KR20240078963A - Press forming method - Google Patents

Abstract

The present invention relates to a press forming method capable of drawing and forming a metal sheet, comprising: a seating step of seating a metal sheet to be drawn and formed between a first mold and a second mold of a forming mold; the first mold and A forming step in which the second mold is molded and the punch unit formed in the first mold presses at least a portion of one surface of the molding area of the metal sheet to form drawing, and after the first mold and the second mold are opened, drawing is performed. A removal step of removing the molded product after the molding has been completed from the mold, wherein in the molding step, the first mold and the second mold are joined together to draw and mold the metal sheet, The applied load value is applied as time series data, a mold load pattern image that is an image of the time series data is calculated, and the mold load pattern image is compared with a previously stored reference pattern image to determine whether there is an abnormality in the drawing molding.

Description

Press forming method {Press forming method}

The present invention relates to a press forming method, and more specifically, to a press forming method capable of drawing forming a metal sheet.

Press processing is a type of plastic processing that can be molded into products with a predetermined shape by cutting or changing the shape of a plate-shaped material using a press molding device, for example, shearing and punching. It can be processed into various types of materials such as trimming, bending, burring, curling and drawing.

In this press molding device, raw materials such as a plate are provided between the upper and lower molds of a mold mounted inside, and then the upper mold is brought into close contact with the lower mold and then pressed to mold the product. Due to the characteristics of the press molding equipment, once the mass production conditions are set, the press molding equipment continues to operate under the same conditions. However, due to material deviations, process deviations, or wear of the molding mold that occur in real time during mass production, mass production is performed under mass production conditions. Quality dispersion is occurring in manufactured products.

However, with this conventional press molding method, it is impossible to check changes or abnormalities such as material deviation, process deviation, or wear of the mold in real time during the mass production process, so defects must be confirmed through visual inspection of the final product. There was a problem. In addition, there was a problem in that it was difficult to identify the cause of defects when defects occurred due to defect determination based on confirmation of the final product, resulting in a significant consumption of time and cost to re-establish mass production conditions.

The present invention is intended to solve various problems including the problems described above, and aims to provide a press forming method that can detect molding defects in real time even during the mass production process and feedback control the blank holding force of the blank holder part. . However, these tasks are illustrative and do not limit the scope of the present invention.

According to one embodiment of the present invention, a press forming method is provided. The press forming method includes a seating step of seating a metal sheet to be drawn between a first mold and a second mold of a forming mold; A molding step in which the first mold and the second mold are joined together so that a punch portion formed in the first mold presses at least a portion of one surface of the molding area of the metal sheet to draw molding; And after the first mold and the second mold are opened, a take-out step of taking out the molded product on which drawing molding has been completed from the mold, wherein in the molding step, the first mold and the second mold are molded. In addition, when drawing and forming the metal plate, the load value applied to the molding mold is received from the load sensor unit as time series data, a mold load pattern image imaging the time series data is calculated, and the mold load pattern image and the previously stored By comparing the reference pattern image, it is possible to determine whether there is an abnormality in the drawing molding.

According to one embodiment of the present invention, the forming step includes a load signal detection step of outputting the change in the load value applied over time from the load sensor unit as the time series data during drawing forming of the metal sheet. ; A pattern image calculation step of calculating the mold load pattern image by imaging the load value applied as the time series data into a shaded texture image; and an abnormality detection step of comparing the mold load pattern image and the reference pattern image to determine whether drawing molding is abnormal.

According to an embodiment of the present invention, the abnormality detection step calculates and stores in advance a pattern image when in a normal state, a pattern image when a molding is defective, and a pattern image when a process is abnormal as the reference pattern image. , When drawing forming the metal sheet, the mold load pattern image calculated in real time in the pattern image calculation step is compared with the reference pattern image to determine whether there is a molding defect of the metal sheet being drawn and formed or a process abnormality of the forming mold. It can be determined.

According to one embodiment of the present invention, the abnormality detection step is performed by collecting the pattern image when in a normal state, the pattern image when there is a molding defect, and the pattern image when there is a process abnormality through an artificial neural network (ANN). Through the learned discrimination algorithm, it is possible to determine whether there is a molding defect of the metal sheet or a process abnormality of the molding mold from the mold load pattern image.

According to one embodiment of the present invention, the pattern image calculation step is to image the time series data for the load value into the texture image of shading by using a Recurrence Plot (RP) algorithm to produce the mold load pattern image. can be calculated.

According to one embodiment of the present invention, in the forming step, the forming mold is installed on both sides of the punch unit and the punch unit for drawing forming by pressing at least a portion of one surface of the forming area of the metal sheet to form the punch. The first mold including a blank holder part that presses and holds one surface of a holding area on both sides of the forming area of the metal sheet when drawing and forming the metal sheet; The second mold is provided opposite to the first mold and includes a die portion that supports the other surface of the holding area of the metal sheet when the punch portion draws and forms the metal sheet; and at least one of the load sensor units installed in the first mold or the second mold to measure the load applied to the mold including the first mold and the second mold. there is.

According to one embodiment of the present invention, in the forming step, the load sensor unit is coupled to the punch part of the first mold near the blank holder part or near the die part of the second mold in the form of a bolt. It may be a piezo bolt that measures the load applied to the mold.

According to one embodiment of the present invention, in the molding step, the position value of the second mold moving up and down toward the first mold can be received from a linear sensor unit installed on the first mold or the second mold. .

According to one embodiment of the present invention, after completion of drawing forming of the metal sheet, during drawing forming of another metal sheet, drawing is performed in the forming mold using a correction function calculated based on the mold load pattern image calculated in the forming step. During molding, a blank holding force correction step of correcting the blank holding force of the blank holder portion holding the metal sheet may be further included.

According to one embodiment of the present invention, in the blank holding force correction step, the time series data for the load value is calculated by dividing the length of the vertical or horizontal width of the shaded area in the mold load pattern image imaged as a shaded texture image. A correction function calculation step of calculating the correction function as a reference; and a correction step of correcting the blank holding force by substituting the correction function calculated in the correction function calculation step into the blank holding force.

According to an embodiment of the present invention, in the correction function calculation step, the correction function is [Formula 1] (α: correction function, A: compensation coefficient, Max. vertical line length: maximum length of the vertical width of the shaded area, Avg. vertical line length: average length of the vertical width of the shaded area) there is.

According to an embodiment of the present invention, in the correction step, the blank holding force is [Formula 2] It can be corrected by (BHF _i : correction blank holding force, α _i-1 : correction function of the previous step, BHF _i-1 : blank holding force of the previous step).

According to one embodiment of the present invention, in the blank holding force correction step, when a forming defect occurs in the drawing forming of the other metal sheet to which the corrected blank holding force corrected in the correction step is applied, the correction function calculation step is performed. It may further include a correction function reduction step of reducing the correction function to a 1/n ratio.

According to an embodiment of the present invention made as described above, the signal of the mold load measured in time series is imaged through a monitoring system for detecting defects in the press molding process, and molding defects are detected in real time based on the image, and the image Depending on the pattern type, defect factors can be determined simultaneously. In addition, by applying an algorithm that can feedback control the blank holding force of the blank holder according to the molding quality determined by the imaged signal, the molding quality can be improved by determining the optimal pressing force of the metal sheet through feedback control.

Accordingly, immediate response measures are possible through real-time detection of molding defects and process abnormalities in the mass production press molding process, and the productivity of the press molding process is improved by applying a feedback algorithm for appropriate blank holding force control according to molding quality. It is possible to implement a press forming method that can have the effect of improving and minimizing cost loss. Of course, the scope of the present invention is not limited by this effect.

1 is a flow chart sequentially showing a press forming method according to an embodiment of the present invention.
FIGS. 2 to 4 are cross-sectional views schematically showing the configuration of a press molding apparatus in which the press molding method of FIG. 1 is implemented.
Figure 5 is an image schematically showing the pattern image calculation step of the press forming method of Figure 1.
Figure 6 is a flow chart sequentially showing a press forming method according to another embodiment of the present invention.
FIG. 7 is a cross-sectional view schematically showing the configuration of a press molding apparatus in which the press molding method of FIG. 6 is implemented.
FIG. 8 is an image schematically showing the blank holding force correction step of the press forming method of FIG. 6.
FIG. 9 is a flowchart schematically showing the correction function reduction step of the press forming method of FIG. 6.

Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the attached drawings.

The embodiments of the present invention are provided to more completely explain the present invention to those skilled in the art, and the following examples may be modified into various other forms, and the scope of the present invention is as follows. It is not limited to the examples. Rather, these embodiments are provided to make the present disclosure more faithful and complete and to fully convey the spirit of the present invention to those skilled in the art. Additionally, the thickness and size of each layer in the drawings are exaggerated for convenience and clarity of explanation.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will now be described with reference to drawings that schematically show ideal embodiments of the present invention. In the drawings, variations of the depicted shape may be expected, for example, depending on manufacturing technology and/or tolerances. Accordingly, embodiments of the present invention should not be construed as being limited to the specific shape of the area shown in this specification, but should include, for example, changes in shape resulting from manufacturing.

Figure 1 is a flow chart sequentially showing a press forming method according to an embodiment of the present invention, and Figures 2 to 4 are cross-sectional views schematically showing the configuration of the press forming apparatus 100 in which the press forming method of Figure 1 is implemented. , FIG. 5 is an image schematically showing the pattern image calculation step (S220) of the press forming method of FIG. 1.

First, as shown in FIG. 1, the press forming method according to an embodiment of the present invention may largely include a seating step (S100), a forming step (S200), and an extraction step (S300).

For example, as shown in FIGS. 1 and 2, in the seating step (S100), the metal sheet (S) to be drawn is placed in the first mold (10) and the second mold (20) of the forming mold (M). ), and in the forming step (S200), the first mold 10 and the second mold 20 are joined together so that the punch portion 11 formed in the first mold 10 is used to form the metal sheet S. At least a portion of one side of the area is pressed to perform drawing molding, and then, in the extraction step (S300), after the first mold 10 and the second mold 20 are opened, the molded product on which the drawing molding is completed is placed in a molding mold (M). ) can be taken out from.

To describe in more detail the press forming apparatus 100 in which this press forming method is implemented, as shown in FIG. 2, the press forming apparatus 100 largely includes a first mold 10 and a second mold. It may include a mold 20, a load sensor unit 30, a linear sensor unit 40, a discrimination control unit 50, and a main body 60 in which they are installed.

As shown in FIG. 2, the press molding device 100 attaches a molding mold (M) to the fixing plate 61 of the main body 60 and the press ram 62 that moves up and down and reciprocates toward the fixing plate 61. It can be formed into a structure that presses or hits the plate (S) seated between the first mold (10) and the second mold (20) of the forming mold (M) by attaching it.

More specifically, the press forming device 100 can be classified into crank press, knuckle press, friction press, screw press, rack press, link press, cap press, etc., depending on the structure for raising and lowering the press ram 62. , a crank press such as the press molding device 100 of the present invention installs a driving part on the upper part of the main body 60 and connects it with the press ram 62 using the rotation axis of the driving part as a crankshaft, so that the main body 60 As the press ram 62 rises and lowers, the plate (S) is moved between the first mold (10) of the forming mold (M) installed on the fixing plate (61) and the second mold (20) installed on the press ram (62). It can be drawn and formed by pressing. In this embodiment, the press forming device 100 is driven by a crank press structure as an example, but it is not necessarily limited to this and the various driving methods described above that can draw and form the sheet material S by the forming mold M are used. It can be applied.

To describe in more detail the structure of the forming mold M that is mounted on the press forming device 100 and capable of drawing and forming the plate S, as shown in FIG. 2, the forming mold M is, It is fixed to the fixing plate 61 of the main body 60 and is installed on both sides of the punch unit 11 and the punch unit 11 to draw and form by pressing at least a portion of one side of the forming area of the plate S. (11) A first mold 10 including a blank holder portion 12 that presses and holds one surface of the holding area on both sides of the forming area of the plate S when drawing and forming the plate S, and the first mold 10 It is fixed to the press ram 62 so as to face the mold 10, and includes a die part 21 that supports the other side of the holding area of the sheet S when the punch part 11 draws and forms the sheet S. It may include a second mold 20.

In this way, the second mold 20 is fixed to the press ram 62 that moves up and down and reciprocates toward the fixing plate 61 on which the first mold 10 is fixed, and the up and down reciprocation of the press ram 62 Depending on the movement, it may descend toward the first mold 10 and be joined to the first mold 10, or, conversely, may rise and be separated from the first mold 10. At this time, in this embodiment, the first mold 10 is fixed to the fixing plate 61 of the main body 60 as the lower mold of the forming mold M, and the second mold 20 is fixed to the press ram 62 as the upper mold. Although it is fixed as an example, it is not necessarily limited to this, and on the contrary, the first mold 10 is fixed to the press ram 62 as an upper mold, and the second mold 20 is fixed to the fixing plate 61 as a lower mold. It may be possible.

In addition, when the first mold 10 and the second mold 20 of the mold M are joined, a load sensor unit 30 is installed on the mold M so that the mold load can be patterned according to time series. It can be.

At least one load sensor unit 30 is installed in the first mold 10 or the second mold 20, and is attached to the mold M including the first mold 10 and the second mold 20. The applied load can be measured.

More specifically, as shown in FIG. 2, the load sensor unit 30 can be coupled near the punch unit 11 or blank holder unit 12 of the first mold 10 in the form of a bolt. there is.

For example, as shown in FIG. 3, the load sensor unit 30 may be coupled in the form of a bolt to the punch plate 11b supporting the punch 11a with the punch 11a of the punch unit 11 as the center. there is. At this time, at least one load sensor unit 30 may be installed on the punch plate 11b, and when a plurality of load sensor units 30 are installed, they may be arranged radially around the punch 11a to measure the mold load for each area.

More specifically, as shown in FIG. 4, the load sensor unit 30 may be a piezo bolt in the form of a bolt having a head 31 and a body 32 with threads formed in at least some sections. there is. At this time, the punch plate 11b to which the load sensor unit 30 is coupled is composed of an upper plate (P1) and a lower plate (P2), and the body 32 of the load sensor unit 30 passes through the upper plate (P1). A through hole (H1) is formed, and a screw hole (H2) in which a female thread is formed so that the load sensor unit 30 can be screwed can be formed at a position corresponding to the through hole (H1) of the lower plate (P2). .

Accordingly, the load sensor unit 30 coupled to the upper plate (P1) and lower plate (P2) of the vertically stacked punch plate (11b) presses the upper surface of the upper plate (P1) with the lower surface of the head portion (31). In this form, the mold load generated when the first mold 10 and the second mold 20 are joined is directly transmitted to the load sensor unit 30 through the punch plate 11b, Through this, the load applied to the mold (M) can be measured.

This load sensor unit 30 is installed in the first mold 10 as an example, but is not necessarily limited thereto. Although not shown, it is coupled near the die portion 21 of the second mold 20 to form a mold. The load applied to (M) can also be measured.

In addition, as shown in FIG. 2, a linear sensor unit 40 is also installed on one side of the mold M, so that the mold fit between the first mold 10 and the second mold 20 of the mold M The mold load may be monitored in real time, and the measured mold load may be patterned according to the distance at which the second mold 20 descends toward the first mold 10.

The mold load value patterned according to the distance can be used by the load sensor unit 30 to verify or correct the mold load value patterned according to time series.

In addition, as shown in FIG. 2, the determination control unit 50 is electrically connected to the load sensor unit 30 and the first mold 10 and the second mold 20 are joined to draw and form the plate material S. At this time, the load value applied to the mold M can be received from the load sensor unit 30 in real time.

Accordingly, in the press forming method, in the forming step (S200), when the first mold 10 and the second mold 20 are joined together to form the sheet material S, the forming mold is applied from the load sensor unit 30. The load value applied to (M) is applied in real time, the change in the load value is applied as time series data (LS in FIG. 5) patterned over time, and the time series data (LS) is imaged. By calculating a load pattern image (RP in FIG. 5) and comparing the mold load pattern image (RP) with a previously stored reference pattern image, it is possible to determine whether there is an abnormality in the drawing molding.

More specifically, as shown in FIGS. 1 and 5, the forming step (S200) is performed by measuring the load value applied over time from the load sensor unit 30 during drawing forming of the metal plate S. A load signal detection step (S210) for outputting changes as the time series data (LS), and the load value received as the time series data (LS) is imaged as a shaded texture image to produce a mold load pattern image (RP). It may include a pattern image calculation step (S220) of calculating and an abnormality detection step (S240) of comparing the mold load pattern image (RP) and the reference pattern image to determine whether there is an abnormality in the drawing molding.

For example, in the pattern image calculation step (S220), the time series data (LS) for the load value is imaged into the shaded texture image using a recurrence plot (RP) algorithm, thereby producing a mold load pattern image (RP). can be calculated.

The recurrence plot algorithm is a visualization tool that aims to explore m-dimensional phase space trajectories through a two-dimensional representation of the regression of data values, and the main idea is to indicate which trajectories return to the previous state. there is. Here, recurrence may be the time when the trajectory returns to a previously visited location.

In this way, through the pattern image calculation step (S220), the time series data (LS) for the load value is imaged into the texture image of shading by the Recurrence Plot (RP) algorithm, thereby performing an abnormality detection step to be described later. In (S240), by applying an artificial neural network (ANN) such as a convolutional neural network (CNN), a deep learning model that shows very high performance in the image recognition field, to time series data (LS), The accuracy of discrimination in the detection step (S240) can be further increased.

For example, in the abnormality detection step (S240), the pattern image in the normal state, the pattern image in the case of molding defects such as fractures or wrinkles, and the pattern image in the case of process abnormalities are converted into the reference pattern image. Calculated and stored in advance, and when drawing and forming a metal sheet (S), the mold load pattern image (RP) calculated in real time in the pattern image calculation step (S220) is compared with the reference pattern image, and the metal being drawn and formed is It is possible to determine whether there is a molding defect in the plate (S) or a process abnormality in the mold (M).

At this time, in the abnormality detection step (S240), as described above, the pattern image when in a normal state, the pattern image when the molding is defective, and the pattern image when the process is abnormal are processed through an artificial neural network (ANN). Through the learned discrimination algorithm, it is possible to determine whether there is a molding defect in the metal plate (S) or a process abnormality in the molding mold (M) from the mold load pattern image (RP).

As such, the press molding method according to an embodiment of the present invention monitors for defect detection of the press molding process, which consists of a load signal detection step (S210), a pattern image calculation step (S220), and an abnormality detection step (S240). By equipping the system, molding defects can be detected in real time by imaging the signal of the mold load measured during the molding step (S200), and defect factors can be determined at the same time.

Therefore, according to the press molding method according to an embodiment of the present invention, the signal of the mold load measured in time series is imaged through a monitoring system for detecting defects in the press molding process, and molding defects are detected in real time based on the image. , defect factors can be determined simultaneously depending on the pattern type of the image.

Therefore, immediate response measures can be taken through real-time detection of molding defects and process abnormalities in the mass production press molding process, making it possible to implement a press molding method that can improve the productivity of the press molding process and minimize cost losses.

Figure 6 is a flow chart sequentially showing a press forming method according to another embodiment of the present invention, Figure 7 is a cross-sectional view schematically showing the configuration of the press forming apparatus 200 in which the press forming method of Figure 6 is implemented, and Figure 8 is an image schematically showing the blank holding force correction step (S400) of the press forming method of FIG. 6, and FIG. 9 is a flowchart schematically showing the correction function reduction step (S430) of the press forming method of FIG. 6.

As shown in Figure 6, in the press forming method according to another embodiment of the present invention, after completing the drawing forming of the metal sheet (S), when drawing forming another metal sheet, the mold load calculated in the forming step (S200) A blank holding force correction step (S400) in which the blank holding force of the blank holder portion 12 that holds the metal plate S is corrected during drawing forming in the forming mold M using a correction function calculated based on the pattern image RP. It may further include.

Here, the seating step (S100), the forming step (S200), and the taking out step (S300) may be the same as the press forming method according to an embodiment of the present invention described above. Therefore, detailed description is omitted.

As shown in FIG. 7, this blank holding force correction step (S400) is performed in the feedback control system 70 electrically connected to the discrimination control unit 50, and the blank holding force corrected according to the calculated correction function is converted into a holding force. It can be input to the press mold device 200 through the control unit 80.

More specifically, as shown in FIGS. 6 and 8, in the blank holding force correction step (S400), the time series data (LS) for the load value is obtained from the mold load pattern image (RP) imaged as a shaded texture image. , a correction function calculation step (S410) of calculating the correction function based on the vertical line or horizontal width of the shaded area, and the correction function calculated in the correction function calculation step (S410) is applied to the blank. It may include a correction step (S420) of correcting the blank holding force by substituting the holding force.

For example, in the correction function calculation step (S410), the correction function may be calculated using the following [Equation 1].

[Formula 1]

α: correction function

A: Compensation coefficient

Max. vertical line length: maximum length of the vertical width of the shaded area

Avg. vertical line length: average length of the vertical width of the shaded area

Additionally, in the correction step (S420), the blank holding force can be corrected by the following [Equation 2].

[Formula 2]

BHF _i : Compensated blank holding force

α _i-1 : correction function from previous step

BHF _i-1 : Blank holding force of previous stage

Accordingly, in the blank holding force correction step (S400), after completing the drawing forming of the metal sheet (S), when drawing forming another metal sheet, the calculation is based on the mold load pattern image (RP) calculated in the forming step (S200). By applying an algorithm that can feedback control the blank holding force of the blank holder unit 12 using the correction function, the molding quality can be improved by determining the optimal pressing force of the metal plate S.

In addition, as shown in FIG. 9, the blank holding force correction step (S400) is a correction function calculation step when forming defects occur in the drawing forming of the other metal sheet to which the corrected blank holding force corrected in the correction step (S420) is applied. It may further include a correction function reduction step (S430) of reducing the correction function calculated in (S410) to a 1/n ratio.

For example, after press forming by applying the corrected blank holding force, as in steps (a) and (b) of FIG. 9, formability can be evaluated as in step (c). Subsequently, as in step (d) of FIG. 9, if it is determined that no molding defect has occurred in the formability evaluation, as in step (e), the correction function calculated in the correction function calculation step (S410) is continued to be used as is. The blank holding force is corrected, and if it is determined that a molding defect has occurred in the formability evaluation, the correction function calculated in the correction function calculation step (S410) is reduced to a ratio of 1/n, as in step (f). The blank holding force can be corrected. This process of reducing the correction function in step (f) of FIG. 9 to a ratio of 1/n can be performed until it is determined that no molding defect has occurred in step (d), and the reduction ratio 1/n is, Considering that the smaller the ratio is set, the adjustment precision of the correction function increases, but the calculation process may increase, and as the ratio is set larger, the adjustment precision of the correction function decreases, but the calculation process may decrease, so the operator can arbitrarily You can set it.

Therefore, according to the press molding method according to another embodiment of the present invention, the signal of the mold load measured in time series is imaged through a monitoring system for detecting defects in the press molding process, and molding defects are detected in real time based on the image. , defect factors can be determined simultaneously depending on the pattern type of the image. In addition, by applying an algorithm that can feedback control the blank holding force of the blank holder according to the molding quality determined by the imaged signal, the molding quality can be improved by determining the optimal pressing force of the metal sheet through feedback control.

Accordingly, immediate response measures are possible through real-time detection of molding defects and process abnormalities in the mass production press molding process, and the productivity of the press molding process is improved by applying a feedback algorithm for appropriate blank holding force control according to molding quality. A press forming method that can have the effect of improving and minimizing cost loss can be implemented.

The present invention has been described with reference to the embodiments shown in the drawings, but these are merely exemplary, and those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true scope of technical protection of the present invention should be determined by the technical spirit of the attached patent claims.

10: first mold
11: Punch unit
11a: punch
11b: Punch plate
12: Blank holder part
20: second mold
21: Daibu
30: Load sensor unit
31: head
32: body
40: Linear sensor unit
50: Discrimination control unit
60: body
61: fixing plate
62: press ram
70: Feedback control system
80: holding force control unit
S: metal plate
M: forming mold
100: Press forming device

Claims (13)

Drawing: A seating step of seating the metal sheet to be formed between the first mold and the second mold of the forming mold;
A molding step in which the first mold and the second mold are joined together so that a punch portion formed in the first mold presses at least a portion of one surface of the molding area of the metal sheet to draw molding; and
After the first mold and the second mold are opened, a take-out step of taking out the molded product on which drawing molding has been completed from the mold,
In the forming step,
When drawing and forming the metal sheet by combining the first mold and the second mold, the load value applied to the forming mold is received as time series data from the load sensor unit, and a mold load pattern image is generated by imaging the time series data. A press forming method that determines whether drawing forming is abnormal by calculating and comparing the mold load pattern image with a previously stored reference pattern image. According to claim 1,
The forming step is,
When drawing and forming the metal sheet, a load signal detection step of outputting the change in the load value applied over time from the load sensor unit as the time series data;
A pattern image calculation step of calculating the mold load pattern image by imaging the load value applied as the time series data into a shaded texture image; and
An abnormality detection step of comparing the mold load pattern image and the reference pattern image to determine whether there is an abnormality in drawing molding;
Including, press molding method. According to claim 2,
In the abnormality detection step,
A pattern image when in a normal state, a pattern image when a molding is defective, and a pattern image when a process is abnormal are calculated and stored in advance as the reference pattern image, and when drawing and forming the metal sheet, the pattern image calculation step is performed. A press forming method that compares the mold load pattern image calculated in real time with the reference pattern image to determine whether there is a molding defect in the metal sheet being drawn and formed or a process abnormality in the forming mold. According to claim 3,
In the abnormality detection step,
The pattern image in a normal state, the pattern image in a molding defect, and the pattern image in a process abnormality are obtained from the mold load pattern image through a discrimination algorithm learned through an artificial neural network (ANN). A press forming method for determining whether there is a defect in the forming of a sheet or an abnormal process of the forming mold. According to claim 2,
In the pattern image calculation step,
A press forming method for calculating the mold load pattern image by imaging the time series data for the load value into the texture image of shading using a Recurrence Plot (RP) algorithm. According to claim 1,
In the forming step, the forming mold is,
The punch unit presses and forms at least a portion of one surface of the forming area of the metal sheet and is installed on both sides of the punch unit, so that the punch part holds both sides of the forming area of the metal sheet when drawing forming the metal sheet. The first mold including a blank holder portion that presses and holds one surface of the area;
The second mold is provided opposite to the first mold and includes a die portion that supports the other surface of the holding area of the metal sheet when the punch portion draws and forms the metal sheet; and
The load sensor unit, at least one of which is installed in the first mold or the second mold, measures the load value applied to the molding mold including the first mold and the second mold;
Including, press molding method. According to claim 6,
In the forming step, the load sensor unit,
A piezo bolt in which the punch part of the first mold is coupled to the vicinity of the blank holder part or the die part of the second mold in the form of a bolt to measure the load value applied to the forming mold. ), press forming method. According to claim 6,
In the forming step,
A press molding method in which the position value of the second mold moving up and down toward the first mold is applied from a linear sensor unit installed on the first mold or the second mold. According to claim 1,
After completion of drawing forming of the metal sheet, during drawing forming of another metal sheet, holding the metal sheet during drawing forming in the forming mold with a correction function calculated based on the mold load pattern image calculated in the forming step. A blank holding force correction step of correcting the blank holding force of the blank holder portion;
Further comprising a press forming method. According to clause 9,
The blank holding force correction step is,
A correction function calculation step of calculating the correction function based on the length of the vertical or horizontal width of the shaded area in the mold load pattern image in which the time series data for the load value is imaged as a shaded texture image; and
A correction step of correcting the blank holding force by substituting the correction function calculated in the correction function calculation step into the blank holding force;
Including, press molding method. According to claim 10,
In the correction function calculation step,
The press forming method wherein the correction function is calculated by the following [Equation 1].
[Formula 1]

α: correction function
A: Compensation coefficient
Max. vertical line length: maximum length of the vertical width of the shaded area
Avg. vertical line length: average length of the vertical width of the shaded area According to claim 10,
In the correction step,
A press forming method in which the blank holding force is corrected by the following [Equation 2].
[Formula 2]

BHF _i : Compensated blank holding force
α _i-1 : correction function from previous step
BHF _i-1 : Blank holding force of previous stage According to claim 10,
The blank holding force correction step is,
If forming defects occur in drawing forming of the other metal sheet to which the corrected blank holding force corrected in the correction step is applied, a correction function reduction step of reducing the correction function calculated in the correction function calculation step to a ratio of 1/n;
Further comprising a press forming method.

Images