Bending forming control device, bending forming control method, and computer program

Abstract

To correct a shape of a bend-molded elongated member so as to be close to a target shape during molding.SOLUTION: A bend-molding control device includes: an acquisition unit for acquiring an actually measured value of a displacement amount of an intermediate part of an elongated member which is positioned between a support tool and a bending tool in a feeding direction; a prediction unit for predicting a prediction value according to a relative position using a prediction model in which a relative position of the bending tool to the support tool and a prediction value of the displacement amount of the elongated member are preliminarily made to correspond to each other; and a correction unit for correcting the prediction model so that a prediction value predicted by the prediction unit and an actually measured value acquired by the acquisition unit are matched at the time of bend-molding of the elongated member.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a bending forming control device, a bending forming control method, and a computer program.

Background Art

[0002] Conventionally, in machine tools for shaping a workpiece to a desired shape, various techniques have been proposed to improve the shaping accuracy. For example, in Patent Document 1, in a manufacturing apparatus for manufacturing a bent member by bending a long member, based on the measurement result of the three-dimensional shape of the portion of the long member where bending is performed, by pre-correcting the three-dimensional shape of the portion to be bent later, a technique for suppressing dimensional changes in the cross section associated with bending is disclosed. Also, in Patent Document 2, in a numerical control device for numerically controlling a machine tool, by inputting an inference data set to an inference unit having a learned inference model obtained by machine learning, a technique for determining the occurrence of chatter vibration that causes deterioration of machining accuracy is disclosed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] Here, the technology disclosed in Patent Document 1 assumes that the degree of bending of the elongated member is constant, and that the cross-section of the elongated member being bent is constant. In other words, the technology disclosed in Patent Document 1 cannot be applied when the degree of bending of the bent portion (the portion being measured) of the elongated member or the cross-section before processing differs from the degree of bending or the cross-section of the portion that is later bent. For this reason, there was room for improvement in the technology that uses the measurement results of the bent elongated member for correction. Furthermore, the technology disclosed in Patent Document 2 determines the occurrence of chatter vibration in a machine tool and does not consider the bending of elongated members at all. In particular, when bending metal materials, the bending angle of the metal material is restored from the time of bending due to springback, so there was room for improvement in bending metal materials to make the angle after restoration the target bending angle.

[0005] The present invention has been made to solve at least some of the above-mentioned problems, and aims to correct the shape of a bent, elongated member during the molding process so that it approaches the desired shape. [Means for solving the problem]

[0006] The present invention has been made to solve at least some of the above-mentioned problems and can be realized in the following forms. Bending forming apparatus for forming elongated members, comprising: a support tool for supporting the elongated member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the elongated member, and pressing against the elongated member to bend the elongated member, the bending forming control device for controlling the bending forming apparatus, comprising: an acquisition unit that acquires measured values ​​of the displacement amount of the intermediate portion of the elongated member located between the support tool and the bending tool in the feeding direction for each first feed amount which is a predetermined feed amount of the elongated member; and a prediction model in which the relative position of the bending tool with respect to the support tool and the predicted value of the displacement amount of the elongated member are pre-associated, according to the relative position A bending forming control device comprising: a prediction unit that predicts the predicted value; a correction unit that, during the bending of the elongated member, corrects the prediction model using the measured value for each first feed amount so that the predicted value predicted by the prediction unit matches the measured value acquired by the acquisition unit; and a movement control unit that, during the bending of the elongated member, controls the movement of the bending tool using the prediction model corrected by the correction unit so that the measured value of the elongated member becomes a preset target value, wherein the movement control unit does not move the bending tool until the elongated member is transported by the first feed amount, and moves the bending tool after the elongated member has been transported by the first feed amount. In addition, the present invention can also be realized in the following forms.

[0007] (1) According to one embodiment of the present invention, a bending forming apparatus for forming an elongated member is provided, comprising: a support tool for supporting the elongated member; and a bending tool positioned at least on the upstream and downstream sides of the support tool in the feeding direction of the elongated member, and bending the elongated member by being pressed against it. The bending forming control device controls an apparatus comprising: an acquisition unit for acquiring measured values ​​of the displacement of an intermediate portion of the elongated member located between the support tool and the bending tool in the feeding direction; a prediction unit for predicting the predicted value according to the relative position, using a prediction model in which the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member are pre-associated; and a correction unit for correcting the prediction model so that the predicted value predicted by the prediction unit matches the measured value acquired by the acquisition unit when the elongated member is being bent.

[0008] In this configuration, if the predicted value predicted using the prediction model differs from the measured value acquired by the acquisition unit, the prediction model is corrected so that the predicted value and the measured value match. Since the measured value is the displacement of the intermediate portion between the support tool and the bending tool in the elongated member, the bending angle can be changed before the intermediate portion of the elongated member is bent by the bending tool by changing the position of the bending tool after acquiring the measured value. In other words, in this configuration, the bending tool is controlled according to the acquired measured value without waiting for the load to be removed from the portion that has passed through the support tool and bending tool and completed bending. This allows for immediate adjustment of the displacement of the elongated member, bringing the shape of the elongated member closer to the target shape during forming. Furthermore, by predicting the deformation state of the intermediate portion and the entire elongated member after bending by the support tool and bending tool, the shape error of the elongated member after forming can be predicted from the target shape to be formed.

[0009] (2) In the bending forming control device according to the above embodiment, the device may further include a movement control unit that, during the bending of the elongated member, controls the movement of the bending tool using the prediction model corrected by the correction unit so that the measured value of the elongated member becomes a preset target value, thereby bending the elongated member. In this configuration, the movement control unit controls the position of the bending tool, which changes the measured value of the elongated member bent by the support tool and the bending tool. Therefore, by controlling the movement of the bending tool by the movement control unit, the measured value of the elongated member after bending can be brought closer to the target value compared to when the prediction model is not corrected.

[0010] (3) In the bending forming control device according to the above embodiment, the acquisition unit acquires the measured value for each first feed amount, which is a predetermined feed amount of the elongated member; the correction unit corrects the prediction model using the measured value for each first feed amount; and the movement control unit does not move the bending tool until the elongated member has been transported by the first feed amount, and may move the bending tool after the elongated member has been transported by the first feed amount. With this configuration, the prediction model is corrected each time the first feed amount is passed. Therefore, compared to a case where the prediction model is corrected in real time, the number of processing steps performed by the correction unit is reduced, and variations in the amount of displacement that occur in the elongated member during the first feed amount can be suppressed. As a result, the error between the shape of the elongated member after bending and the target shape can be suppressed.

[0011] (4) In the bending forming control device according to the above embodiment, the correction unit may correct the prediction model, which is created based on the relationship between the relative position measured in advance and the measured value, so that the prediction value and the measured value match when the elongated member is bent. With this configuration, the predictive model created by pre-measurement is corrected so that the predicted values ​​during bending match the measured values. As a result, the error between the predicted and measured values ​​during bending is small, the amount of correction required for the predictive model during bending is small, and the error between the shape of the elongated member after forming and the target shape can be further suppressed.

[0012] (5) In the bending forming control device according to the above embodiment, the prediction model includes a forming condition model that represents the relationship between the curvature of the intermediate portion calculated from the prediction value and the curvature of the elongated member after bending, and a deformation state model that represents the relationship between the curvature of the intermediate portion calculated from the prediction value and parameters related to the relative position, and the correction unit may, when bending the elongated member, correct the deformation state model so that the prediction value and the measured value match, without correcting the forming condition model. In this configuration, the prediction model includes a forming condition model that represents the relationship between the predicted value and the shape of the elongated member after bending, and a deformation state model that represents the relationship between the predicted value and the relative position. By correcting only the deformation state model without correcting the forming condition model according to the measured value, the number of parameters to be corrected can be reduced while still bringing the shape closer to that of the elongated member after bending represented by the forming condition model.

[0013] Furthermore, the present invention can be realized in various forms, for example, as a bending forming control device, a bending forming apparatus, a bending forming system, a bending forming control method, a bending forming method, a control method for these apparatuses and systems, a computer program executed in these apparatuses and systems, a server device for distributing this computer program, a non-temporary storage medium storing the computer program, and so on. [Brief explanation of the drawing]

[0014] [Figure 1] This is a block diagram of a bending system equipped with a bending control device as one embodiment of the present invention. [Figure 2] This is an explanatory diagram of the molding apparatus. [Figure 3] It is an explanatory diagram of the prediction model. [Figure 4] It is an explanatory diagram of the prediction model. [Figure 5] It is an explanatory diagram of the prediction model. [Figure 6] It is an explanatory diagram of the prediction model. [Figure 7] It is an explanatory diagram of the correction of the prediction model by the correction unit. [Figure 8] It is an explanatory diagram of the correction of the prediction model by the correction unit. [Figure 9] It is an explanatory diagram of the correction of the prediction model by the correction unit. [Figure 10] It is a flowchart of the control method of the molding device. [Figure 11] It is an explanatory diagram of the effect by the molding control device of the present embodiment. [Figure 12] It is an explanatory diagram of the effect by the molding control device of the present embodiment. [Figure 13] It is an explanatory diagram of the effect by the molding control device of the present embodiment. [Figure 14] It is an explanatory diagram of the effect by the molding control device of the present embodiment. [Figure 15] It is an explanatory diagram of the effect by the molding control device of the present embodiment. [Figure 16] It is an explanatory diagram of the effect by the molding control device of the present embodiment. [Figure 17] It is an explanatory diagram of the sensor of the modification example.

Mode for Carrying Out the Invention

[0015] <Embodiment> Figure 1 is a block diagram of a bending system 500 equipped with a bending control device 100 as one embodiment of the present invention. The bending system 500 of this embodiment includes a forming device (bending device) 200 for bending long members such as pipes, and a forming control device (bending control device) 100 for controlling the forming device 200. The forming control device 100 bends the long member using a prediction model that associates the measured value of the displacement of the long member during bending with the predicted value of the displacement predicted from a prediction model. The forming control device 100 corrects the prediction model in real time during the bending of the long member so that the measured value and the predicted value match. As a result, the forming control device 100 can bend the long member during forming so that its shape approaches the target shape during forming.

[0016] Figure 2 is an explanatory diagram of the molding apparatus 200. Figure 2 shows a schematic front view of the molding apparatus 200. The molding apparatus 200 bends a long member LG while feeding it along the feed direction DR1 from left to right in the figure. As shown in Figure 2, the molding apparatus 200 is equipped with a support tool 220 positioned upstream in the feed direction DR1 and a bending tool 210 positioned downstream of the support tool 220 in the feed direction DR1. The Cartesian coordinate system CS shown in Figure 2 corresponds to the Cartesian coordinate system CS shown in Figure 3 and subsequent figures.

[0017] The support tool 220 and the bending tool 210 bend the elongated member LG by pressing them against it. The support tool 220 has upstream rotating rollers RL1 and RL2 positioned to sandwich the elongated member LG in the feed direction DR1, and rotating rollers RL3 and RL4 positioned downstream along the feed direction DR1 at a different position from the rotating rollers RL1 and RL2, and also sandwiching the elongated member LG. The four rotating rollers RL1 to RL4 rotate around an axis parallel to the Y-axis.

[0018] The bending tool 210 consists of rotating rollers RL5 and RL6 positioned to grip the elongated member LG, similar to the rotating rollers RL1 to RL4 of the support tool 220. The bending tool 210 is movable on the ZX plane under the control of the molding control device 100. Specifically, the bending tool 210 changes its relative position to the support tool 220 on the ZX plane. The sensor 140 included in the molding control device 100 shown in Figure 2 will be described later.

[0019] In the state shown in Figure 2, the elongated member LG is bent by three-point support: point P1 in contact with the rotating roller RL2, point P2 in contact with the rotating roller RL3, and point P3 in contact with the rotating roller RL6. In other words, the elongated member LG is bent by three-point bending with point P2, where the downstream rotating roller RL3 of the support tool 220 contacts the elongated member LG, acting as the fulcrum.

[0020] As shown in Figure 1, the molding control device 100 comprises a CPU (Central Processing Unit) 110, an input unit 120, a display unit 130, a sensor 140, and a storage unit 150. The molding control device 100 is composed of, for example, a personal computer. The input unit 120 consists of a keyboard and mouse that receive input from the operator and transmits the input information to the CPU 110. The display unit 130 is a monitor capable of displaying various images and displays various information according to the information transmitted from the CPU 110.

[0021] As shown in Figure 2, the sensor 140 is a displacement sensor that detects the measured value of the displacement of the intermediate portion of the elongated member LG located between the support tool 220 and the bending tool 210 in the feed direction DR1. The sensor 140 in this embodiment detects the measured value at predetermined feed amounts FA.

[0022] The storage unit 150 shown in Figure 1 is composed of a hard disk drive (HDD) and the like. The storage unit 150 stores a prediction model that predicts the amount of bending of the elongated member LG after molding, according to the relative position of the bending tool 210. In the prediction model, the relative position of the bending tool 210 with respect to the support tool 220 is associated with the predicted value of the displacement of the middle portion of the elongated member LG detected by the sensor 140.

[0023] The memory unit 150 includes a first model database (first model DB) 151 for storing the first regression model, a second model database (second model DB) 152 for storing the second regression model, and a third model database (third model DB) 153 for storing the third regression model. Each regression model is created based on the relationship between the relative position of the bending tool 210, which has been measured in advance, and the measured values ​​of the elongated member LG. Details of the first to third regression models as prediction models will be described later.

[0024] The CPU 110 executes a program by loading the program stored in ROM (Read Only Memory), which is not shown, into RAM (Random Access Memory). The CPU 110 functions as a movement control unit 111, a prediction unit 112, and a correction unit 113. The prediction unit 112 uses a prediction model to predict the displacement amount according to the relative position of the bending tool 210 with respect to the support tool 220.

[0025] Figures 3 through 6 are explanatory diagrams of the prediction model. Figure 3 shows an explanatory diagram of the relationship between the curvature of the elongated member LG during molding and the curvature of the elongated member LG after molding. In Figure 3, the central axis OL1 of the elongated member LG during molding is shown by a solid line, and the central axis OL2 of the elongated member LG after molding is shown by a dashed line.

[0026] In the state shown in Figure 3, the change in the relative position of the bending tool 210 with respect to the support tool 220 is represented by the amount of movement D1 of the bending tool 210 in the positive Z-axis direction and the rotation angle θ1 of the bending tool 210 around the Y-axis. The amount of movement D1 and the rotation angle θ1 are associated with the curvature Kunload of the elongated member LG as the target value. To make this association, first, the target curvature Ktarget is calculated from the target value. The curvature Kload during molding is calculated so that the curvature Kunload after molding is the target curvature Ktarget. The shape of the elongated member LG during molding is created from the curvature Kload during molding. From the shape of the created elongated member LG and the geometric positional relationship between the support tool 220 and the bending tool 210, the amount of movement D1 and the rotation angle θ1 are calculated. The first regression model stored in the first model DB151 is a regression model (molding condition model) that represents the relationship between the curvature Kunload after molding and the curvature Kload during molding of the elongated member LG, which is calculated from the displacement D1 and rotation angle θ1.

[0027] Figure 4 shows the change in the curvature Kunload after molding on the vertical axis, with the horizontal axis representing the curvature Kload during molding, for the first regression model. The curvature Kload during molding is determined by controlling the displacement D1 and rotation angle θ1, which represent the relative position of the bending tool 210. From the relationship shown in Figure 4, the curvature Kunload after molding can be predicted from the curvature Kload during molding.

[0028] Figure 5 shows a regression model representing the relationship between the curvature Kload during molding and the inclination A1 shown in Figure 3, as a second regression model. In this embodiment, inclination A1 represents the angle at which the central axis OL2 of the elongated member LG, which is the target shape after molding, is rotated in Figure 3 to become the central axis OL1 of the elongated member LG during molding, in accordance with the support tool 220 and bending tool 210 during actual molding. When changing the central axis OL2 to the central axis OL1, first, the position of the feed direction DR1 (longitudinal position) of the elongated member LG is aligned. Furthermore, the inclination of the tangent to the central axis OL2 on the Z axis passing through the support point P2 is adjusted to be parallel to the X axis. Subsequently, the inclination when changing the central axis OL2 to the central axis OL1 in accordance with the positions of the support tool 220 and bending tool 210 during actual bending is A1. Furthermore, the amount of movement along the Z axis direction generated from the deflection of the elongated member LG during molding with actual three-point support is the indentation amount PP, which will be described later. In Figure 5, as a second regression model, the change in slope A1 is shown on the vertical axis, with the horizontal axis representing the curvature Kload during molding.

[0029] Figure 6 shows a regression model representing the relationship between the curvature Kload during molding and the amount of indentation PP shown in Figure 3, as the third regression model. In Figure 6, the change in the amount of indentation PP is shown on the vertical axis, with the horizontal axis representing the curvature Kload during molding, as the third regression model. Note that the amount of indentation PP and the slope A1 are parameters calculated using the geometry of the three-point bend, and can therefore be rephrased as parameters related to the relative position of the bending tool 210 with respect to the support tool 220. Furthermore, the second regression model shown in Figure 5 and the third regression model shown in Figure 6 correspond to deformation state models. As will be described in detail later, the second and third regression models are corrected during molding according to the measured values ​​of the sensor 140.

[0030] The prediction unit 112 shown in Figure 1 uses a first regression model (Figure 4) to predict the curvature Kload during molding from the target curvature Kunload after molding. The prediction unit 112 uses the predicted curvature Kload during molding, a second regression model (Figure 5), and a third regression model (Figure 6) to predict the inclination A1 and the amount of indentation PP during molding. The movement control unit 111 moves the bending tool 210 by a movement amount D1 and rotates it by a rotation angle θ1 to correspond to the predicted inclination A1 and amount of indentation PP.

[0031] The correction unit 113 corrects the prediction models of the second and third regression models so that the predicted value of the displacement amount during bending of the elongated member LG, as predicted by the prediction unit 112, matches the measured value of the displacement amount obtained by the sensor 140. The correction unit 113 corrects the second and third regression models, but not the first regression model. The movement control unit 111 controls the movement of the bending tool 210 during bending of the elongated member LG using the prediction models corrected by the correction unit 113 so that the measured value of the elongated member LG becomes the target value.

[0032] Figures 7 to 9 are explanatory diagrams illustrating the correction of the prediction model by the correction unit 113. Figure 7 shows a schematic front view of the elongated member LG during molding, where the first feed amount FA1, which is a predetermined feed amount FA of the elongated member LG, is indicated by hatching. Figure 8 shows a schematic front view of the elongated member LG during molding when the feed amount FA is fed from the state shown in Figure 7. In Figure 8, in addition to the first feed amount FA1, the second feed amount FA2, which is a predetermined feed amount FA immediately following the first feed amount FA1, is indicated by different hatching than the first feed amount FA1. Figure 9 shows a schematic front view of the elongated member LG during molding, where a corresponding amount of the elongated member LG has been fed along the feed direction DR1 from the state shown in Figure 8. In Figure 9, in addition to the first feed amount FA1 and the second feed amount FA2, the third feed amount FA3 is indicated by different hatching than the first feed amount FA1 and the second feed amount FA2, respectively.

[0033] In the state shown in Figures 7-9, for example, when the first feed amount FA1 passes through the detection range of the sensor 140, the sensor 140 detects the displacement of the portion of the elongated member LG that is fed first by the first feed amount FA1 as an actual value. Subsequently, when the feed amount FA is fed, the sensor detects the actual value of the portion of the elongated member LG that is fed first by the second feed amount FA2. In other words, the sensor 140 of this embodiment detects the actual value of the elongated member LG each time a predetermined feed amount FA is fed.

[0034] In this embodiment, the correction unit 113 calculates a correction coefficient so that the predicted values ​​predicted using the second and third regression models match the measured values ​​detected by the sensor 140. Specifically, the correction unit 113 corrects the second regression model by multiplying the slope A1 predicted from the second regression model by a correction coefficient C1 so that it matches the measured values. Similarly, the correction unit 113 corrects the third regression model by multiplying the amount of concave PP predicted from the third regression model by a correction coefficient C2 so that it matches the measured values. In order to correct the prediction model according to the measured values ​​of the sensor 140, the correction unit 113 corrects the second and third regression models using the measured values ​​each time the feed amount FA is fed. The correction unit 113 stores the corrected second regression model in the second model DB 152 and the corrected third regression model in the third model DB 153.

[0035] During the bending of the elongated member LG, the movement control unit 111 changes the relative position of the bending tool 210 with respect to the support tool 220 according to the predicted values ​​of the second and third regression models, which are corrected prediction models by the correction unit 113. The correction of the prediction model by the correction unit 113 is performed each time a predetermined feed amount FA is fed. Therefore, once the movement control unit 111 moves the bending tool 210 in accordance with the correction of the prediction model, it does not move the bending tool 210 again until the elongated member LG has been fed by the predetermined feed amount FA. On the other hand, after the elongated member LG has been fed by the predetermined feed amount FA, the movement control unit 111 moves the bending tool 210 again in accordance with the correction of the prediction model.

[0036] Figure 10 is a flowchart of the control method for the molding apparatus 200. In the control flow shown in Figure 10, initial settings are performed first (step S1). In the initial settings, the settings for the elongated member LG to be molded by the molding apparatus 200 and the target shape of the elongated member LG after bending are set via the input unit 120. Once the initial settings are performed, the molding apparatus 200 starts molding the elongated member LG based on the settings (step S2).

[0037] When molding begins, the prediction unit 112 performs a prediction step (step S3) to predict the curvature Kunload during molding using a prediction model so that the elongated member LG takes on the desired shape. The movement control unit 111 changes the relative position of the bending tool 210 so that the curvature Kunload during molding is as predicted in the prediction step. Next, the sensor 140 performs an acquisition step (step S4) to acquire the actual displacement amount of the elongated member LG being molded as it is fed in the feed direction DR1. The sensor 140 detects the displacement amount of the elongated member LG as an actual value each time the feed amount FA is fed.

[0038] The correction unit 113 performs a correction process (step S5) to correct the prediction model so that the predicted value predicted by the prediction unit 112 matches the measured value detected by the sensor 140. The correction unit 113 corrects the second and third regression models respectively using correction coefficients C1 and C2 so that the slope A1 in the second regression model and the amount of concave PP in the third regression model match the measured values.

[0039] The movement control unit 111 controls the movement of the relative position of the bending tool 210 according to the prediction model corrected by the prediction unit 112 (step S6). The sensor 140 detects the actual value and determines whether a predetermined feed amount FA has been fed or not (step S7). If it is determined that the feed amount FA has not been fed (step S7: NO), the sensor 140 waits without detecting the actual value until the feed amount FA is fed.

[0040] In step S7, if it is determined that the feed amount FA has been fed (step S7: YES), it is determined whether the bending of the elongated member LG being formed has been completed (step S8). If it is determined that the bending has not been completed (step S8: NO), the prediction unit 112 performs a prediction process to predict the predicted value using the corrected prediction model (step S3). Subsequently, the processes from step S4 onwards are repeated. In step S8, if it is determined that the bending of the elongated member LG has been completed (step S8: YES), the bending control flow ends.

[0041] Figures 11 to 16 are explanatory diagrams illustrating the effects of the molding control device 100 of this embodiment. Figure 11 shows schematic front views of three target shapes SP1 to SP3 of the elongated member LG after molding. Figure 12 shows the dimensional relationships along each coordinate axis in the target shapes of the elongated member LG. In Figure 12, the horizontal axis represents the position on the X coordinate axis, and the vertical axis represents the position on the Z coordinate axis. In Figure 12, shape SP1 is shown by the solid curve CV1, shape SP2 is shown by the dashed curve CV2, and shape SP3 is shown by the dashed curve CV3.

[0042] Figure 13 shows the curvature of the elongated member LG according to its position along the longitudinal direction, based on the target shapes SP1 to SP3 shown in Figure 12. Note that the curvature shown in Figure 13 is the outer curvature (the larger curvature) of the elongated member LG. In Figure 13, the curvature of shape SP1 is shown by the solid line LN1, the curvature of shape SP2 is shown by the dashed line LN2, and the curvature of shape SP3 is shown by the dashed line LN3.

[0043] Figure 14 shows the change in the relative position of the bending tool 210 according to the feed amount of the elongated member LG in the embodiment and comparative examples 1 and 2. In Figure 14, the relative position of the bending tool 210 when controlled by the molding control device 100 of this embodiment, when the target shape is shape SP1, is shown by the solid curve C11 as the embodiment. Also in Figure 14, the relative position of the bending tool 210 in comparative example 1, where the correction of the prediction model by the correction unit 113 is not performed during molding, is shown by the dashed curve C12 as the embodiment. In comparative example 2, the relative position of the bending tool 210 when the curvature is calculated without using the prediction model itself is shown by the dashed curve C13 as the embodiment. As shown in Figure 14, the relative position of the bending tool 210 differs between the embodiment and comparative examples 1 and 2.

[0044] Figure 15 shows the change in the measured value of sensor 140 in accordance with the feed amount of the elongated member LG in the bending process of shape SP1 in the embodiment and comparative examples 1 and 2. In Figure 15, the measured value of sensor 140 in embodiment 1 is shown by the solid curve C21, the measured value of sensor 140 in comparative example 1 is shown by the dashed curve C22, and the measured value of sensor 140 in comparative example 2 is shown by the dashed curve C23, for the relative position of the bending tool 210 shown in Figure 14. As shown in Figure 14, the relative position of the bending tool 210 is different in the embodiment and comparative examples 1 and 2, so the measured value of sensor 140 will also be different.

[0045] Figure 16 shows the difference in hatching between the Example and Comparative Examples 1 and 2, representing the error between the target shape and the molded shape for each of the shapes SP1 to SP3. As shown in Figure 16, the error in the Example was smaller than that in Comparative Examples 1 and 2 for all of the shapes SP1 to SP3. In other words, the Example was able to perform bending with the smallest error relative to the target shape.

[0046] As described above, in the molding control device 100 of this embodiment, the sensor 140 detects the measured displacement of the intermediate portion of the elongated member LG located between the support tool 220 and the bending tool 210 in the feed direction DR1. The prediction unit 112 uses a prediction model to predict the displacement according to the relative position of the bending tool 210. The correction unit 113 corrects the prediction model during bending of the elongated member LG so that the predicted value for molding predicted by the prediction unit 112 matches the measured value obtained by the sensor 140. Therefore, in this embodiment, if the predicted value predicted using the prediction model differs from the measured value detected by the sensor 140, the prediction model is corrected so that the predicted value and the measured value match. The measured value is the displacement of the intermediate portion of the elongated member LG between the support tool 220 and the bending tool 210. Therefore, by changing the position of the bending tool 210 after detecting the measured value, the bending angle can be changed before the intermediate portion of the elongated member LG is bent by the bending tool 210. In other words, in this embodiment, the bending tool 210 is controlled according to the detected measured value without waiting for the load to be removed from the elongated member LG after it has passed through the support tool 220 and the bending tool 210 and the bending form has been completed. This allows the displacement amount of the elongated member LG to be adjusted immediately, and the shape of the elongated member LG can be brought closer to the target shape during the forming process. Furthermore, by predicting the deformation state of the intermediate portion and the entire elongated member LG after bending form by the support tool 220 and the bending tool 210, the shape error of the elongated member LG after forming can be predicted from the target shape to be formed.

[0047] Furthermore, in this embodiment, the movement control unit 111 controls the movement of the bending tool 210 during the bending of the elongated member LG, using a prediction model corrected by the correction unit 113, so that the measured value of the elongated member LG becomes the target value. As a result, the movement control unit 111 controls the position of the bending tool 210, which changes the measured value of the elongated member LG that is bent by the support tool 220 and the bending tool 210. Therefore, by controlling the movement of the bending tool 210 by the movement control unit 111, the measured value of the elongated member LG after bending can be brought closer to the target value compared to the case where the prediction model is not corrected.

[0048] Furthermore, the sensor 140 in this embodiment detects the actual value of the elongated member LG each time a predetermined feed amount FA is sent. The correction unit 113 corrects the prediction model using the actual value each time the feed amount FA is sent. The movement control unit 111 moves the bending tool 210 again in accordance with the correction of the prediction model after the elongated member LG has been sent by the predetermined feed amount FA, and does not move the bending tool 210 until the elongated member LG has been sent by the predetermined feed amount FA. In other words, in this embodiment, the prediction model is corrected each time a predetermined feed amount FA is sent. Therefore, compared to the case where the prediction model is corrected in real time, the number of processing operations performed by the correction unit 113 is reduced, and variations in the amount of displacement that occur in the molding of the elongated member LG for each feed amount FA can be suppressed. As a result, the error between the shape of the elongated member LG after bending and the target shape can be suppressed.

[0049] Furthermore, the correction unit 113 of this embodiment corrects the prediction model, which is created based on the relationship between the relative position of the bending tool 210 measured in advance and the actual measured value of the elongated member LG, so that the predicted value from the prediction unit 112 matches the actual measured value from the sensor 140. With this configuration, the prediction model created by pre-measurement is corrected so that the predicted value and the actual measured value match during bending. As a result, the error between the predicted value and the actual measured value during bending is small, the amount of correction required for the prediction model during bending is small, and the error between the shape of the elongated member LG after molding and the target shape can be further suppressed.

[0050] Furthermore, the prediction model of this embodiment includes first to third regression models. The first regression model is a regression model (forming condition model) that represents the relationship between the curvature Kunload after forming and the curvature Kload of the elongated member LG during forming, which is calculated from the amount of movement D1 and the rotation angle θ1. The second regression model is a regression model that represents the relationship between the curvature Kload during forming and the inclination A1, which is a parameter related to the relative position of the bending tool 210. The third regression model is a regression model that represents the relationship between the curvature Kload during forming and the amount of indentation PP, which is a parameter related to the relative position of the bending tool 210. In other words, the prediction model of this embodiment has a first regression model that represents the relationship between the predicted value and the shape of the elongated member LG after bending and forming, and a second regression model and a third equipment model that represent the relationship between the predicted value and the relative position. By correcting the second and third regression models without correcting the first regression model according to the measured value, the number of parameters to be corrected can be reduced. Furthermore, in this embodiment, the shape of the elongated member LG after bending and forming can be made closer to the shape represented by the first regression model.

[0051] <Modified examples of embodiments> The present invention is not limited to the embodiments described above, and can be implemented in various forms without departing from its spirit. For example, the following modifications are possible. Furthermore, in the above embodiments, some of the configurations implemented by hardware may be replaced with software, and conversely, some of the configurations implemented by software may be replaced with hardware.

[0052] <Example 1> The molding control device 100 in the above embodiment is just one example, and the configuration of the molding control device 100 and the control of each configuration can be modified within the range in which the correction unit 113 corrects the prediction model so that the predicted value and the measured value match. In the above embodiment, as shown in Figure 2, elongated members LG with the same cross-sectional shape were given as an example of the object to be bent forming, but the elongated members to be processed may be members with different cross-sectional shapes, and may be rods or plates. By predicting predicted values ​​using the prediction model according to different cross-sectional shapes, the molding control device 100 can process elongated members with different cross-sectional shapes.

[0053] The molding apparatus 200 of the above embodiment includes a support tool 220 having two sets of rotating rollers RL1, RL2 and rotating rollers RL3, RL4 arranged upstream along the feed direction DR1, and a movable bending tool 210 arranged downstream. However, the configuration of the support tool 220 and the bending tool 210 is deformable. For example, the support tool 220 may have rotating rollers RL1, RL2 but not RL3, RL4. Alternatively, the bending tool 210 may be arranged upstream along the feed direction DR1, and the support tool 220 may be arranged downstream. The support tool 220 may be arranged to be movable in the ZX plane in Figure 1, similar to the bending tool 210. Furthermore, of the rotating rollers RL5, RL6 of the support tool 220 shown in Figure 1, the bending tool 210 may be composed only of rotating roller RL6, without rotating roller RL5. The support tool 220 and the bending tool 210 can take on well-known configurations that enable bending and molding of elongated members LG without rotating rollers.

[0054] In the above embodiment, the correction unit 113 corrected the prediction model using the measured value detected by the sensor 140 each time a predetermined feed amount FA was fed, but the frequency at which the sensor 140 detects the measured value can be modified. For example, the sensor 140 may detect the curvature of all elongated members LG in the section of the predetermined feed amount FA. Alternatively, the sensor 140 may detect the curvature of the intermediate portion of the elongated member LG in real time, and the correction unit 113 and the movement control unit 111 may form the elongated member LG using the measured value detected in real time and the predicted value. The frequency at which the sensor 140 detects the measured value may change depending on the feed speed of the elongated member LG (rotation speed of the rotating rollers RL1 to RL6) and the curvature as the target value.

[0055] <Modification 2> Figure 17 is an explanatory diagram of a modified sensor 140a. Figure 17 shows a schematic front view of the molding apparatus 200 corresponding to Figure 2 of the above embodiment and the modified sensor 140a. In the modified example shown in Figure 17, the sensor 140a differs in that, in addition to the sensor 140 that detects the amount of displacement in the above embodiment, it is equipped with a load sensor 141 that detects the load applied to the rotating roller RL3 during molding. In the modified example, the configuration and control that differ from the above embodiment will be described, and the description of the same configuration as the above embodiment will be omitted.

[0056] The modified prediction model, like the second and third regression models shown in Figures 5 and 6, includes a fourth regression model that represents the relationship between the curvature Kload during molding and the load detected by the load sensor 141. Therefore, the correction unit of the modified model corrects the fourth regression model during the molding of the elongated member LG so that the predicted value from the prediction unit using the fourth regression model matches the measured value detected by the load sensor 141. Thus, the type and number of prediction models can be modified. In addition, other well-known sensors besides the sensor 140 and load sensor 141 can be used as the acquisition unit for obtaining the displacement amount of the elongated member LG in the middle section.

[0057] <Variation 3> In the above embodiment, the correction unit 113 corrected the second and third regression models without correcting the first regression model as a prediction model during molding. However, the models to be corrected and the correction method are modifiable. For example, only the first regression model may be stored in the storage unit 150 as a prediction model, and the correction unit 113 may correct the first regression model according to the measured values ​​of the sensor 140. Alternatively, the correction unit 113 may correct only the second regression model without correcting the first and third regression models. As a method for correcting the prediction model, a well-known correction method may be used instead of correction using correction coefficients C1 and C2.

[0058] The embodiments of this specification have been described above based on the embodiments and modifications described above. The embodiments described above are for the purpose of facilitating understanding of this specification and do not limit it. This specification may be modified and improved without departing from its spirit and the scope of the claims, and equivalents thereof are included in this specification. Furthermore, any technical features that are not described as essential in this specification may be deleted as appropriate.

[0059] The present invention can also be realized in the following forms. [Application Example 1] A bending forming apparatus for forming a long, rectangular member, comprising: a support tool for supporting the long, rectangular member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the long, rectangular member, and pressed against the long, rectangular member to bend it; and a bending forming control device for controlling the bending forming apparatus, An acquisition unit that acquires measured values ​​of the displacement of the intermediate portion of the elongated member located between the support tool and the bending tool in the feed direction, A prediction unit predicts the predicted value corresponding to the relative position using a prediction model in which the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member are pre-associated. During the bending process of the elongated member, a correction unit corrects the prediction model so that the predicted value predicted by the prediction unit matches the measured value obtained by the acquisition unit. A bending forming control device equipped with the following: [Application Example 2] The bending forming control device described in Application Example 1, further, A bending control device comprising a movement control unit that, during the bending of the elongated member, controls the movement of the bending tool so that the measured value of the elongated member becomes a preset target value using the prediction model corrected by the correction unit, thereby bending the elongated member. [Application Example 3] A bending forming control device as described in Application Example 1 or Application Example 2, The acquisition unit acquires the measured value for each first feed amount, which is a predetermined feed amount of the elongated member. The correction unit corrects the prediction model using the measured values ​​for each of the first feed amounts. The movement control unit does not move the bending tool until the elongated member is transported by the first feed amount, and moves the bending tool after the elongated member has been transported by the first feed amount. [Application Example 4] A bending forming control device according to any one of Application Examples 1 to 3, The correction unit corrects the prediction model, which is created based on the relationship between the relative position measured in advance and the measured value, so that the predicted value and the measured value match when the elongated member is bent. [Application Example 5] A bending forming control device according to any one of Application Examples 1 to 4, The aforementioned prediction model, A molding condition model that represents the relationship between the curvature of the intermediate portion calculated from the predicted value and the curvature of the elongated member after bending, A deformation state model representing the relationship between the curvature of the intermediate portion calculated from the predicted value and the parameters related to the relative position, Includes, The correction unit is a bending forming control device that, when bending the elongated member, corrects the deformation state model so that the predicted value and the measured value match, without correcting the forming condition model. [Application Example 6] A bending forming apparatus for forming a long, rectangular member, comprising: a support tool for supporting the long, rectangular member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the long, rectangular member, and pressed against the long, rectangular member to bend it; and a bending forming control method for controlling the bending forming apparatus, wherein a computer controls the bending forming apparatus, An acquisition step to acquire measured values ​​of the displacement amount of the intermediate portion of the elongated member located between the support tool and the bending tool in the feed direction, A prediction step of predicting the predicted value corresponding to the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member, using a prediction model in which these values ​​are pre-associated; During the bending process of the elongated member, a correction step is performed to correct the prediction model so that the predicted value predicted by the prediction step matches the measured value obtained by the acquisition step. A bending form control method that performs this operation. [Application Example 7] A computer program for controlling a bending forming apparatus for forming an elongated member, comprising: a support tool for supporting the elongated member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the elongated member, and pressing against the elongated member to bend it. The acquisition function acquires the measured displacement of the intermediate portion of the elongated member located between the support tool and the bending tool in the feed direction, A prediction function that predicts the predicted value according to the relative position, using a prediction model in which the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member are pre-associated. During the bending process of the elongated member, a correction function is provided to correct the prediction model so that the predicted value predicted by the prediction process matches the measured value obtained by the acquisition process. A computer program that causes a computer to execute something. [Explanation of symbols]

[0060] 100... Molding control device (bending molding control device) 110…CPU 111...Movement Control Unit 112... Prediction section 113...Correction section 120...Input section 130...Display section 140...Sensor 140a...Sensor of the comparative example 141... Load sensor 150...Storage section 151…Model 1 DB 152…2nd Model DB 153…Third Model DB 200... Molding equipment (bending equipment) 210... Bending tool 220...Support Tools 500…Molding system C1, C2... Correction coefficients CS… Cartesian coordinate system D1…Movement amount DR1... Feed direction FA... Predetermined feed rate FA1...First feed amount FA2...Second feed amount FA3...Third feed amount Kload... Curvature during molding Ktarget... Curvature of the target Kunload... Curvature after molding LG...Long-shaped member PP… amount of indentation RL1~RL6... Rotating rollers θ1... Rotation angle

Claims

1. A bending forming apparatus for forming a long, rectangular member, comprising: a support tool for supporting the long, rectangular member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the long, rectangular member, and pressed against the long, rectangular member to bend it; and a bending forming control device for controlling the bending forming apparatus, An acquisition unit acquires the measured displacement of the intermediate portion of the elongated member located between the support tool and the bending tool in the feed direction, for each first feed amount which is a predetermined feed amount of the elongated member. A prediction unit predicts the predicted value corresponding to the relative position using a prediction model in which the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member are pre-associated. During the bending process of the elongated member, a correction unit corrects the prediction model using the measured values ​​for each of the first feed amounts so that the predicted values ​​predicted by the prediction unit match the measured values ​​obtained by the acquisition unit. A bending control device for bending an elongated member by controlling the movement of a bending tool so that the measured value of the elongated member becomes a preset target value, using the prediction model corrected by the correction unit during the bending of the elongated member, the movement control device further comprising: a movement control device that does not move the bending tool until the elongated member is transported by a first feed amount, and moves the bending tool after the elongated member has been transported by the first feed amount.

2. A bending forming control device according to claim 1, The correction unit corrects the prediction model, which is created based on the relationship between the relative position measured in advance and the measured value, so that the predicted value and the measured value match when the elongated member is bent.

3. A bending forming control device according to claim 2, The aforementioned prediction model, A molding condition model that represents the relationship between the curvature of the intermediate portion calculated from the predicted value and the curvature of the elongated member after bending, A deformation state model representing the relationship between the curvature of the intermediate portion calculated from the predicted value and the parameters related to the relative position, Includes, The correction unit is a bending forming control device that, when bending the elongated member, corrects the deformation state model so that the predicted value and the measured value match, without correcting the forming condition model.

4. A bending forming apparatus for forming a long, rectangular member, comprising: a support tool for supporting the long, rectangular member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the long, rectangular member, and pressed against the long, rectangular member to bend it; and a bending forming control method for controlling the bending forming apparatus, wherein a computer controls the bending forming apparatus, The acquisition step involves obtaining measured values ​​of the displacement of the intermediate portion of the elongated member located between the support tool and the bending tool in the feed direction, for each first feed amount which is a predetermined feed amount of the elongated member. A prediction step of predicting the predicted value corresponding to the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member, using a prediction model in which these values ​​are pre-associated; During the bending process of the elongated member, a correction step is performed to correct the prediction model using the measured values ​​for each of the first feed amounts so that the predicted values ​​predicted by the prediction step match the measured values ​​obtained by the acquisition step. A movement control step for bending the elongated member during the bending of the elongated member, wherein the movement of the bending tool is controlled using the prediction model corrected by the correction step so that the measured value of the elongated member becomes a preset target value, the movement control step comprising: not moving the bending tool until the elongated member is transported by the first feed amount, and moving the bending tool after the elongated member has been transported by the first feed amount; A bending form control method that performs this operation.

5. A computer program for controlling a bending forming apparatus for forming an elongated member, comprising: a support tool for supporting the elongated member; and a bending tool positioned at least on the upstream and downstream side of the support tool in the feeding direction of the elongated member, and pressing against the elongated member to bend it. The acquisition function acquires the measured displacement of the intermediate portion of the elongated member located between the support tool and the bending tool in the feed direction, for each predetermined feed amount of the elongated member, which is a first feed amount. A prediction function that predicts the predicted value according to the relative position, using a prediction model in which the relative position of the bending tool with respect to the support tool and the predicted value of the displacement of the elongated member are pre-associated. During the bending process of the elongated member, a correction function is provided to correct the prediction model using the measured values ​​for each of the first feed amounts so that the predicted values ​​predicted by the prediction function match the measured values ​​obtained by the acquisition function. A computer program that causes a computer to execute a movement control function for bending a long elongated member during the bending of the long elongated member, by controlling the movement of the bending tool so that the measured value of the long elongated member becomes a preset target value, using the prediction model corrected by the correction function, wherein the movement control function does not move the bending tool until the long elongated member is transported by a first feed amount, and moves the bending tool after the long elongated member has been transported by the first feed amount.

Images