be described with reference to FIG. 5. The process in the figure is the process executed by the computer device 10 in FIG. 4. When the press 3 performs press working as trial working, the computer device 10 reads out processing data representing the processing amount and the like of each adjustment part of the mold, and also captures from the press 3 the measured data representing the processing dimensions (processing dimensions by trial working) of a plurality of measurement parts (for example, the parts A to E in FIG. 3) preset for the product 5 (S101). Then, the computer device 10 determines whether the processing dimensions of each measurement part of the product 5 are within a preset standard (range of target dimensions) (S102). In step S101 above, it is also possible for the operator to input the processing amount and the processing dimensions to the computer device 10.

[0024] For a plurality of measurement parts, if each processing dimension by trial working is within the standard (S102: YES), the computer device 10 ends the process as it is and ends the selection operation of the processing amount. In this case, thereafter, the processing amount of the trial working is set, and the mode shifts to producing the product 5, and the press working by the press 3 is repeatedly executed, and the product 5 is produced.

[0025] On the other hand, if the processing dimension by trial working is out of the standard for any of the measurement parts (S102: NO), the computer device 10 executes a process of predicting the dimension change amount (an example of the predicted value of the change amount of the processing dimension) when the processing amount by trial working is changed (S103). The process of predicting the dimension change amount will be described in detail later.

[0026] The computer device 10 selects a recommended processing amount for a plurality of adjustment parts of the mold (S104). The computer device 10 displays, for example, on the display screen of the liquid crystal display, the predicted dimension, which is the predicted value of the processing dimension, together with the recommended processing amount of each adjustment part (S105).

[0027] Next, the content of the process of predicting the dimensional change amount (1) (the process of step S103 in FIG. 5) and the process of selecting the recommended processing amount (2) (the process of step S104 in the same figure) when the processing amount of press working is changed will be described. These processes are all processes executed by the computer device 10 (FIG. 4).

[0028] (1) Process of predicting the dimensional change amount The support system 1 in this example uses the neural network model (NN model) illustrated in FIG. 6 to obtain the dimensional change amount (an example of a predicted value), which is the change amount of the processing dimension when the processing amount of press working as trial processing is changed. The NN model in the same figure is an example of a learning model that outputs the dimensional change amount predicted when the processing amount is changed based on the processing amount and processing dimension (measured data) of the actual press working as trial processing. The input data of the NN model is the processing amount and processing dimension of the trial processing, and the change amount from the processing amount of the trial processing (changed processing amount). The processing amount and the changed processing amount are the processing amounts applied to a plurality of adjustment locations (for example, locations No. 1 to 5) set in the mold. The processing dimension is the measured data of the dimensions of a plurality of measurement locations (for example, locations A to E in FIG. 3) set in the product 5. The output data of the NN model is the dimensional change amount of a plurality of measurement locations (for example, locations A to E in FIG. 3) set in the product 5, that is, the predicted value of the change amount from the processing dimension by the trial processing.

[0029] Note that the NN model used by the support system 1 is a model that has been previously learned using the data of the actually performed press working as teacher data (learning data). The data of the actually performed press working includes the processing amount and processing dimension by the reference press working, the changed processing amount, which is the change amount from the processing amount by the reference press working, and the dimensional change amount when the processing amount is changed.

[0030] In this example, for the learned NN model of FIG. 6, by inputting the processing amount and processing dimensions (measured data) by trial processing, together with the additional processing amounts at each adjustment location (for example, locations No. 1 to 5), the predicted value of the dimensional change amount is obtained. The data of the processing amount by the actual press working as the trial processing is the processing amount set at a plurality of adjustment locations (for example, No. 1 to 5) set in the mold. Also, the processing dimensions (measured data) of the trial processing are the processing dimensions measured at a plurality of measurement locations (for example, locations A to E in FIG. 3.) preset in the product 5. The additional processing amount is a combination of the additional processing amounts at a plurality of adjustment locations of the mold. The combination of the additional processing amounts is basically a combination of all possible additional processing amounts belonging to the search range set at each adjustment location. For example, FIG. 7 illustrates ○, □, ×, and △ as part of the combination of the additional processing amounts at adjustment locations No. 1 to 5.

[0031] Note that as the learning data of the NN model, instead of the above configuration, learning data that includes the processing amount and processing dimensions (measured data) of the actually performed press working but does not include the additional processing amount and the dimensional change amount may be used. When the data (processing amount and processing dimensions) of any press working is used as the reference data, the difference in the processing amount between the data of other press workings can be treated as the additional processing amount and used as part of the learning data. Further, the difference between the processing dimensions related to the reference data and the processing dimensions related to the data of other press workings can be treated as the dimensional change amount and used as part of the learning data. According to such handling, it is only necessary to prepare the processing amount and processing dimensions of the actual press working as the learning data, and the learning of the NN model illustrated in FIG. 6 becomes easy.

[0032] (2) Process of selecting the recommended processing amount In the support system 1 of this example, for a plurality of adjustment locations (for example, locations No. 1 to 5) set in the mold, a search range is set for each, and the amount of change machining is selectively set from within the search range. Then, in this support system 1, the amount of dimensional change (an example of a predicted value) is obtained for each combination of the amounts of change machining set for the plurality of adjustment locations, and the recommended combination of the amounts of change machining is selected as the recommended machining amount. Here, the amount of change machining is the amount of change from the amount of machining by actual press working as trial machining.

[0033] As the search range for each adjustment location (for example, No. 1 to 5), the allowable range of the amount of machining at each adjustment location and the range in which the NN model (Fig. 6) has learned the actual data are set by manual input or automatic input. By setting the search range in this way, it is possible to avoid unnecessary search processing in unrealistic and impossible ranges, unlearned ranges where the teacher data has not been learned, etc., and to shorten the time required for the arithmetic processing. Also, in the unlearned range, it is highly likely that the prediction performance by the NN model cannot be sufficiently ensured. By excluding the unlearned range from the search range, high prediction performance can be ensured.

[0034] Also, in the support system 1 of this example, an allowable range is set for the difference in the amount of machining between different adjustment locations (for example, the adjustment location No. 1 and the adjustment location No. 3, etc.). In the support system 1, it is determined whether or not the combination of the amounts of change machining for different adjustment locations belongs to the above-mentioned allowable range of the difference. If the combination of the amounts of change machining for different adjustment locations belongs to the above-mentioned allowable range of the difference, the process of obtaining the amount of dimensional change is executed. On the other hand, if the combination of the amounts of change machining for different adjustment locations is outside the above-mentioned allowable range of the difference, even if each amount of change machining belongs to the above-mentioned search range, it is excluded from the combination of the amounts of change machining input to the NN model.

[0035] That is, in the configuration of this example, the combination of the machining change amounts set for different adjustment points among the plurality of adjustment points being within the range of the difference in the allowable machining amounts between the different adjustment points is the execution condition for the process of obtaining the dimensional change amount. In the configuration of this example, by excluding combinations of machining amounts that are physically impossible in this way, unnecessary search processing is avoided beforehand.

[0036] The computer device 10 obtains the dimensional change amounts at a plurality of measurement points (points A to E in FIG. 3) for various combinations of the machining change amounts (for example, the combinations of ○, □, ×, and △ in FIG. 7) of a plurality of adjustment points (for example, No. 1 to 5) set in the mold. Then, for each of the plurality of adjustment points, the computer device 10 obtains the predicted dimension after machining by adding the dimensional change amount, which is an example of the predicted value, to the machining dimension by the actual press machining as trial machining.

[0037] The table illustrated in FIG. 8 is a table that illustrates the deviations from the reference value of the predicted dimensions at the measurement points A to E (see FIG. 3) for the combinations of the machining change amounts ○, □, ×, and △ in FIG. 7. As an example of the reference value that is an example of the target dimension, for example, the median value (specification median value) of the specification, which is the allowable range of the dimension, can be adopted. In the figure, the values indicated by the broken lines are the upper limit value and the lower limit value of the specification. Note that at the measurement points A to E, the size of the specification range (the difference between the upper limit value and the lower limit value) is the same.

[0038] When selecting the recommended machining amount, the computer device 10 first excludes combinations that include predicted dimensions that deviate from the specification from among the combinations of the machining change amounts (for example, the combinations of ○, □, ×, and △ in FIG. 8) of each adjustment point (for example, No. 1 to 5). For example, in the case of FIG. 8, the combination of △ that includes a predicted dimension that deviates from the specification is excluded.

[0039] For the remaining combinations, namely the combinations of ○, □, and ×, the computer device 10 calculates the sum of the magnitudes of the differences between the predicted dimensions and the reference values for each measurement location (Fig. 9). Here, the sum of the magnitudes of the differences between the predicted dimensions and the reference values for each measurement location is an example of an index representing the degree of deviation between the predicted dimensions of multiple measurement locations and the reference values of the multiple measurement locations. The computer device 10 selects the combination with the minimum sum of the magnitudes of the differences. For example, in the case of Fig. 9, the sum of the magnitudes of the differences related to the □ combination is the minimum. Therefore, the combination of the modification processing amounts of □ in Fig. 7 is selected. Here, the magnitude of the difference means the absolute value of the difference obtained by subtracting the reference value from the predicted dimension.

[0040] In this way, the computer device 10 selects the combination of modification processing amounts with the minimum sum of the magnitudes of the above differences, and determines the recommended processing amount by adding the modification processing amount related to the selected combination to the processing amount of the actual press working as trial processing. The computer device 10 displays the selected recommended processing amount and modification processing amount on a display screen such as a liquid crystal display, and inputs the data to the press machine 3.

[0041] As described above, according to the support system 1 and the press work support method of this example, it is a useful system that can efficiently select the processing amount during press working. By using this support system 1, when the work material 59 is subjected to press working (trial processing), by inputting the processing amount and processing dimensions (measured data) into the NN model (Fig. 6), it is possible to receive a presentation of the recommended processing amount and the like. The processing dimensions input to the NN model may deviate from the specifications. The support system 1 of this example can select and present a processing amount that meets the specifications based on data such as the processing dimensions actually obtained by press working (trial processing).

[0042] In this example, a configuration is adopted in which the computer device 10 inputs the recommended processing amount and the modification processing amount to the press machine 3, and the processing amount of the mold is automatically adjusted. However, instead of this configuration, a configuration in which the operator adjusts the mold to adjust the processing amount may be adopted.

[0043] In this example, when the processing dimensions deviate from the standard during the actual press working as trial processing, the NN model is used to select the processing amount. Instead of this, even when the actual processing dimensions as trial processing fall within the standard, the NN model may be used to select the recommended processing amount. In this case, it is possible to receive a presentation of a processing amount closer to the best.

[0044] In the configuration of this example, the standard median value, which is the median of the standard range, is used as a reference value (target dimension) for comparing with the predicted dimension. Instead of this, it is also possible to set a value offset from the standard median value as the reference value. For example, if it is possible to grasp the tendency that the processing dimension fluctuates, such as fluctuating to the plus side as production progresses, it is also possible to offset the reference value of the predicted dimension from the standard median value. For example, if it is grasped that the processing dimension gradually increases as the source of the workpiece 59 shifts to the inner peripheral side of the coil material 21, it is advisable to set a value offset to the lower side than the standard median value as the reference value of the predicted dimension. In this case, after setting the processing amount of the press working and starting production, the number of products that can be produced until the processing dimension deviates from the standard can be increased.

[0045] In this example, for each combination of the changed processing amounts at a plurality of adjustment points, the total magnitude of the difference between the reference value and the predicted dimension is obtained. As the difference, instead of the value obtained by subtracting the reference value from the predicted dimension, a value normalized by the magnitude of the standard range may be used. For example, if the difference is 2 mm and the magnitude of the standard range is 10 mm, the value of the normalized difference is 2 mm ÷ 10 mm = 0.2. In this case, it is advisable that a circuit such as a hard disk or ROM provided in the computer device 10 stores the standards for each of the plurality of measurement points, that is, the range of the target dimensions. By normalizing the magnitude of the difference in this way, even if the magnitude of the standard range is different for each measurement point, the recommended processing amount can be appropriately selected. Note that the total sum of the normalized differences is an example of an index representing the degree of deviation between the predicted dimensions at a plurality of measurement points and the reference values at the plurality of measurement points.

[0046] As an index representing the degree of deviation between the predicted dimensions of a plurality of measurement points and the reference values of the plurality of measurement points, instead of the sum of the exemplified differences, it is also possible to adopt the Euclidean distance, the Manhattan distance, etc. Although detailed explanations of these indexes are omitted, they have been well known in the field of machine learning algorithms. For example, in the case of the Euclidean distance shown by the following formula, the larger the Euclidean distance, the greater the degree of deviation.

[0047] [Number] Here, PD(A), STD(A), Rmax(A), and Rmin(A) are the predicted dimension, the reference value, the upper limit value of the standard within the range of the target dimension, and the lower limit value of the same standard of measurement point A, respectively. Equation 1 is an example of an arithmetic expression for obtaining the Euclidean distance d for measurement points A to E.

[0048] In this example, when changing the processing amount of the actual press working as a trial working, an NN model that outputs the amount of change (dimension change amount) from the processing dimension by the trial working as a predicted value is illustrated (see Fig. 6). Instead of this, it is also possible to adopt an NN model that directly outputs the predicted dimension of each measurement point as a predicted value.

[0049] In this example, the search range of the changed processing amount by the NN model is a range included in the allowable range of the processing amount at each adjustment point and also included in the range in which the teacher data is learned. Instead of this, it is also possible to set only one of the conditions that it is the allowable range of the processing amount at each adjustment point and that it is the range in which the teacher data is learned. Also, in the configuration of this example, an allowable range is provided for the difference in the processing amount between different adjustment points, and combinations of changed processing amounts that deviate from this allowable range are excluded. The condition regarding the allowable range of the difference may be omitted. Furthermore, it is also possible to omit conditions such as the allowable range of the processing amount at each adjustment point and the range in which the teacher data is learned, and set only the condition regarding the allowable range of the difference in the processing amount between different adjustment points.

[0050] In this example, when selecting the recommended processing amount, first, combinations including out-of-spec predicted dimensions are excluded from the combinations of the processing amounts for changes at each adjustment point. Excluding combinations including out-of-spec predicted dimensions in this way is not an essential configuration. For example, it is also possible to rank the combinations of the processing amounts for changes according to the sum of the magnitudes of the differences between the predicted dimensions at each measurement point and the reference value (an example of an index representing the degree of deviation), and recommend the top-ranked combinations. It may happen that combinations with predicted dimensions outside the specifications are recommended. In such a case, combinations with predicted dimensions outside the specifications may be excluded separately.

[0051] Also, in this example, the support system 1 for selecting the processing amount of one-time bending (press working) is illustrated. Instead of this, it is also possible to apply the support system 1 of this example to the second and subsequent bending processes in the process of processing a product by bending multiple times. For example, it is also possible to apply the support system 1 of this example to the process of giving a predetermined correction amount (an example of the processing amount) by the second bending process for adjusting the processing dimensions after the bending process for forming the first approximate shape. In this case, the change correction amount, which is the amount of change in the correction amount at each adjustment point, can be efficiently selected using the support system 1. As the input data of the NN model, it is preferable to set the data of the correction amount and the processing dimensions during the second bending (trial processing), and the change correction amount (processing amount for change) for each adjustment point. Regarding this NN model, it is preferable to learn the learning data including the data during the second bending (trial processing) as teacher data.

[0052] As described above, specific examples of the present invention have been described in detail as in the embodiments. However, these specific examples merely disclose an example of the technology included in the claims. Needless to say, the claims should not be construed in a limited manner depending on the configurations, numerical values, etc. of the specific examples. The claims include technologies obtained by variously modifying, changing, or appropriately combining the specific examples using known technologies and the knowledge of those skilled in the art.

Explanation of Reference Numerals

[0053] 1 Support system 10 Computer device 21 Coil material 22 Uncoiler 23 Precision level 3 Press machine 31 Lower die 33 Upper die (punch) 5 Product 59 Workpiece

Claims

1. A support system for predicting the processed dimensions, which are the dimensions of a product or component after performing press working on a workpiece, comprising: a learning model for obtaining a predicted value of the processed dimension or the change amount of the processed dimension when the processing amount is changed by a predetermined change amount based on the processing amount of the press working performed on the workpiece and the processed dimension after the press working; a circuit that, when the change amount is set for a plurality of adjustment locations preset in a die for press working, uses the learning model to obtain the predicted values of a plurality of measurement locations preset in the product; a circuit that stores the target dimensions of each of the plurality of measurement locations and stores the range of the target dimensions set for each measurement location; a circuit for obtaining an index of the degree of deviation between the predicted dimension, which is the processed dimension based on the predicted values of the plurality of measurement locations, and the target dimensions of the plurality of measurement locations; a circuit for selecting a combination of change amounts that results in the smallest degree of deviation among the combinations of change amounts of the plurality of adjustment locations; and the learning model is a model that has learned, as teacher data, learning data including the processing amount of the press working actually performed on the workpiece and the processed dimension after the press working; the circuit for obtaining an index representing the degree of deviation is configured to, for each of the plurality of measurement locations, obtain the range of the target dimension from the circuit that stores the range of the target dimension, normalize the difference between the target dimension and the predicted dimension by the magnitude of the range of the target dimension, and obtain the index based on the normalized differences of the respective measurement locations. A support system for press working.

2. A support system for predicting the processed dimensions, which are the dimensions of a product or component after performing press working on a workpiece, comprising: a learning model for obtaining a predicted value of the processed dimension or the change amount of the processed dimension when the processing amount is changed by a predetermined change amount based on the processing amount of the press working performed on the workpiece and the processed dimension after the press working; a circuit that, when the change amount is set for a plurality of adjustment locations preset in a die for press working, uses the learning model to obtain the predicted values of a plurality of measurement locations preset in the product; a circuit for setting a search range that is the range of the change amount; means for setting an allowable range of the difference in the processing amount between different adjustment locations among the plurality of adjustment locations. The learning model is a model that learns, as teacher data, learning data including the processing amount of the press working actually performed on the workpiece and the processed dimensions after the press working. The circuit that sets the search range sets the search range so as to be included in at least one of the allowable range of the processing amount applicable to the workpiece and the range in which the learning model has learned the teacher data. The circuit that obtains the predicted value obtains the predicted value for the change amount belonging to the search range, while Among the combinations of change amounts set for different adjustment points among the plurality of adjustment points, for the combination of change amounts in which the difference in the processing amount between the different adjustment points is outside the allowable range, even if it is a combination of change amounts belonging to the search range, it is configured to exclude it from the combinations for obtaining the predicted value. A support system for press working.

3. A support system for predicting the processed dimensions of a product or part after performing press working on a workpiece, A learning model for obtaining a predicted value of the processed dimensions or the amount of change in the processed dimensions when the processing amount is changed by a predetermined amount of change based on the processing amount of the press working performed on the workpiece and the processed dimensions after the press working, A circuit that uses the learning model to obtain the predicted values of a plurality of measurement points preset in the product when the change amount is set for a plurality of adjustment points preset in the die for press working, A circuit that stores the standard that is the allowable range of the dimensions of each measurement point among the plurality of measurement points, A circuit that obtains an index of the degree of deviation between the predicted dimensions that are the processed dimensions based on the predicted values of the plurality of measurement points and the target dimensions of the plurality of measurement points, Including a circuit that selects a combination of change amounts in which the degree of deviation is the smallest among the combinations of change amounts of the plurality of adjustment points. The learning model is a model that learns, as teacher data, learning data including the processing amount of the press working actually performed on the workpiece and the processed dimensions after the press working. The workpiece is unwound from a coil material around which a strip-shaped metal plate is wound, and the degree of curvature changes as the unwinding position moves from the outer circumference to the inner circumference of the coil material in response to the unwinding from the coil material, and the processed dimensions change as the degree of curvature changes. It is a material. In the circuit for obtaining the index of the degree of deviation, a value offset from the median value of the standard is set as the target dimension so that a combination of the change amounts that can increase the number of press working times until the processing dimension deviates from the standard can be selected according to the change in the processing dimension accompanying the change in the degree of bending of the workpiece after starting the press working by unwinding the workpiece from the coil material, a support system for press working.

4. The support system for press working according to any one of claims 1 to 3, wherein the workpiece is made of a metal material.

5. A method for supporting press work by predicting a processing dimension which is a dimension of a product or a part after performing press working on a workpiece, Using a model learned with learning data including the processing amount of the press working actually performed on the workpiece and the processing dimension after the press working as teacher data, and based on the processing amount of the press working performed on the workpiece and the processing dimension after the press working, obtaining a predicted value of the processing dimension or the change amount of the processing dimension when the processing amount is changed by a predetermined change amount, a process of obtaining the predicted values of a plurality of measurement locations preset in a die for press working when the change amount is set for the plurality of adjustment locations; A process of obtaining an index of the degree of deviation between a predicted dimension which is a processing dimension based on the predicted values of the plurality of measurement locations and the target dimensions of the plurality of measurement locations; A process of selecting a combination of change amounts in which the degree of deviation is minimized from among combinations of change amounts of the plurality of adjustment locations; A process of storing the target dimension of each measurement location among the plurality of measurement locations and storing the range of the target dimension set for each measurement location, When obtaining an index representing the degree of deviation, for each measurement location among the plurality of measurement locations, normalizing the difference between the target dimension and the predicted dimension according to the size of the range of the target dimension stored in advance; Based on the normalized differences of the respective measurement locations, obtaining an index representing the degree of deviation between the predicted dimensions of the plurality of measurement locations and the target dimensions of the plurality of measurement locations, a method for supporting press work.

6. A method for supporting press work by predicting a processing dimension which is a dimension of a product or a part after performing press working on a workpiece, A model that learns learning data including the amount of press working actually applied to a workpiece and the processed dimensions after the press working as teacher data, and based on the amount of press working applied to the workpiece and the processed dimensions after the press working, a learning model that obtains a predicted value of the processed dimensions or the amount of change in the processed dimensions when the amount of working is changed by a predetermined amount of change is used, and when the amount of change is set for a plurality of adjustment locations preset in a die for press working, a process of obtaining the predicted values of a plurality of measurement locations preset in a product, and a process of setting a search range so as to be included in at least one of the allowable range of the amount of working applicable to the workpiece and the range in which the learning model has learned learning data including the amount of press working actually applied to the workpiece and the processed dimensions after the press working as teacher data, and while obtaining the predicted values for the amounts of change belonging to the search range, A method for assisting press work, in which, among combinations of amounts of change set for different adjustment locations among the plurality of adjustment locations, a combination of amounts of change in which the difference in the amount of working between the different adjustment locations is outside the allowable range is excluded from the combinations of amounts of change belonging to the search range even if they are combinations of amounts of change belonging to the search range for which the predicted values are obtained.

7. A method for assisting press work by predicting the processed dimensions, which are the dimensions of a product or a part after press working is performed on a workpiece, wherein the workpiece is unwound from a coil material around which a strip-shaped metal plate is wound, and the degree of bending changes as the unwinding position moves from the outer circumference to the inner circumference of the coil material in response to the unwinding from the coil material, and a change occurs in the processed dimensions as the degree of bending changes, and the workpiece is a material in which A model that learns learning data including the amount of press working actually applied to a workpiece and the processed dimensions after the press working as teacher data, and based on the amount of press working applied to the workpiece and the processed dimensions after the press working, a learning model that obtains a predicted value of the processed dimensions or the amount of change in the processed dimensions when the amount of working is changed by a predetermined amount of change is used, and when the amount of change is set for a plurality of adjustment locations preset in a die for press working, a process of obtaining the predicted values of a plurality of measurement locations preset in a product, and a process of storing a standard range, which is the allowable range of the dimensions of each measurement location among the plurality of measurement locations, and A process for obtaining an index of the degree of deviation between a predicted dimension, which is a processed dimension based on predicted values of the plurality of measurement points, and target dimensions of the plurality of measurement points; A process for selecting a combination of change amounts having the smallest degree of deviation among combinations of change amounts of the plurality of adjustment points, the process including: In the process for obtaining the index of the degree of deviation, a value offset from the median of the standard range is set as the target dimension so that a combination of change amounts that can increase the number of press working times until the processed dimension deviates from the standard can be selected according to the change in the processed dimension accompanying the change in the degree of curvature of the workpiece after starting the press working by unwinding the workpiece from the coil material. A method for supporting press work.

8. The method for supporting press work according to any one of Claims 5 to 7, wherein the workpiece is made of a metal material.

Images