JP2019126823A - Press manufacturing condition collection system and sensor plate for press making machine - Google Patents

Abstract

To estimate generation of defectiveness event caused by deterioration of die set to a press making machine in high speed and high accuracy, and to suppress generation of defective moldings in minimum.SOLUTION: A press manufacturing condition collection system comprises: a sensor plate which is arranged to face to a die on a slide of a press making machine and multiple pressure sensors placed with spaces on a surface at right angles for a direction receiving load; a data analyzer generating data point which is time series of detection level every pressure sensor, and estimating defectiveness event of the die on the basis of the data point; and a data collection device collecting the estimated results provided by the data analyzer.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a press manufacturing condition collecting system for estimating failure events of a press molding machine, collecting estimation results and the like, and a sensor plate for a press molding machine used therefor.

  The forming quality by the press forming machine gradually decreases due to deterioration such as wear of parts such as a punch and a die in the mold along with repeated press processing. If pressing is repeated without replacing the deteriorated parts at an appropriate timing, many defective products will be generated, resulting in an increase in waste costs.

  It has been conventionally practiced to detect the state of the press molding machine by various sensors and collect and display the detected data. As an example, hydraulic pressure pressure and ram (slide) position data while the press molding machine is operating are automatically collected at predetermined sampling times using sensors such as position sensors and pressure sensors, and the collected data is displayed on the screen. There are known data collection devices that can be displayed as waveforms or can be digitized and saved as files. In addition, the data collection device compares the values detected by various sensors with a predetermined alarm setting range, and determines the occurrence of defective products when the detected value deviates from the alarm setting range. (See Patent Document 1).

JP-A-2004-164635 (Paragraphs [0020] to [0024])

  However, detection of a partial failure event that occurs due to deterioration of a mold, such as punch breakage or punching two, for example, is generally performed by visual inspection or the like. Therefore, there is a possibility that a large number of defective molded articles may be produced after the detection of the occurrence of the defect. In particular, in recent years, the molding cycle of a press molding machine is very short (about 0.5 to 2 seconds), and the number of defective molded articles tends to increase as all rods become defective molded articles. In addition, since it takes a long time to know the defective part of a mold after the occurrence of a defective product is known, it takes a long time to stop production for mold repair, which reduces productivity. I was invited. Furthermore, the frequent occurrence of these molding defects promotes excessive wear of mold parts such as punches, and also has a significant effect on the reduction of mold life. Therefore, in order to improve productivity, it is necessary to take accurate preventive measures by quickly detecting the sign that a defective molding will occur, rather than dealing with it after a defective molding is detected. Yes.

  In addition, there is also a problem that it is practically difficult to cope with mounting a new sensor or detection device to an existing mold in consideration of the man-hour for remodeling the mold and the remodeling cost. Furthermore, if a large number of sensors are attached to the mold and the state of the mold is detected for each part, the amount of data to be processed becomes enormous, making real-time data processing difficult and detecting the occurrence of defects. However, there was a risk of delay.

  In view of the circumstances as described above, the object of the present invention is to enable high-speed and high-accuracy estimation of the occurrence of a defective event due to deterioration of a mold set in a press molding machine, and to generate a defective molding product. It is an object of the present invention to provide a press manufacturing condition collecting system and a sensor plate for a press molding machine which can contribute to minimizing the pressure.

  In order to achieve the above object, a press manufacturing condition collecting system according to an aspect of the present invention is disposed on a slide of a press molding machine so as to face a mold and separated from each other in a plane orthogonal to a load receiving direction. A sensor plate on which a plurality of pressure sensors are disposed, and waveform data which is a time series of detection values of each pressure sensor, and data analysis for estimating a defect event of the mold based on the waveform data The apparatus comprises: an apparatus; and a data collection apparatus for collecting the estimation result obtained by the data analysis apparatus.

  According to this press manufacturing condition collection system, the occurrence of defective events due to deterioration of the mold set in the press molding machine can be accelerated by estimating the defective events of the mold of the press molding machine by the data analysis device. And, it is possible to estimate with high accuracy, and it becomes possible to minimize the occurrence of molding defects.

In the press manufacturing condition collection system according to one aspect of the present invention, the data analysis device estimates the failure event for each part of the mold based on the waveform data for each pressure sensor. Good.
Thereby, since the defective event for every site | part of a metal mold | die can be estimated, the time cost required for the components replacement | exchange of a metal mold | die can be reduced.

  In the press manufacturing condition collection system according to an aspect of the present invention, the data analysis apparatus supplies a command for stopping press molding to the press molding machine when the defective event of the mold is estimated. This makes it possible to more reliably minimize the occurrence of molding defects.

  In the press manufacturing condition collection system according to one aspect of the present invention, the data analysis device transmits the generated waveform data for each of the pressure sensors to the data collection device, and the data collection device receives the data analysis device from the data analysis device. The pressure distribution of the mold may be displayed on a display unit in real time based on the transmitted waveform data of each pressure sensor.

  In the press manufacturing condition collecting system according to one aspect of the present invention, the data collecting device accumulates the waveform data of the collected pressure sensors, and the data collecting device collects the waveform data of the accumulated pressure sensors based on the accumulated waveform data. The sign of the failure event for each part of the mold may be estimated.

  The data acquisition device may calculate the coordinate position at which the failure event has occurred based on the waveform data of the plurality of pressure sensors collected.

  In the press manufacturing condition collection system according to an aspect of the present invention, the data analysis device estimates a defective event for each part of the mold by analyzing the waveform data for each pressure sensor using an AI function. It may be.

In addition, the sensor plate for a press forming machine according to another aspect of the present invention is capable of being opposed to the mold on the slide of the press forming machine, and separated from each other in a plane orthogonal to the load receiving direction. A plurality of pressure sensors, and an output terminal for outputting a detection signal for each pressure sensor to the outside.
The sensor plate for a press molding machine is mounted on the slide of the press molding machine so as to face the mold, and a plurality of pressure sensors are disposed apart from each other in a plane orthogonal to the load receiving direction. Therefore, no matter what mold is used, it is possible to detect the load for each part received by the mold.

  According to the present invention, it is possible to estimate the occurrence of a defective event caused by deterioration of a mold set in a press molding machine at high speed and with high accuracy, and to minimize the occurrence of defective molding.

It is a block diagram showing composition of press manufacture condition collection system 100 of one embodiment concerning the present invention. It is sectional drawing which shows the structure of the press molding machine 10 with which the sensor plate 2 used for the press manufacturing condition collection system 100 of FIG. 1 was mounted. 2 is a cross-sectional view of a sensor plate 2. FIG. 5 is a plan view showing the arrangement of a plurality of pressure sensors 2 b in a sensor plate 2 from the direction of receiving a load. 2 is a block diagram showing a configuration of a data analysis device 3. FIG. It is a figure which shows the example of the normal waveform data 51 obtained from the pressure sensor 2b. It is a figure which shows the example of waveform data 71 of a shear load at the time of generating "two sheets" of material by scrape rising compared with usual waveform data 51. It is a figure which shows the example of waveform data 81 of a shear load when "punch breakage" has occurred compared with waveform data 51 at the time of normal. It is a figure which shows the relationship between the position of the several pressure sensor 2b in the sensor plate 2, and the punching position by the punch which "2 sheet | seat removal" generate | occur | produced. It is a figure which shows the example of a display of the pressure distribution (die-cutting force distribution) of the whole metal mold | die. It is the figure which visualized the waveform data of a certain pressure sensor 2b accumulated over a certain period in the storage device of the data acquisition device 4. FIG. 7 is a diagram for explaining a method of calculating a defective portion based on a plurality of waveform data by the data acquisition device 4; It is a figure explaining the estimation method of the defective event of the press molding machine.

Hereinafter, embodiments according to the present invention will be described.
FIG. 1 is a block diagram showing a configuration of a press manufacturing condition collection system 100 according to an embodiment of the present invention.

  This press manufacturing condition collection system 100 is generated from the sensor plate 2 mounted on the side of the slide 13 of the press molding machine 10 facing the mold (upper mold 11) and the output of each pressure sensor 2b of the sensor plate 2. Analysis of waveform data to estimate defective events for each part of the mold, and results of estimation of defective events for each part of the mold by the data analysis apparatus 3 and collection of waveform data for each pressure sensor 2b And a data collection device 4 that performs, for example, accumulation, display, and determination of a sign of a failure event for each part of the mold based on analysis of accumulated waveform data.

Next, each configuration will be described.
(Sensor plate 2)
FIG. 2 is a cross-sectional view showing the configuration of the press molding machine 10 to which the sensor plate 2 is mounted.
Reference numeral 11 denotes an upper mold of the mold, 12 denotes a lower mold of the mold, and 13 denotes a slide 13 of the press molding machine 10. The sensor plate 2 is mounted on the lower surface of the slide 13 of the press molding machine 10 in a state of being sandwiched between the upper surface of the upper mold 11 (the upper surface of the backing plate 11a). The sensor plate 2 includes a plate body 2a and a plurality of pressure sensors 2b embedded in the plate body 2a. The plurality of pressure sensors 2b are arranged on the plate body 2a so as to be separated from each other in a plane orthogonal to the direction in which the load is received. As the pressure sensor 2b, a highly sensitive semiconductor sensor based on a piezo effect is used. The present invention is not limited to the type of pressure sensor, but is required to be highly sensitive and capable of pressure detection with a cycle sufficiently short with respect to the cycle of driving the slide 13 (pressing cycle). Ru.

3 is a cross-sectional view of the sensor plate 2, and FIG. 4 is a plan view showing the arrangement of the plurality of pressure sensors 2b on the sensor plate 2 from the direction of receiving a load.
As shown in these figures, in the sensor plate 2, a plurality of pressure sensors 2b are directed against the load receiving direction (Z direction) so that the stress distribution of the entire mold (upper mold 11) can be obtained. In the orthogonal plane (X / Y-axis plane), they are dispersedly arranged in the respective axial directions. For example, the sensor plate 2 shown in FIG. 4 has a total of 24 pressure sensors 2b, six in the X-axis direction and four in the Y-axis direction. The present invention is not limited to the number of pressure sensors 2b of the sensor plate 2, but more pressure sensors 2b may be more densely arranged if higher detection resolution is required.

  In addition, the sensor plate 2 is provided with an amplifier 2c that amplifies the output of the pressure sensor 2b. The output of each pressure sensor 2b is amplified by the corresponding amplifier 2c, and can be output to the outside (data analyzer 3) through the terminal 2d for external connection. In FIG. 4, symbols P 1, P 2, P 3 and P 4 indicate punching positions of the work by the punch of the upper mold 11.

  The sensor plate 2 is not a dedicated one manufactured according to the mold set in the present press molding machine 10, but can be used for molds of various configurations. Accordingly, in the present embodiment, the plurality of pressure sensors 2b disposed on the sensor plate 2 may be arranged on the surface where the sensor plate 2 receives a load, for example, with a predetermined interval in each axial direction. It is not always necessary to arrange them at a predetermined interval.

(Data analysis device 3)
FIG. 5 is a block diagram showing the configuration of the data analysis device 3.
As shown in the figure, the data analysis device 3 includes an amplifier 2c that amplifies the output of each pressure sensor 2b, an A / D conversion unit 31 that performs A / D conversion and data conversion of the output of the amplifier 2c, and A / D conversion. An analysis AI unit 32 that analyzes the output of each D conversion unit 31 using the AI function and a data transmission unit 33 are included. The A / D conversion unit 31 converts the output of the pressure sensor 2b from an analog signal to a digital signal, and further converts it into digital data having a predetermined number of bits, for example, 8 bits.

  The A / D conversion unit 31 performs A / D conversion at a sufficiently short cycle with respect to the drive cycle (press working cycle) of the slide 13, and outputs waveform data that is a time series of detection values for each pressure sensor 2b. .

  The analysis AI unit 32 analyzes the characteristics of the waveform data obtained from the A / D conversion unit 31 by using the AI (Artificial Intelligence) function, so that “two pieces” that may occur in press molding, A failure event such as “punch breakage” is generated as an estimation result of the failure event based on the waveform data for each pressure sensor 2 b. The analysis AI unit 32 analyzes the waveform data as image data by using the AI function, for example, and can control the GPU (Graphics Processing Unit) that can analyze the characteristics of the waveform data and the GPU control. It is comprised by the GPU control part etc. to perform. In addition to a general neural network, deep learning or the like can be adopted as the AI.

  The data transmission unit 33 packetizes the estimation result of the defective event obtained by each analysis AI unit 32 together with the waveform data to be analyzed, and transmits the packet to the data collection device 4 through the network device such as the hub 5 and the network 6. . The network 6 may be a LAN (Local Area Network) or a WAN (Wide Area Network) such as the Internet. Further, the network communication may be wireless communication or wired communication.

(Data acquisition device 4)
Returning to FIG. 1, the data collection device 4 is connected to be able to communicate with the data analysis device 3 through the network 6 and the hub 5. The data collection device 4 may be, for example, a computer such as a central processing unit (CPU), a memory such as a random access memory (RAM), a storage device such as a hard disk drive (HDD), or a display unit such as a liquid crystal display (LCD) monitor. It can be configured using the following hardware: The data collection device 4 may be configured by a general personal computer.

  The data collection device 4 stores the estimation result of the defective event based on the waveform data for each pressure sensor 2b received from the data analysis device 3 through the network 6 and the waveform data to be analyzed in a storage device in association with each other.

The CPU of the data collection device 4 can perform, for example, the following processing based on the data collected from the data analysis device 3.
1. Display of defective events for each part of the mold estimated by the data analysis device 3.
2. Display of pressure distribution etc. of the whole die using collected waveform data.
3. Predictive judgment of failure events for each part of the mold.
Details of these processes will be described later.

  Next, the operation of each part in the press manufacturing condition collecting system 100 of the present embodiment will be described.

[Defect event estimation based on waveform data by data analysis device 3]
First, failure event estimation based on waveform data by the data analysis device 3 will be described.
The data analyzer 3 performs A / D conversion on the output of the pressure sensor 2b at a sufficiently short cycle with respect to the drive cycle (press cycle) of the slide 13 to obtain data of a predetermined number of bits. As a result, waveform data that is a time series of pressure data for each short cycle is obtained, the analysis is performed in the analysis AI unit 32, and various problems occurred in press molding due to problems such as mold deterioration. Bad events are estimated. In order to estimate various failure events from waveform data in the analysis AI unit 32, each analysis AI unit 32 appears in the output of the pressure sensor 2b when the failure event occurs for each of various types of failure events. What has been subjected to machine learning using various learning data including waveform data that has a tendency or actually appears is used.

Here, an example of how a defect event of a mold appears in waveform data is illustrated.
When the material is punched out by the punch, the surface of the punch and the debris are in a vacuum state, so that the debris adheres to the surface of the punch and is sent together with the material to induce "two-out", and the punch (especially at the corner) Causes "burrs" due to accelerated wear. The defect which collectively named these is called "sump up". These influences also lead to damage to the mold such as punch life reduction, punch breakage, and damage to the press molding machine 10. Defective events such as the above “two-sheet removal” and “punch breakage” appear as waveform changes in waveform data which is time-series data of data detected in a cycle sufficiently shorter than the press forming cycle by the pressure sensor 2b. .

  FIG. 6 is a diagram showing an example of normal (normal) waveform data 51 obtained from one pressure sensor 2b. Here, the horizontal axis is the rotation angle of the main gear of the crank mechanism that drives the slide 13, and the vertical axis is the detection signal level of the pressure sensor 2b. In this example, from time t1, the punch starts to penetrate the material (work), and the shear load gradually increases. The shear load reaches a peak at time t2 immediately before the shearing starts. When the shearing starts, the shearing load decreases due to a breakthrough phenomenon in which the elastic deformation energy due to the compressive stress stored in the mold or the like is released at once, and the shearing is completed. . When a failure event such as “dual sheet removal” or “punch breakage” occurs, waveform data reflecting the failure event is generated.

  First, the case where “two sheets” of material has occurred will be described. The “two-sheet removal” is generated, for example, when the burrs of the removal debris are adsorbed to the tip end face of the punch due to the wear of the punch and dropped and placed on the die. FIG. 7 is a view showing the waveform data 71 of the shear load in the case where the “two-sheet removal” of the material occurs, in comparison with the waveform data 51 at the normal time. In the waveform data 71 when the “two sheets” of the material is generated, the peak shear load is about twice as large as the waveform data 51 at the normal time, and the generation time of the shear load is about twice as long. Specific changes such as spreading (not shown) appear.

  Next, a case where “punch breakage” occurs will be described. FIG. 8 is a diagram showing an example of the waveform data 81 in the case where the “punch breakage” occurs, in comparison with the waveform data 51 at the normal time. The waveform data 81 in the case where the “punch breakage” has occurred, for example, the shear load finely vibrates up and down during the period in which the punch is in contact with the material (workpiece), and is clear as the waveform data 51 in the normal state. There are features such as no peaks appearing.

  As described above, defective events such as “two punches” and “punched break” are caused by the waveform data of the pressure sensor 2 b closest to the problematic punch among the plurality of pressure sensors 2 b in the sensor plate 2. Appears prominently. Therefore, the data analysis device 3 can estimate the occurrence of the failure event and the type thereof for each part of the mold.

[Forced stop of press forming triggered by defect event]
When the data analysis device 3 estimates a predetermined failure event based on waveform data, the data analysis device 3 outputs a command for forcibly stopping press forming which is continuously executed to the controller 15 of the press forming machine 10 Do. For example, when occurrence of “punch breakage” is estimated by any of the analysis AI units 32 of the data analysis device 3, a forced stop instruction is supplied to the controller 15 of the press forming machine 10 so as to immediately stop press forming. Do.

  As described above, each analysis AI unit 32 of the data analysis apparatus 3 estimates a failure event with respect to the waveform data of the corresponding individual pressure sensor 2b, so that the occurrence of the failure event can be detected with higher speed and accuracy. . For this reason, the operation of the press molding machine 10 can be forcibly stopped in a short elapsed time from the occurrence of the defect event, and the number of defective molding products can be minimized.

Next, the operation of the data acquisition device 4 will be described.
(Display of failure events for each part of the mold)
The CPU of the data collection device 4 can cause the display unit to display the estimation result of the failure event for each part of the mold collected from the data analysis device 3 in real time. Under the present circumstances, the presumed result of a defect event may be displayed with the information which shows the site | part of the metal mold | die specified from the position of the pressure sensor 2b. For example, as shown in FIG. 9, when “two punching” occurs in the punched portion of P4, the pressure sensor 2b in which the influence of the defective event appears most in the waveform data is the sensor at the position (x4, y1). It is. In this case, at least a defect event of "two out" is estimated by the analysis AI unit 32 corresponding to the pressure sensor 2b, and a portion of the mold specified by the position (x4, y1) of the pressure sensor 2b Is displayed to the user through the display unit that "two out" has occurred.

(Display of pressure distribution (mold force distribution) of the whole mold)
The CPU of the data acquisition device 4 collects the respective waveform data output from each analysis AI unit 32 of the data analysis device 3 and displays the values of the same time of each of the waveform data on the display according to, for example, a three-dimensional graph. It is possible to display it as the pressure distribution (die cutting force distribution) of the entire mold. FIG. 10 shows that the output of the pressure sensor 2b at the position (x4, y1) is significantly increased by the occurrence of “two sheets removed” at the punching portion by the punch P4 shown in FIG. As a result, the user can identify the occurrence and part of the defective event “extract two sheets” at a glance.

(Predicting the occurrence of a bad event)
From the data analysis device 3 to the data collection device 4, not only the estimation result of the defective event but also the analyzed waveform data itself is transmitted. The CPU of the data collection device 4 accumulates the waveform data of each pressure sensor 2b transmitted from the data analysis device 3 in the storage device.

  FIG. 11 is a diagram visualizing waveform data of a certain pressure sensor 2 b accumulated in the storage device of the data acquisition device 4. Here, the horizontal axis is the cumulative number of detections by the pressure sensor, and the vertical axis is the detection signal level. As described above, as the cumulative number of times of detection by the pressure sensor increases, that is, as the cumulative number of press forming increases, deterioration due to wear of the punch proceeds, and the moving average value 111 of the detection signal level gradually increases. The CPU of the data collection device 4 indicates that the evaluation value such as the moving average value 111 of the detection signal level has exceeded the threshold value 113. Is determined to be present, and this is presented to the user through a display unit or the like.

  The sign determination of the failure event is individually performed on the waveform data of all the pressure sensors 2b. In this way, it is possible to perform a prognostic determination of a failure event for each part of the mold.

  Alternatively, the CPU of the data collection device 4 may determine that the mold has reached the end of its life from the evaluation value such as the moving average value 111 of the detection signal level and the number of occurrences or occurrence frequency of the defective event. Good.

[Calculation of defective part based on multiple waveform data]
The CPU of the data acquisition device 4 can also calculate higher-accuracy coordinate data of the defect event occurrence site in the mold based on the detection signal levels of the plurality of pressure sensors 2b of the sensor plate 2 at the same timing. It is.

  For example, as shown in FIG. 12, a case is assumed in which a defective event such as “two punching” occurs in a punched portion of one of the two punches P5 and P6. The CPU of the data collection device 4 has the same timing value in the waveform data of each pressure sensor 2b for the pressure sensor 2b at the position (x1, y1) where the peak value of the shear load is highest and the plurality of pressure sensors 2b around it. Based on the above, the coordinates of the position where the defective event has occurred are calculated. This makes it possible to identify the problematic punch (P6) and the non-problematic punch (P5) with high accuracy.

(About estimation of failure event of press molding machine 10)
FIG. 13 is a diagram comparing the waveform data of the two pressure sensors 2b. In this example, since the slide gib that slides and guides the reciprocating motion of the slide 13 is loosened, the relationship between the expansion and contraction of the distortion of the two waveform data is reversed in the period 131 in which the shear load is released after the shearing of the material is completed. doing. The data acquisition device 4 can thus take the difference between the two waveform data to estimate the occurrence of the slide give slack and present the result to the user through the display.

(Summary of effects of this embodiment)
1. As described above, according to the press manufacturing condition collection system 100 of the present embodiment, the shear load received by the mold is simultaneously detected by the plurality of pressure sensors 2 b in the sensor plate 2, and the shear detected by each pressure sensor 2 b By analyzing the waveform data of the load in each analysis AI unit 32 of the data analysis device 3, occurrence of failure events such as “two-piece removal” and “punch breakage” due to deterioration of the mold can be performed at high speed and with high accuracy Can be estimated. Furthermore, since a defective event for each part of the mold can be estimated, it is possible to reduce the time cost required for replacing the parts of the mold.

  2. According to the press manufacturing condition collection system 100 of the present embodiment, when a predetermined failure event such as “punch break” is estimated in the analysis AI unit 32 of the data analysis device 3, the press molding machine 10 is forced to perform press molding. Since the stop instruction is given, it is possible to minimize the occurrence of molding defects as much as possible.

  3. According to the press manufacturing condition collection system 100 of the present embodiment, the sensor plate 2 is mounted on the slide 13 of the press molding machine 10 so as to face the mold (lower mold 11), and is a surface orthogonal to the direction of receiving the load. Since the plurality of pressure sensors 2b are spaced apart from each other, the load for each part received by the mold can be detected no matter what mold is used.

(Modification)
In the above embodiment, the data analysis device 3 sends an instruction of forced stop of press forming to the controller 15 of the press forming machine 10 in response to the estimation of the defect event. In response to the determination of the sign or the sign thereof, a request to send a command to forcibly stop press forming may be sent to the data analysis device 3 through the network 6.

  As shown in FIG. 1, the press manufacturing condition collecting system 100 according to the present embodiment connects a plurality of data analysis devices 3 to the hub 5 and obtains the data collection device 4 by the plurality of data analysis devices 3. , And can be configured to collect an estimation result of defect events for each part of the mold and waveform data.

  As mentioned above, although "two sheets" and "punch breakage" were illustrated as inferior events, it is needless to say that it is good also as an object of estimation also about inferior events other than these.

DESCRIPTION OF SYMBOLS 2 ... Sensor plate 2b ... Pressure sensor 3 ... Data analysis device 4 ... Data acquisition device 10 ... Press molding machine 11 ... Upper mold 12 ... Lower mold 13 ... Slide 31 ... A / D conversion part 32 ... Analysis AI part 33 ... Data transmission Part 100 ... press manufacturing condition collection system

Claims (8)

A sensor plate which is disposed on the slide of the press forming machine so as to face the mold, and in which a plurality of pressure sensors are disposed apart from each other in a plane orthogonal to the load receiving direction;
A data analysis device that generates waveform data that is a time series of detection values for each pressure sensor and estimates a defective event of the mold based on the waveform data;
A press manufacturing condition collection system comprising: a data collection device for collecting the estimation results obtained by the data analysis device. The press manufacturing condition collecting system according to claim 1, wherein
The data analysis apparatus estimates the failure event for each part of the mold based on the waveform data for each pressure sensor. The press manufacturing condition collecting system according to claim 1 or 2, wherein
The data analysis system supplies a command for stopping press forming to the press forming machine when it estimates a defect event of the mold. The press manufacturing condition collection system according to any one of claims 1 to 3,
The data analysis device transmits the generated waveform data for each pressure sensor to the data collection device,
The data collection apparatus displays a pressure distribution of the mold on a display unit based on waveform data of each pressure sensor transmitted from the data analysis apparatus. The press manufacturing condition collection system according to any one of claims 1 to 4,
The data collection device accumulates the waveform data of each pressure sensor collected, and estimates a sign of a failure event of each part of the mold based on the waveform data of each pressure sensor accumulated. Manufacturing condition collection system. The press manufacturing condition collection system according to any one of claims 1 to 5,
The data collection device calculates a coordinate position where the defective event has occurred based on the collected waveform data for each of the pressure sensors. The press manufacturing condition collection system according to any one of claims 1 to 6,
The data analysis apparatus estimates a defective event for each part of the mold by analyzing the waveform data for each pressure sensor using an AI function. A plurality of pressure sensors arranged on the surface perpendicular to the direction of receiving a load, and arranged apart from each other, on the slide of the press molding machine, facing the mold;
A sensor plate for a press molding machine, comprising: an output terminal that outputs a detection signal for each pressure sensor to the outside.

Images