Trained model and injection molding machine control device

Abstract

To improve safety.SOLUTION: A trained model according to one embodiment includes: an input layer; one or more intermediate layers connected to the input layer; and an output layer connected to the intermediate layers. A machine learning is performed based on first data indicating results detected while a product is being produced by an injection molding machine or the results measured after the product is produced, and correct answer information that divides quantities related to the production of the product into two or more intervals and indicates, for each interval, whether the value included in that interval is correct or not. When second data indicating results detected during production or results measured after production is input from the input layer, a computer is caused to function so as to output from the output layer, for each of the intervals, a degree to which the value included in that interval is correct.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a trained model and a management device for an injection molding machine. [Background technology]

[0002] Conventionally, in order to properly produce a product using an injection molding machine, it is necessary to set the appropriate settings for that product. To set the appropriate settings, the quality of the product produced with the settings is checked, and the settings are repeatedly readjusted until the appropriate settings are derived. For this reason, the settings require a lot of time and effort from craftsmen.

[0003] In recent years, along with the improvement of computer processing power, there is a tendency for artificial intelligence to develop. For example, Patent Document 1 proposes a technology for estimating quality by machine learning. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2021-191619 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the technology described in Patent Document 1, quality is estimated from detection data detected during molding using a trained model, and it is difficult to grasp how reliable the estimated quality is.

[0006] One aspect of the present invention provides a technology that enables setting and control that takes reliability into consideration by dividing the amount of product production into two or more intervals and outputting the degree of accuracy for each interval. [Means for solving the problem]

[0007] A trained model according to one embodiment of the present invention has an input layer, one or more intermediate layers connected to the input layer, and an output layer connected to the intermediate layer, and performs machine learning based on first data indicating results detected while a product is being produced by an injection molding machine or results measured after the product is produced, and correct answer information that divides quantities related to the production of the product into two or more intervals and indicates, for each interval, whether the value included in the interval is correct or not, and causes a computer to function such that, when second data indicating results detected during production or results measured after production is input from the input layer, the degree to which the value included in each interval is correct for each interval is output from the output layer. Effect of the Invention

[0008] According to one aspect of the present invention, safety is improved by taking reliability into consideration. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram showing a state when mold opening of an injection molding machine according to an embodiment is completed. [Diagram 2] FIG. 2 is a diagram showing a state during mold clamping of the injection molding machine according to one embodiment. [Diagram 3] FIG. 3 is a conceptual diagram showing cooperation regarding a trained model between the learning device according to the first embodiment and the control device of the injection molding machine. [Figure 4] FIG. 4 is a diagram illustrating an example of a functional configuration of the learning device and the control device according to the first embodiment. [Diagram 5] FIG. 5 is a diagram illustrating various parameters for each shot number used in the evaluation acquisition unit and the correct answer information calculation unit according to the first embodiment. [Figure 6] FIG. 6 is a diagram showing an example of a calculation method of information on a correct answer of a mold clamping force set value by the correct answer information calculation unit according to the first embodiment. [Figure 7] FIG. 7 is a diagram illustrating information about a correct answer generated by the correct answer information calculation unit according to the first embodiment. [Figure 8] FIG. 8 is a diagram illustrating a structure of a trained model used in the learning device and the control device according to the first embodiment. [Figure 9] FIG. 9 is a diagram illustrating information based on row vectors output from the inference unit according to the first embodiment. [Figure 10] FIG. 10 is a diagram showing a first example of a graph showing the confidence level for each section, output from the inference unit according to the first embodiment. [Figure 11] FIG. 11 is a diagram showing a second example of a graph showing the confidence level for each section output from the inference unit according to the first embodiment. [Figure 12] FIG. 12 is a conceptual diagram showing inference performed by a trained model installed in a control device according to a modified example. [Figure 13] FIG. 13 is a diagram illustrating the configuration of a learning device, a group management device, and an injection molding machine according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The embodiment described below is an example, not a limitation of the invention, and all features and combinations described in the embodiment are not necessarily essential to the invention. In addition, the same or corresponding components in each drawing are denoted by the same or corresponding reference numerals, and the description may be omitted.

[0011] FIG. 1 is a diagram showing a state of the injection molding machine according to the first embodiment when mold opening is completed. FIG. 2 is a diagram showing a state of the injection molding machine according to the first embodiment when mold clamping is performed. In this specification, the X-axis direction, the Y-axis direction, and the Z-axis direction are perpendicular to each other. The X-axis direction and the Y-axis direction represent the horizontal direction, and the Z-axis direction represents the vertical direction. When the mold clamping device 100 is of a horizontal type, the X-axis direction is the mold opening / closing direction, and the Y-axis direction is the width direction of the injection molding machine 10. The negative side of the Y-axis direction is called the operation side, and the positive side of the Y-axis direction is called the anti-operation side.

[0012] As shown in FIG. 1 and FIG. 2, the injection molding machine 10 includes a mold clamping device 100 that opens and closes a mold device 800, an ejector device 200 that ejects a molded product molded by the mold device 800, an injection device 300 that injects a molding material into the mold device 800, a moving device 400 that moves the injection device 300 forward and backward relative to the mold device 800, a control device 700 that controls each component of the injection molding machine 10, and a frame 900 that supports each component of the injection molding machine 10. The frame 900 includes a mold clamping device frame 910 that supports the mold clamping device 100, and an injection device frame 920 that supports the injection device 300. The mold clamping device frame 910 and the injection device frame 920 are each installed on the floor 2 via a leveling adjuster 930. The control device 700 is disposed in the internal space of the injection device frame 920. Hereinafter, each component of the injection molding machine 10 will be described.

[0013] (mold clamping device) In the description of the clamping unit 100, the moving direction of the movable platen 120 during mold closing (for example, the positive direction of the X-axis) is the front, and the moving direction of the movable platen 120 during mold opening (for example, the negative direction of the X-axis) is the rear.

[0014] The mold clamping device 100 performs mold closing, pressurization, mold clamping, depressurization, and mold opening of the mold device 800. The mold device 800 includes a fixed mold 810 and a movable mold 820. The mold clamping device 100 is, for example, a horizontal type, and the mold opening and closing direction is horizontal. The mold clamping device 100 has a fixed platen 110 to which the fixed mold 810 is attached, a movable platen 120 to which the movable mold 820 is attached, and a movement mechanism 102 that moves the movable platen 120 in the mold opening and closing direction relative to the fixed platen 110.

[0015] The stationary platen 110 is fixed to the mold clamping unit frame 910. A stationary mold 810 is attached to the surface of the stationary platen 110 that faces the movable platen 120.

[0016] The movable platen 120 is disposed so as to be movable in the mold opening and closing direction relative to the mold clamping unit frame 910. A guide 101 for guiding the movable platen 120 is disposed on the mold clamping unit frame 910. A movable mold 820 is attached to the surface of the movable platen 120 facing the fixed platen 110.

[0017] The moving mechanism 102 advances and retreats the movable platen 120 relative to the fixed platen 110, thereby performing mold closing, pressurization, mold clamping, depressurization, and mold opening of the mold apparatus 800. The moving mechanism 102 has a toggle support 130 disposed at a distance from the fixed platen 110, a tie bar 140 connecting the fixed platen 110 and the toggle support 130, a toggle mechanism 150 that moves the movable platen 120 in the mold opening and closing direction relative to the toggle support 130, a mold clamping motor 160 that operates the toggle mechanism 150, a motion conversion mechanism 170 that converts the rotational motion of the mold clamping motor 160 into linear motion, and a mold thickness adjustment mechanism 180 that adjusts the distance between the fixed platen 110 and the toggle support 130.

[0018] The toggle support 130 is disposed at a distance from the fixed platen 110, and is placed on the mold clamping unit frame 910 so as to be freely movable in the mold opening and closing direction. The toggle support 130 may be disposed so as to be freely movable along a guide laid on the mold clamping unit frame 910. The guide of the toggle support 130 may be the same as the guide 101 of the movable platen 120.

[0019] In this embodiment, the fixed platen 110 is fixed to the mold clamping unit frame 910, and the toggle support 130 is arranged so as to be freely movable in the mold opening and closing direction relative to the mold clamping unit frame 910, but the toggle support 130 may be fixed to the mold clamping unit frame 910, and the fixed platen 110 may be arranged so as to be freely movable in the mold opening and closing direction relative to the mold clamping unit frame 910.

[0020] The tie bar 140 connects the fixed platen 110 and the toggle support 130 at a distance L in the mold opening / closing direction. A plurality of tie bars 140 (for example, four) may be used. The plurality of tie bars 140 are arranged parallel to the mold opening / closing direction and extend according to the mold clamping force. At least one tie bar 140 may be provided with a tie bar strain detector 141 that detects strain of the tie bar 140. The tie bar strain detector 141 sends a signal indicating the detection result to the control device 700. The detection result of the tie bar strain detector 141 is used for detecting the mold clamping force, etc.

[0021] In this embodiment, the tie bar strain detector 141 is used as the clamping force detector for detecting the clamping force, but the present invention is not limited to this. The clamping force detector is not limited to the strain gauge type, and may be a piezoelectric type, a capacitive type, a hydraulic type, an electromagnetic type, or the like, and the attachment position thereof is also not limited to the tie bar 140.

[0022] The toggle mechanism 150 is disposed between the movable platen 120 and the toggle support 130, and moves the movable platen 120 relative to the toggle support 130 in the mold opening / closing direction. The toggle mechanism 150 has a crosshead 151 that moves in the mold opening / closing direction, and a pair of link groups that bend and stretch with the movement of the crosshead 151. Each of the pair of link groups has a first link 152 and a second link 153 that are connected to bendable and stretchable by a pin or the like. The first link 152 is attached to the movable platen 120 by a pin or the like so as to be swingable. The second link 153 is attached to the toggle support 130 by a pin or the like so as to be swingable. The second link 153 is attached to the crosshead 151 via a third link 154. When the crosshead 151 is advanced or retreated relative to the toggle support 130, the first link 152 and the second link 153 bend and stretch, and the movable platen 120 advances or retreats relative to the toggle support 130.

[0023] The configuration of toggle mechanism 150 is not limited to the configuration shown in Figures 1 and 2. For example, although the number of joints in each link group is five in Figures 1 and 2, it may be four, and one end of third link 154 may be connected to a joint between first link 152 and second link 153.

[0024] The mold clamping motor 160 is attached to the toggle support 130 and operates the toggle mechanism 150. The mold clamping motor 160 advances and retreats the crosshead 151 relative to the toggle support 130, thereby bending and extending the first link 152 and the second link 153 and advancing and retreating the movable platen 120 relative to the toggle support 130. The mold clamping motor 160 is directly connected to the motion conversion mechanism 170, but may also be connected to the motion conversion mechanism 170 via a belt, a pulley, or the like.

[0025] The motion conversion mechanism 170 converts the rotational motion of the mold clamping motor 160 into the linear motion of the crosshead 151. The motion conversion mechanism 170 includes a screw shaft and a screw nut that screws onto the screw shaft. A ball or roller may be interposed between the screw shaft and the screw nut.

[0026] The mold clamping unit 100 performs a mold closing process, a pressure increase process, a mold clamping process, a pressure release process, a mold opening process, and the like under the control of the control device 700.

[0027] In the mold closing process, the mold clamping motor 160 is driven to move the crosshead 151 forward at a set moving speed to a mold closing completion position, thereby moving the movable platen 120 forward and bringing the movable mold 820 into contact with the fixed mold 810. The position and moving speed of the crosshead 151 are detected by using, for example, a mold clamping motor encoder 161. The mold clamping motor encoder 161 detects the rotation of the mold clamping motor 160 and sends a signal indicating the detection result to the control device 700.

[0028] The crosshead position detector that detects the position of the crosshead 151 and the crosshead movement speed detector that detects the movement speed of the crosshead 151 are not limited to the mold clamping motor encoder 161, and general devices can be used. Furthermore, the movable platen position detector that detects the position of the movable platen 120 and the movable platen movement speed detector that detects the movement speed of the movable platen 120 are not limited to the mold clamping motor encoder 161, and general devices can be used.

[0029] In the pressure increasing step, the mold clamping motor 160 is further driven to move the crosshead 151 further forward from the mold closing completion position to the mold clamping position, thereby generating a mold clamping force.

[0030] In the mold clamping process, the mold clamping motor 160 is driven to maintain the position of the crosshead 151 at the mold clamping position. In the mold clamping process, the mold clamping force generated in the pressure increase process is maintained. In the mold clamping process, a cavity space 801 (see FIG. 2) is formed between the movable mold 820 and the fixed mold 810, and the injection device 300 fills the cavity space 801 with liquid molding material. The filled molding material is solidified to obtain a molded product.

[0031] The number of cavity spaces 801 may be one or more. In the latter case, multiple molded products are obtained at the same time. An insert material may be placed in a part of the cavity space 801, and another part of the cavity space 801 may be filled with a molding material. A molded product in which the insert material and the molding material are integrated is obtained.

[0032] In the depressurization process, the mold clamping motor 160 is driven to move the crosshead 151 back from the mold clamping position to the mold opening start position, thereby moving the movable platen 120 back and reducing the mold clamping force. The mold opening start position and the mold closing completion position may be the same position.

[0033] In the mold opening process, the mold clamping motor 160 is driven to move the crosshead 151 backward at a set moving speed from the mold opening start position to the mold opening completion position, thereby moving the movable platen 120 backward and separating the movable mold 820 from the fixed mold 810. After that, the ejector unit 200 ejects the molded product from the movable mold 820.

[0034] The setting conditions in the mold closing process, the pressure increase process, and the mold clamping process are set together as a series of setting conditions. For example, the moving speed and position of the crosshead 151 in the mold closing process and the pressure increase process (including the mold closing start position, the moving speed switching position, the mold closing completion position, and the mold clamping position) and the mold clamping force are set together as a series of setting conditions. The mold closing start position, the moving speed switching position, the mold closing completion position, and the mold clamping position are arranged in this order from the rear side to the front, and represent the start point and the end point of the section in which the moving speed is set. The moving speed is set for each section. There may be one or more moving speed switching positions. The moving speed switching position does not have to be set. Only one of the mold clamping position and the mold clamping force may be set.

[0035] The setting conditions in the depressurization process and mold opening process are set in the same manner. For example, the movement speed and position of the crosshead 151 in the depressurization process and mold opening process (mold opening start position, movement speed switching position, and mold opening completion position) are set together as a series of setting conditions. The mold opening start position, movement speed switching position, and mold opening completion position are arranged in this order from the front to the rear, and represent the start and end points of the section in which the movement speed is set. The movement speed is set for each section. There may be one or more movement speed switching positions. The movement speed switching position does not have to be set. The mold opening start position and the mold closing completion position may be the same position. Also, the mold opening completion position and the mold closing start position may be the same position.

[0036] Note that the moving speed and position of the movable platen 120 may be set instead of the moving speed and position of the crosshead 151. Moreover, the clamping force may be set instead of the position of the crosshead (e.g., the clamping position) or the position of the movable platen.

[0037] Incidentally, the toggle mechanism 150 amplifies the driving force of the mold clamping motor 160 and transmits it to the movable platen 120. The amplification ratio is also called the toggle ratio. The toggle ratio changes according to the angle θ between the first link 152 and the second link 153 (hereinafter also referred to as the "link angle θ"). The link angle θ is determined from the position of the crosshead 151. When the link angle θ is 180°, the toggle ratio is maximum.

[0038] When the thickness of the mold device 800 changes due to replacement of the mold device 800 or a temperature change of the mold device 800, a mold thickness adjustment is performed so that a predetermined clamping force is obtained during mold clamping. In the mold thickness adjustment, for example, the distance L between the fixed platen 110 and the toggle support 130 is adjusted so that the link angle θ of the toggle mechanism 150 becomes a predetermined angle at the mold touch time when the movable mold 820 touches the fixed mold 810.

[0039] The mold clamping device 100 has a mold thickness adjustment mechanism 180. The mold thickness adjustment mechanism 180 adjusts the mold thickness by adjusting the distance L between the fixed platen 110 and the toggle support 130. The mold thickness adjustment is performed, for example, between the end of a molding cycle and the start of the next molding cycle. The mold thickness adjustment mechanism 180 has, for example, a screw shaft 181 formed at the rear end of the tie bar 140, a screw nut 182 held rotatably and immovably on the toggle support 130, and a mold thickness adjustment motor 183 that rotates the screw nut 182 screwed onto the screw shaft 181.

[0040] The screw shaft 181 and the screw nut 182 are provided for each tie bar 140. The rotational driving force of the mold thickness adjustment motor 183 may be transmitted to the multiple screw nuts 182 via a rotational driving force transmission unit 185. The multiple screw nuts 182 can be rotated synchronously. Note that by changing the transmission path of the rotational driving force transmission unit 185, it is also possible to rotate the multiple screw nuts 182 individually.

[0041] The rotational drive force transmission unit 185 is composed of, for example, gears. In this case, a driven gear is formed on the outer periphery of each screw nut 182, a drive gear is attached to the output shaft of the mold thickness adjustment motor 183, and an intermediate gear that meshes with the multiple driven gears and the drive gear is rotatably held in the center of the toggle support 130. Note that the rotational drive force transmission unit 185 may be composed of a belt, a pulley, or the like, instead of gears.

[0042] The operation of the mold thickness adjustment mechanism 180 is controlled by a control device 700. The control device 700 drives the mold thickness adjustment motor 183 to rotate the screw nut 182. As a result, the position of the toggle support 130 relative to the tie bar 140 is adjusted, and the interval L between the fixed platen 110 and the toggle support 130 is adjusted. Note that a plurality of mold thickness adjustment mechanisms may be used in combination.

[0043] The distance L is detected using a mold thickness adjustment motor encoder 184. The mold thickness adjustment motor encoder 184 detects the amount of rotation and the direction of rotation of the mold thickness adjustment motor 183, and sends a signal indicating the detection result to the control device 700. The detection result of the mold thickness adjustment motor encoder 184 is used to monitor and control the position of the toggle support 130 and the distance L. Note that the toggle support position detector that detects the position of the toggle support 130 and the distance detector that detects the distance L are not limited to the mold thickness adjustment motor encoder 184, and general types can be used.

[0044] The mold clamping apparatus 100 may have a mold temperature regulator that regulates the temperature of the mold apparatus 800. The mold apparatus 800 has a flow path for a temperature control medium therein. The mold temperature regulator regulates the temperature of the mold apparatus 800 by regulating the temperature of the temperature control medium supplied to the flow path of the mold apparatus 800.

[0045] Although the mold clamping apparatus 100 of this embodiment is of a horizontal type in which the mold opening and closing direction is horizontal, it may be of a vertical type in which the mold opening and closing direction is vertical.

[0046] Although the mold clamping apparatus 100 of the present embodiment has a mold clamping motor 160 as a drive source, it may have a hydraulic cylinder instead of the mold clamping motor 160. Also, the mold clamping apparatus 100 may have a linear motor for opening and closing the mold, and an electromagnet for mold clamping.

[0047] (Ejector device) In describing the ejector unit 200, similar to the description of the mold clamping unit 100, the direction of movement of the movable platen 120 during mold closing (e.g., the positive direction of the X-axis) will be referred to as the forward direction, and the direction of movement of the movable platen 120 during mold opening (e.g., the negative direction of the X-axis) will be referred to as the rearward direction.

[0048] The ejector unit 200 is attached to the movable platen 120 and moves forward and backward together with the movable platen 120. The ejector unit 200 has an ejector rod 210 that ejects a molded product from the mold device 800, and a drive mechanism 220 that moves the ejector rod 210 in the movement direction of the movable platen 120 (X-axis direction).

[0049] The ejector rod 210 is arranged so as to be able to move forward and backward in a through hole of the movable platen 120. The front end of the ejector rod 210 contacts an ejector plate 826 of the movable mold 820. The front end of the ejector rod 210 may or may not be connected to the ejector plate 826.

[0050] The drive mechanism 220 includes, for example, an ejector motor and a motion conversion mechanism that converts the rotational motion of the ejector motor into the linear motion of the ejector rod 210. The motion conversion mechanism includes a screw shaft and a screw nut that screws onto the screw shaft. A ball or roller may be interposed between the screw shaft and the screw nut.

[0051] The ejector unit 200 performs an ejection process under the control of the control unit 700. In the ejection process, the ejector rod 210 advances from the standby position to the ejection position at a set moving speed, thereby advancing the ejector plate 826 and ejecting the molded product. After that, the ejector motor is driven to move the ejector rod 210 backward at the set moving speed, and the ejector plate 826 backward to the original standby position.

[0052] The position and moving speed of the ejector rod 210 are detected, for example, by using an ejector motor encoder. The ejector motor encoder detects the rotation of the ejector motor and sends a signal indicating the detection result to the control device 700. Note that the ejector rod position detector that detects the position of the ejector rod 210 and the ejector rod moving speed detector that detects the moving speed of the ejector rod 210 are not limited to the ejector motor encoder, and general types can be used.

[0053] (injection device) In the description of the injection device 300, unlike the description of the mold clamping device 100 and the description of the ejector device 200, the movement direction of the screw 330 during filling (e.g., the negative X-axis direction) is described as the forward direction, and the movement direction of the screw 330 during metering (e.g., the positive X-axis direction) is described as the rearward direction.

[0054] The injection device 300 is installed on a slide base 301, and the slide base 301 is disposed so as to be movable forward and backward with respect to the injection device frame 920. The injection device 300 is disposed so as to be movable forward and backward with respect to the mold device 800. The injection device 300 touches the mold device 800, and fills the molding material measured in the cylinder 310 into a cavity space 801 in the mold device 800. The injection device 300 has, for example, a cylinder 310 for heating the molding material, a nozzle 320 provided at the front end of the cylinder 310, a screw 330 disposed in the cylinder 310 so as to be movable forward and backward and rotatable, a metering motor 340 for rotating the screw 330, an injection motor 350 for moving the screw 330 forward and backward, and a load detector 360 for detecting a load transmitted between the injection motor 350 and the screw 330.

[0055] Cylinder 310 heats the molding material supplied to the inside from supply port 311. The molding material includes, for example, resin. The molding material is formed, for example, in the shape of pellets, and is supplied to supply port 311 in a solid state. Supply port 311 is formed at the rear of cylinder 310. A cooler 312 such as a water-cooled cylinder is provided on the outer periphery of the rear of cylinder 310. A heater 313 such as a band heater and a temperature detector 314 are provided on the outer periphery of cylinder 310, forward of cooler 312.

[0056] Cylinder 310 is divided into a plurality of zones in the axial direction (e.g., X-axis direction) of cylinder 310. Each of the plurality of zones is provided with heater 313 and temperature detector 314. A set temperature is set for each of the plurality of zones, and control device 700 controls heater 313 so that the detected temperature of temperature detector 314 becomes the set temperature.

[0057] The nozzle 320 is provided at the front end of the cylinder 310, and is pressed against the mold device 800. A heater 313 and a temperature detector 314 are provided on the outer periphery of the nozzle 320. The control device 700 controls the heater 313 so that the detected temperature of the nozzle 320 becomes the set temperature.

[0058] The screw 330 is disposed in the cylinder 310 so as to be rotatable and movable forward and backward. When the screw 330 is rotated, the molding material is sent forward along the spiral groove of the screw 330. As the molding material is sent forward, it is gradually melted by heat from the cylinder 310. As the liquid molding material is sent forward of the screw 330 and accumulates in the front part of the cylinder 310, the screw 330 is moved backward. Thereafter, when the screw 330 is moved forward, the liquid molding material accumulated in front of the screw 330 is injected from the nozzle 320 and filled into the mold device 800.

[0059] A check ring 331 is attached to the front of the screw 330 so as to be movable back and forth as a check valve for preventing the backflow of molding material from the front to the rear of the screw 330 when the screw 330 is pushed forward.

[0060] When the screw 330 is moved forward, the backflow prevention ring 331 is pushed backward by the pressure of the molding material in front of the screw 330, and moves backward relatively to the screw 330 to a blocking position (see FIG. 2) where the flow path of the molding material is blocked. This prevents the molding material accumulated in front of the screw 330 from flowing backward.

[0061] On the other hand, when the screw 330 is rotated, the backflow prevention ring 331 is pushed forward by the pressure of the molding material sent forward along the spiral groove of the screw 330, and moves forward relatively to the screw 330 to an open position (see FIG. 1) where the flow path of the molding material is opened. As a result, the molding material is sent forward of the screw 330.

[0062] The backflow prevention ring 331 may be either a co-rotating type that rotates together with the screw 330 or a non-co-rotating type that does not rotate together with the screw 330.

[0063] The injection device 300 may have a drive source for moving the backflow prevention ring 331 back and forth between the open position and the closed position relative to the screw 330.

[0064] The metering motor 340 rotates the screw 330. The drive source for rotating the screw 330 is not limited to the metering motor 340, and may be, for example, a hydraulic pump.

[0065] The injection motor 350 advances and retreats the screw 330. A motion conversion mechanism that converts the rotational motion of the injection motor 350 into the linear motion of the screw 330 is provided between the injection motor 350 and the screw 330. The motion conversion mechanism has, for example, a screw shaft and a screw nut that screws onto the screw shaft. A ball or roller may be provided between the screw shaft and the screw nut. The drive source that advances and retreats the screw 330 is not limited to the injection motor 350, and may be, for example, a hydraulic cylinder.

[0066] The load detector 360 detects the load transmitted between the injection motor 350 and the screw 330. The detected load is converted into pressure by the control device 700. The load detector 360 is provided on a load transmission path between the injection motor 350 and the screw 330, and detects the load acting on the load detector 360.

[0067] The load detector 360 sends a signal of the detected load to the control device 700. The load detected by the load detector 360 is converted into pressure acting between the screw 330 and the molding material, and is used to control and monitor the pressure that the screw 330 receives from the molding material, the back pressure on the screw 330, the pressure that the screw 330 acts on the molding material, and the like.

[0068] The pressure detector for detecting the pressure of the molding material is not limited to the load detector 360, and a general detector may be used. For example, a nozzle pressure sensor or a mold internal pressure sensor may be used. The nozzle pressure sensor is installed in the nozzle 320.

[0069] The injection device 300 performs a metering process, a filling process, a pressure holding process, and the like under the control of the control device 700. The filling process and the pressure holding process may be collectively referred to as the injection process.

[0070] In the metering process, the metering motor 340 is driven to rotate the screw 330 at a set rotation speed, and the molding material is sent forward along the spiral groove of the screw 330. As a result, the molding material is gradually melted. As the liquid molding material is sent forward of the screw 330 and accumulates at the front of the cylinder 310, the screw 330 is moved backward. The rotation speed of the screw 330 is detected, for example, by using a metering motor encoder 341. The metering motor encoder 341 detects the rotation of the metering motor 340 and sends a signal indicating the detection result to the control device 700. Note that the screw rotation speed detector that detects the rotation speed of the screw 330 is not limited to the metering motor encoder 341, and a general one can be used.

[0071] In the metering process, in order to restrict abrupt retraction of the screw 330, a set back pressure may be applied to the screw 330 by driving the injection motor 350. The back pressure on the screw 330 is detected, for example, by using a load detector 360. When the screw 330 retracts to the metering completion position and a predetermined amount of molding material is accumulated in front of the screw 330, the metering process is completed.

[0072] The position and rotation speed of the screw 330 in the metering process are set together as a series of setting conditions. For example, a metering start position, a rotation speed switching position, and a metering completion position are set. These positions are arranged in this order from the front to the rear, and represent the start and end points of the section in which the rotation speed is set. The rotation speed is set for each section. There may be one or more rotation speed switching positions. The rotation speed switching positions do not have to be set. In addition, a back pressure is set for each section.

[0073] In the filling step, the injection motor 350 is driven to move the screw 330 forward at a set moving speed, and the liquid molding material accumulated in front of the screw 330 is filled into the cavity space 801 in the mold device 800. The position and moving speed of the screw 330 are detected, for example, by using an injection motor encoder 351. The injection motor encoder 351 detects the rotation of the injection motor 350 and sends a signal indicating the detection result to the control device 700. When the position of the screw 330 reaches a set position, switching from the filling step to the pressure holding step (so-called V / P switching) is performed. The position where the V / P switching is performed is also called the V / P switching position. The set moving speed of the screw 330 may be changed depending on the position of the screw 330, time, etc.

[0074] The position and movement speed of the screw 330 in the filling process are set together as a series of setting conditions. For example, a filling start position (also called an "injection start position"), a movement speed switching position, and a V / P switching position are set. These positions are arranged in this order from the rear side to the front, and represent the start and end points of the section for which the movement speed is set. The movement speed is set for each section. There may be one or more movement speed switching positions. The movement speed switching positions do not have to be set.

[0075] An upper limit value for the pressure of the screw 330 is set for each section in which the moving speed of the screw 330 is set. The pressure of the screw 330 is detected by a load detector 360. When the pressure of the screw 330 is equal to or lower than the set pressure, the screw 330 is advanced at the set moving speed. On the other hand, when the pressure of the screw 330 exceeds the set pressure, the screw 330 is advanced at a moving speed slower than the set moving speed so that the pressure of the screw 330 is equal to or lower than the set pressure, for the purpose of protecting the mold.

[0076] After the position of the screw 330 reaches the V / P switching position in the filling process, the screw 330 may be temporarily stopped at the V / P switching position, and then the V / P switching may be performed. Immediately before the V / P switching, instead of stopping the screw 330, the screw 330 may be moved forward or backward at a slow speed. Furthermore, the screw position detector that detects the position of the screw 330 and the screw movement speed detector that detects the movement speed of the screw 330 are not limited to the injection motor encoder 351, and general detectors may be used.

[0077] In the holding pressure step, the injection motor 350 is driven to push the screw 330 forward, and the pressure of the molding material at the front end of the screw 330 (hereinafter also referred to as "holding pressure") is maintained at a set pressure, and the molding material remaining in the cylinder 310 is pushed toward the mold device 800. The molding material that is insufficient due to cooling contraction in the mold device 800 can be replenished. The holding pressure is detected, for example, by a load detector 360. The set value of the holding pressure may be changed depending on the elapsed time from the start of the holding pressure step. The holding pressure and the holding time for which the holding pressure is held in the holding pressure step may each be set multiple times, or may be set collectively as a series of setting conditions.

[0078] In the dwelling step, the molding material in the cavity space 801 in the mold device 800 is gradually cooled, and when the dwelling step is completed, the entrance to the cavity space 801 is blocked by the solidified molding material. This state is called a gate seal, and prevents the molding material from flowing back from the cavity space 801. After the dwelling step, the cooling step is started. In the cooling step, the molding material in the cavity space 801 is solidified. A measuring step may be performed during the cooling step in order to shorten the molding cycle time.

[0079] The injection device 300 of this embodiment is of an in-line screw type, but may be of a pre-plastication type or the like. A pre-plastication type injection device supplies molding material molten in a plasticization cylinder to an injection cylinder, and injects the molding material from the injection cylinder into a mold device. A screw is disposed in the plasticization cylinder so as to be rotatable but unable to move forward or backward, or a screw is disposed so as to be rotatable and able to move forward or backward. Meanwhile, a plunger is disposed in the injection cylinder so as to be able to move forward or backward.

[0080] Furthermore, the injection device 300 of this embodiment is a horizontal type in which the axial direction of the cylinder 310 is horizontal, but may be a vertical type in which the axial direction of the cylinder 310 is vertical. A mold clamping device to be combined with the vertical injection device 300 may be either a vertical type or a horizontal type. Similarly, a mold clamping device to be combined with the horizontal injection device 300 may be either a horizontal type or a vertical type.

[0081] (Mobile device) In the description of the moving device 400, similar to the description of the injection device 300, the moving direction of the screw 330 during filling (for example, the negative X-axis direction) is the front, and the moving direction of the screw 330 during metering (for example, the positive X-axis direction) is the rear.

[0082] The moving device 400 moves the injection device 300 forward and backward relative to the mold device 800. In addition, the moving device 400 presses the nozzle 320 against the mold device 800 to generate a nozzle touch pressure. The moving device 400 includes a hydraulic pump 410, a motor 420 as a drive source, a hydraulic cylinder 430 as a hydraulic actuator, and the like.

[0083] The hydraulic pump 410 has a first port 411 and a second port 412. The hydraulic pump 410 is a pump that can rotate in both directions, and by switching the rotation direction of the motor 420, it draws in hydraulic fluid (e.g., oil) from one of the first port 411 and the second port 412 and discharges it from the other, thereby generating hydraulic pressure. Note that the hydraulic pump 410 can also draw in hydraulic fluid from a tank and discharge the hydraulic fluid from one of the first port 411 and the second port 412.

[0084] The motor 420 operates the hydraulic pump 410. The motor 420 drives the hydraulic pump 410 with a rotational direction and rotational torque according to a control signal from the control device 700. The motor 420 may be an electric motor or an electric servo motor.

[0085] The hydraulic cylinder 430 has a cylinder body 431, a piston 432, and a piston rod 433. The cylinder body 431 is fixed to the injection device 300. The piston 432 divides the inside of the cylinder body 431 into a front chamber 435 as a first chamber and a rear chamber 436 as a second chamber. The piston rod 433 is fixed to the fixed platen 110.

[0086] A front chamber 435 of the hydraulic cylinder 430 is connected to a first port 411 of the hydraulic pump 410 via a first flow path 401. The hydraulic fluid discharged from the first port 411 is supplied to the front chamber 435 via the first flow path 401, thereby pushing the injection unit 300 forward. The injection unit 300 is moved forward, and the nozzle 320 is pressed against the fixed die 810. The front chamber 435 functions as a pressure chamber that generates a nozzle touch pressure of the nozzle 320 by the pressure of the hydraulic fluid supplied from the hydraulic pump 410.

[0087] On the other hand, the rear chamber 436 of the hydraulic cylinder 430 is connected to the second port 412 of the hydraulic pump 410 via the second flow path 402. The hydraulic fluid discharged from the second port 412 is supplied to the rear chamber 436 of the hydraulic cylinder 430 via the second flow path 402, thereby pushing the injection unit 300 backward. The injection unit 300 is moved backward, and the nozzle 320 is separated from the fixed mold 810.

[0088] In this embodiment, the moving device 400 includes the hydraulic cylinder 430, but the present invention is not limited to this. For example, instead of the hydraulic cylinder 430, an electric motor and a motion conversion mechanism that converts the rotational motion of the electric motor into the linear motion of the injection device 300 may be used.

[0089] (Control device) The control device 700 is configured, for example, by a computer, and has a CPU (Central Processing Unit) 701, a storage medium 702 such as a memory, an input interface 703, an output interface 704, and a communication interface 705 as shown in Fig. 1 and Fig. 2. The control device 700 performs various controls by causing the CPU 701 to execute a program stored in the storage medium 702. The control device 700 also receives signals from the outside via the input interface 703, and transmits signals to the outside via the output interface 704.

[0090] The control device 700 repeatedly performs a metering process, mold closing process, pressure increase process, mold clamping process, filling process, pressure holding process, cooling process, pressure release process, mold opening process, and ejection process, etc., to repeatedly produce molded products. A series of operations to obtain a molded product, for example, an operation from the start of a metering process to the start of the next metering process, is also called a "shot" or a "molding cycle." The time required for one shot is also called the "molding cycle time" or "cycle time."

[0091] One molding cycle includes, for example, a metering process, a mold closing process, a pressure increase process, a mold clamping process, a filling process, a pressure holding process, a cooling process, a pressure release process, a mold opening process, and an ejection process, in this order. The order here refers to the order in which each process starts. The filling process, the pressure holding process, and the cooling process are performed during the mold clamping process. The start of the mold clamping process may coincide with the start of the filling process. The completion of the pressure release process coincides with the start of the mold opening process.

[0092] In addition, in order to shorten the molding cycle time, a plurality of processes may be performed simultaneously. For example, the metering process may be performed during the cooling process of the previous molding cycle, or during the mold clamping process. In this case, the mold closing process may be performed at the beginning of the molding cycle. The filling process may be started during the mold closing process. The ejection process may be started during the mold opening process. In the case where an opening / closing valve for opening and closing the flow path of the nozzle 320 is provided, the mold opening process may be started during the metering process. This is because even if the mold opening process is started during the metering process, the molding material will not leak from the nozzle 320 as long as the opening / closing valve closes the flow path of the nozzle 320.

[0093] One molding cycle may include steps other than the metering step, mold closing step, pressure increase step, mold clamping step, filling step, pressure holding step, cooling step, pressure release step, mold opening step, and ejection step.

[0094] For example, after the dwelling step is completed, a pre-metering suck-back step may be performed in which the screw 330 is retracted to a preset metering start position before the metering step is started. This can reduce the pressure of the molding material accumulated in front of the screw 330 before the metering step starts, and can prevent the screw 330 from suddenly retracting at the start of the metering step.

[0095] Furthermore, after the metering step is completed, a post-metering suck-back step may be performed in which the screw 330 is moved back to a preset filling start position (also called the "injection start position") before the filling step is started. This can reduce the pressure of the molding material accumulated in front of the screw 330 before the filling step is started, and can prevent the molding material from leaking from the nozzle 320 before the filling step is started.

[0096] The control device 700 is connected to an operation device 750 that accepts input operations by a user and a display device 760 that displays a screen. The operation device 750 and the display device 760 may be configured, for example, by a touch panel 770 and may be integrated. The touch panel 770 as the display device 760 displays a screen under the control of the control device 700. For example, the screen of the touch panel 770 may display information such as the settings of the injection molding machine 10 and the current state of the injection molding machine 10. The touch panel 770 is capable of accepting operations in the displayed screen area. In addition, for example, an operation unit such as a button or an input field that accepts an input operation by a user may be displayed in the screen area of ​​the touch panel 770. The touch panel 770 as the operation device 750 detects an input operation on the screen by the user and outputs a signal corresponding to the input operation to the control device 700. As a result, for example, the user can operate the operation unit provided on the screen while checking the information displayed on the screen to perform settings of the injection molding machine 10 (including input of a setting value), etc. Furthermore, the user can operate an operation unit provided on the screen to perform an operation of the injection molding machine 10 corresponding to the operation unit. The operation of the injection molding machine 10 may be, for example, the operation (including stopping) of the clamping device 100, the ejector device 200, the injection device 300, the moving device 400, etc. The operation of the injection molding machine 10 may be, for example, switching of a screen displayed on the touch panel 770 serving as the display device 760, etc.

[0097] In this embodiment, the operation device 750 and the display device 760 are described as being integrated as the touch panel 770, but they may be provided independently. Also, a plurality of operation devices 750 may be provided. The operation device 750 and the display device 760 are disposed on the operation side (Y-axis negative direction) of the mold clamping unit 100 (more specifically, the fixed platen 110).

[0098] (First embodiment) When molding with an injection molding machine, there is a demand to set up using a trained model. In this case, machine learning is required to generate the trained model.

[0099] It is also possible to carry out the learning phase of the machine learning process in an injection molding machine. However, this has the problem that the processing device of the injection molding machine is required to have high processing power. For this reason, it is possible to carry out the learning process in an information processing device different from the injection molding machine.

[0100] Therefore, in this embodiment, an example in which the learning phase and the inference phase are executed in different devices will be described. Specifically, a trained model created in the learning phase of a learning device is mounted on an injection molding machine. Note that, although this embodiment describes an example in which the learning phase and the inference phase are executed in different devices as an example, the present invention is not limited to a method in which the learning phase and the inference phase are executed in different devices, and the learning phase and the inference phase may be executed in the same device.

[0101] FIG. 3 is a conceptual diagram showing cooperation regarding a trained model by the learning device according to this embodiment and the control device of the injection molding machine.

[0102] In the example shown in FIG. 3, a test injection molding machine 1350, a learning device 1300, and a control device 700 of the injection molding machine 10 are shown.

[0103] 3, the test injection molding machine 1350 and the learning device 1300 may be owned by, for example, the manufacturer of the injection molding machine 10, but the method is not limited to ownership by the manufacturer of the injection molding machine, and may be owned by, for example, the destination of the injection molding machine, or may be owned by a service provider that generates a trained model. The injection molding machine 10 and the control device 700 shown in FIG. 3 may show, for example, a state before shipment.

[0104] The test injection molding machine 1350 is an injection molding machine for performing machine learning. The configuration of the test injection molding machine 1350 is similar to that of the injection molding machine 10, and therefore a description thereof will be omitted.

[0105] The test injection molding machine 1350 according to this embodiment produces a molded product (an example of a product) according to settings made by a user. The test injection molding machine 1350 outputs waveform data measured during injection molding to the learning device 1300. In this embodiment, data showing the detection results of a sensor provided in the test injection molding machine 1350 or the injection molding machine 10 in chronological order is referred to as waveform data.

[0106] The learning device 1300 generates teacher data based on the input waveform data, and stores the teacher data in the teacher data storage unit 1321 .

[0107] By the way, in order to utilize artificial intelligence, it is necessary to execute a training phase and an inference phase.

[0108] The learning device 1300 executes a learning phase. To this end, the learning unit 1315 of the learning device 1300 loads the teacher data stored in the teacher data storage unit 1321 into a neural network, and generates a network in which the weights and biases of the synapses are adjusted, as a learned model LM.

[0109] For example, the learning device 1300 reads a large amount of training data and performs machine learning by backpropagation using a neural network, thereby generating a learned model LM.

[0110] The learning device 1300 may be, for example, an on-premise server installed in a factory or the like, or may be a cloud server. Furthermore, the learning device 1300 may be a stationary terminal device arranged in a factory or the like, or a portable terminal device (mobile terminal). The stationary terminal device may include, for example, a desktop PC (Personal Computer). Furthermore, the portable terminal device may include, for example, a smartphone, a tablet terminal, a laptop PC, etc.

[0111] The control device 700 of the injection molding machine 10 executes the inference phase. The control device 700 inputs data to the trained model LM and causes the trained model LM to perform inference. Then, the control device 700 sets predetermined items of the molding conditions based on the output results from the trained model LM.

[0112] In this embodiment, the structure of the trained model LM generated by the learning unit 1315 of the learning device 1300 and the trained model LM used by the inference unit 715 of the control device 700 may be made common. Then, the learning device 1300 may transfer the weight and bias of the trained model LM to the control device 700. The control device 700 can match the trained model LM of the learning unit 1315 of the learning device 1300 by updating the trained model LM based on the received weight and bias. This allows the learning unit 1315 of the learning device 1300 and the inference unit 715 of the control device 700 of the injection molding machine 10 to be implemented in different languages. For example, the learning unit 1315 of the learning device 1300 may be implemented using, for example, Python. On the other hand, the inference unit 715 of the control device 700 may be implemented using, for example, C++.

[0113] Furthermore, the learning device 1300 and the control device 700 can have different CPUs, OSs, development languages, etc. The control device 700 can use a CPU with a lower processing speed than the learning device 1300. This can reduce costs.

[0114] Furthermore, the structural information of the weights and biases to be transferred need only include the values ​​of each weight, each bias, and parameter structure information, and may be, for example, an object in which the values ​​of each weight and each bias are described in accordance with a data structure using a predetermined method (e.g., JSON (JavaScript Object Notation), XML (Extensible Markup Language), ONNX (Open Neural Network Exchange), etc.).

[0115] This enables the learning device 1300 to pass information indicating the structure of the weights and biases to the control device 700 of the injection molding machine 10.

[0116] Note that this embodiment is not limited to a method of transferring information indicating the structure of weights and biases from the learning device 1300 to the injection molding machine 10, and the learned model LM itself may be transferred from the learning device 1300 to the control device 700 of the injection molding machine 10. In this case, the CPU, OS, development language, etc. may be common between the control device 700 and the learning device 1300.

[0117] This allows the control device 700 to perform inference using the learned model LM generated in the learning phase of the learning device 1300.

[0118] Fig. 4 is a diagram showing an example of the functional configuration of the learning device 1300 and the control device 700 according to this embodiment. As shown in Fig. 4, the learning device 1300 and the control device 700 are connected via a communication line NW.

[0119] The functions of the learning device 1300 are realized by any hardware or any combination of hardware and software, etc. For example, as shown in FIG.

[0120] The storage medium 1302 stores various installed programs, as well as files and data necessary for various processes. The storage medium 1302 includes, for example, a hard disk drive (HDD), a solid state drive (SSD), a flash memory, and the like.

[0121] The storage medium 1302 according to this embodiment includes a teacher data storage unit 1321 and a trained model storage unit 1322.

[0122] The training data storage unit 1321 stores training data used for training the trained model LM. The configuration of the training data will be described later.

[0123] The learned model memory unit 1322 stores the learned model LM.

[0124] The communication interface 1303 is used as an interface for communicatively connecting to an external device. This allows the learning device 1300 to communicate with an external device such as an injection molding machine 10 through the communication interface 1303. The communication interface 1303 may have multiple types of communication interfaces depending on the communication method between the connected device and the like.

[0125] A CPU 1301 of the learning device 1300 executes a program stored in a storage medium 1302. As a result, the CPU 1301 includes, as functional units, an acquisition unit 1311, an evaluation acquisition unit 1312, a correct answer information calculation unit 1313, a teacher data generation unit 1314, a learning unit 1315, and a communication control unit 1316.

[0126] The learning device 1300 according to the present embodiment generates a trained model LM by the above-mentioned configuration. For example, when waveform data showing a detection result by a sensor provided in the injection molding machine 10 in a time series is input while the injection molding machine 10 is performing injection molding, the trained model LM divides a range that can be set as an item related to the molding conditions of the injection molding machine 10 into two or more sections and outputs a confidence level for each section. The confidence level indicates the degree to which a value included in the section is correct as a value to be set for the item. Therefore, by referring to the output information, it is possible to estimate a value to be set for the item.

[0127] First, a configuration for the learning device 1300 to generate training data used to generate a trained model LM will be described.

[0128] The acquisition unit 1311 acquires waveform data (an example of the first data) indicating the detection results by a sensor provided in the test injection molding machine 1350 from the test injection molding machine 1350 each time a setting value is set for the item to be set and injection molding is performed.

[0129] In this embodiment, a case where the item to be set is a mold clamping force set value will be described. Then, the acquisition unit 1311 acquires waveform data (one example of collective data) indicating the mold clamping force detected by the mold clamping force detector from the start to the end of the molding cycle from the test injection molding machine 1350 every time the mold clamping force set value is set to mold a molded product. Note that in this embodiment, the acquired waveform data is not limited to waveform data indicating the fastening force from the start to the end of the molding cycle, but may be waveform data indicating the fastening force for each shot. For example, the acquired waveform data may be waveform data indicating a part or the whole of the molding cycle, or may be waveform data that approaches the previous and next shots.

[0130] The learning device 1300 according to this embodiment generates teacher data for outputting information related to the setting of the mold clamping force set value. Next, an example of a method for generating teacher data will be described. Note that the method for generating teacher data is not limited to the generation method shown below, and various methods are possible.

[0131] The learning device 1300 according to this embodiment generates teacher data based on an evaluation of the injection-molded product each time a clamping force set value is set and injection molding is performed based on that clamping force set value.

[0132] Specifically, the evaluation acquisition unit 1312 acquires evaluation information indicating an evaluation of the molded product each time a clamping force setting value is set and injection molding is performed based on the clamping force setting value. For example, the evaluation acquisition unit 1312 acquires a value indicating a measurement result by a sensor provided in the injection molding machine 10 (hereinafter referred to as a measured value) as the evaluation information. The evaluation of the molded product according to this embodiment is based on whether or not burrs are generated on the molded product. In other words, the evaluation acquisition unit 1312 according to this embodiment acquires a measured value that is an index of whether or not burrs are generated on the molded product.

[0133] The measurement may be, for example, the difference in clamping force (detected by a clamping force detector) occurring within a molding cycle.

[0134] Fig. 5 is a diagram illustrating various parameters for each (shot) number used in the evaluation acquisition unit 1312 and the correct information calculation unit 1313 according to this embodiment. In the example shown in Fig. 5, a clamping force set value, a clamping force detection value, a measurement value, a recommended clamping force value, and a clamping force adjustment value are shown in association with each other for each (shot) number. The clamping force detection value is, for example, the highest value (peak value) among the clamping force detection values ​​detected from the start to the end of a molding cycle. In the example shown in Fig. 5, the clamping force set value is made different for each shot, and molding conditions other than the clamping force set value are fixed.

[0135] The evaluation acquisition unit 1312 acquires a measurement value each time a clamping force setting value is set and injection molding is performed, in other words, for each (shot) number. The measurement value according to this embodiment is the difference in clamping force. Specifically, the measurement value is a value obtained by subtracting the detection value of the clamping force detected at the start of filling from the detection value of the clamping force detected at the time before the mold device 800 is opened.

[0136] In this embodiment, the acquired mold clamping force difference (in other words, the increased mold clamping force) is regarded as the amount by which the mold device 800 is opened due to the molding material that has entered the parting line surface. For this reason, the evaluation acquisition unit 1312 calculates the mold clamping force difference and acquires the calculated mold clamping force difference as a measurement value corresponding to the flash generated on the molded product.

[0137] In this embodiment, the measurement value corresponding to the burrs generated on the molded product is not limited to the value obtained by subtracting the detected clamping force from the detected clamping force. For example, the measurement value may be the result of measuring the amount of burrs on the molded product taken out of the molding machine.

[0138] In this embodiment, an example will be described in which the evaluation acquisition unit 1312 acquires a measured value (indicating an evaluation of the presence or absence of burrs) each time a mold clamping force set value is set and injection molding is performed, but the present invention is not limited to a method of acquiring a measured value each time a mold clamping force set value is set and injection molding is performed. For example, as shown in Fig. 5, the evaluation acquisition unit 1312 may acquire a difference in fastening force as a measured value (indicating an evaluation of the presence or absence of burrs) each time injection molding is performed and a mold clamping force detection value is detected.

[0139] The correct information calculation unit 1313 calculates a clamping force setting value that is estimated to prevent burrs from occurring in the molded product based on the measurement values ​​(an example of evaluation information) acquired for each clamping force setting value or clamping force detection value, as information indicating the correct answer.

[0140] FIG. 6 is a diagram showing an example of a calculation method of information on the correct answer of the clamping force set value by the correct answer information calculation unit 1313 according to the present embodiment. In the example shown in FIG. 6, the correct answer information calculation unit 1313 approximates the measurement value for each clamping force with two straight lines, a line 1601 and a line 1602. As an approximation method of the two straight lines, for example, the least square method is used. Then, the correct answer information calculation unit 1313 calculates the intersection of the two straight lines as the recommended clamping force value. In the example shown in FIG. 6, the correct answer information calculation unit 1313 calculates the clamping force 380 [kN], which is the intersection of the line 1601 and the line 1602, as the recommended clamping force value. Note that FIG. 6 shows an example of a calculation method, and is not limited to this calculation method. For example, the correct answer information calculation unit 1313 may perform curve approximation instead of two straight line approximation, and calculate the change point of the curve as the recommended clamping force value.

[0141] In the example shown in Fig. 5, the mold clamping force recommended value of 380 [kN] and the mold clamping force adjustment value for adjusting the mold clamping force detection value are shown for each number. The mold clamping force adjustment value shown in Fig. 5 is a value for adjusting the mold clamping force detection value, but may be a value for adjusting the current mold clamping force setting value. In this embodiment, the mold clamping force adjustment value is used as information on the correct answer.

[0142] In this embodiment, the correct answer information calculation unit 1313 divides the range of the mold clamping force setting value that can be set by the injection molding machine 10 into a predetermined interval (for example, 100 [kN]). Then, for each interval into which the range that can be set as the mold clamping force setting value is divided, the correct answer information calculation unit 1313 generates information on the correct answer indicating whether or not the mold clamping force adjustment value is included in the interval.

[0143] 7 is a diagram illustrating information related to a correct answer generated by the correct answer information calculation unit 1313 according to this embodiment. In the example shown in FIG. 7, whether or not a clamping force adjustment value is included is indicated for each section "100" in the range of clamping force setting values ​​from "-1000" to "1000" that can be set by the injection molding machine 10. In the example shown in FIG. 7, "1" indicates that a clamping force adjustment value is included, and "0" indicates that a clamping force adjustment value is not included. In the example shown in FIG. 7, each section is a half-open section, "[" indicates a closed section, in other words, greater than or equal to, and)" indicates an open section, in other words, smaller than.

[0144] As shown in FIG. 5, (shot) number "1" has a mold clamping force set value of "1000", a mold clamping force adjustment value of "990", a recommended mold clamping force value of "380", and a mold clamping force adjustment value of "-610".

[0145] 7, in the (shot) number "1", the section [-700, -600) is "1", in other words, this section includes the mold clamping force adjustment value. The other sections are "0".

[0146] Then, the correct answer information calculation unit 1313 generates a row vector indicating whether or not the mold clamping force adjustment value is included for each section as information regarding the correct answer. The row vector corresponding to the (shot) number "1" is, for example, [0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. The row vector is a matrix in which "1" or "0" is set for every 100 (half-open section) for the range from "1000" to "-1000". "1" indicates that the mold clamping force adjustment value is included, and "0" indicates that the mold clamping force adjustment value is not included.

[0147] For numbers "2" to "11" as well, the correct answer information calculation unit 1313 generates row vectors. The generation method is the same as in the case of number "1" described above, so a description thereof will be omitted.

[0148] 4, the teacher data generation unit 1314 generates teacher data by combining the waveform data for each (shot) number acquired by the acquisition unit 1311 and the row vector for each (shot) number generated by the correct answer information calculation unit 1313. Then, the teacher data generation unit 1314 stores the generated teacher data in the teacher data storage unit 1321.

[0149] This embodiment is an example of a method for generating teacher data, and teacher data may be generated by a method other than the above. In addition, in this embodiment, an example in which the correct answer information calculation unit 1313 automatically generates information on the correct answer based on the evaluation information has been described, but the method is not limited to automatically generating information based on the evaluation information. For example, the correct answer information calculation unit 1313 may generate information on the correct answer according to information input by the user via an operation device. In addition, after the correct answer information calculation unit 1313 calculates information on the correct answer, the user may visually confirm the information on the correct answer, and the information may be corrected.

[0150] In addition, the present invention is not limited to a method in which the learning device 1300 automatically performs all steps up to the generation of the trained model LM. For example, after the learning device 1300 generates a plurality of pieces of training data, the user may check the contents included in each of the plurality of pieces of training data.

[0151] In the present embodiment, the multiple pieces of teacher data may be referred to as a teacher data set. The teacher data set is a set of teacher data that is a combination of waveform data as input and a row vector indicating whether or not a clamping force adjustment value is included for each section as output.

[0152] After the user has completed checking and correcting the contents of the training data set, the learning device 1300 may generate a trained model LM based on the training data set.

[0153] (Generating trained models, etc.) The learning unit 1315 generates a trained model LM by performing machine learning based on the teacher data set stored in the teacher data storage unit 1321. The trained model LM is generated by applying supervised learning to a base learning model.

[0154] Specifically, the learning unit 1315 generates a learned model LM by performing machine learning based on waveform data indicating the results detected by the clamping force detector for each value set as the clamping force setting value or detected as the clamping force detection value included in the teacher dataset, and a row vector indicating whether or not a clamping force adjustment value is included for each section divided into the range that can be set as the clamping force setting value.

[0155] When waveform data (an example of the second data) showing a time series of the clamping force detected while the injection molding machine 10 or the test injection molding machine 1350 is performing molding is input from the input layer, the trained model LM outputs from the output layer a row vector indicating the confidence that the clamping force adjustment value is included in each interval divided into the range in which the clamping force can be set.

[0156] In addition, the trained model LM may be updated by additionally training a new teacher dataset on an existing trained model LM.

[0157] As the machine learning used to generate the trained model LM, for example, a neural network may be applied, or as another example, machine learning using a deep neural network (DNN) may be applied, and deep learning may be applied. As the deep learning, for example, a convolutional neural network, an RNN (Recurrent Neural Networks), or an LSTM (Long Short Term Memory) may be applied.

[0158] FIG. 8 is a diagram illustrating an example of the structure of a trained model LM used in the learning device 1300 and the control device 700 according to the first embodiment.

[0159] 8 is composed of an input layer 1911, one or more intermediate layers 1912 connected to the input layer, and an output layer 1913 connected to the intermediate layer. The input layer has, for example, 2200 nodes, the intermediate layer has, for example, 200 nodes, and the output layer has 20 nodes.

[0160] When input data 1901 corresponding to 2200 nodes is input, the learned model LM outputs output data 1902 corresponding to 20 nodes.

[0161] The input data 1901 is, for example, waveform data showing actual values ​​that change while injection molding is being performed in chronological order. The input data 1901 may also be, for example, waveform data showing changes in the clamping force, which is the detection result of a clamping force detector, in chronological order while injection molding is being performed.

[0162] The output data 1902 outputs, for example, a row vector for adjusting a predetermined item of the injection molding machine 10. The output data 1902 is, for example, a row vector for adjusting a clamping force setting value set when injection molding is performed by the injection molding machine 10. Specifically, the output data 1902 is a row vector indicating a degree of certainty for each section obtained by dividing a range that can be set as the clamping force setting value. The degree of certainty indicates the degree to which the clamping force adjustment value included in the section is correct as a value for adjusting the clamping force setting value. In this embodiment, the degree of certainty is a numerical value within a range of 0 to 1. If the degree of certainty is "1", it indicates that the probability that the clamping force adjustment value is included in the section is 100%, and if the degree of certainty is "0", it indicates that the probability that the clamping force adjustment value is included in the section is 0%.

[0163] In this way, when waveform data of the clamping force detected by the clamping force detector is input, the learned model LM outputs a row vector indicating the confidence level for each interval in the range that can be set as the clamping force setting value.

[0164] When the learning unit 1315 learns using the teacher data, the waveform data included in the teacher data may be pre-processed. Note that the pre-processing may also be performed when the inference unit 715 (described later) performs inference.

[0165] In this embodiment, for example, pre-processing may be performed on the clamping force set value or the clamping force detection value used in the waveform data, etc. by dividing it by the rated clamping force so that the clamping force set value or the clamping force detection value, etc., falls within the range of 1.0.

[0166] As a modified example different from this embodiment, pre-processing such as label smoothing may be performed on the correct answer information calculated by the correct answer information calculation unit 1313. Label smoothing is a method of alleviating the difference between the correct answer label and the incorrect answer label by distributing a probability value subtracted from the correct answer label, which is a hard correct answer label consisting of only "0" or "1", to the incorrect answer label. Therefore, in the modified example, the correct answer information is expressed not as "0" or "1", but as a numerical value including a decimal between "0" and "1". When label smoothing is performed, for example, the above-mentioned row vector [0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0] becomes a row vector [0.01, 0.01, 0.01, 0.01, 0.81, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01]. Other processing is omitted as it is the same as the embodiment. In this modified example, by applying the preprocessing, overlearning can be suppressed and accuracy can be improved.

[0167] Various mold devices are attached to the test injection molding machine 1350 according to this embodiment, and various molding conditions are set. Then, the learning device 1300 generates a teacher data set based on various detection results of molded products produced by the test injection molding machine 1350.

[0168] In this way, the learning device 1300 prepares a teacher data set based on various molding conditions provided in the test injection molding machine 1350 and various mold devices provided in the test injection molding machine 1350. The learning device 1300 performs machine learning using the teacher data set to generate a trained model LM. This makes it possible to perform inference using the trained model LM regardless of the molding conditions and the type of mold device 800.

[0169] Therefore, the control device 700 of the injection molding machine 10 can change the mold clamping force setting value using the trained model LM, regardless of the mold device 800 provided and the molding conditions that are set.

[0170] In the example shown in Figure 8, the input layer defines the variables corresponding to each node as the input variable vector x 0~2199 In the intermediate layer, the variables corresponding to each node are defined as the intermediate variable vector z 0~199 In the output layer, the variables corresponding to the nodes are defined as the output variable vector y 0~19 It is defined as:

[0171] In this case, the intermediate variable vector z0 of the intermediate layer is calculated by the following formula (1). In formula (1), the weight vector w 0~199,0~2199 is the intermediate variable vector z 0~199 The weight vectors for outputting the values ​​of each of the intermediate variables z 1~199 The bias b can also be calculated using a similar formula, so the explanation is omitted. 0~199 is a bias used in the neural network, although it is omitted in Fig. 8. The activation function φ may be, for example, a ReLU function, but other functions may also be used.

[0172] z0 = φ(w 0,0~2199 x 0~2199 +b0)……(1)

[0173] The output variable vector y0 of the output layer is calculated by the following formula (2). In formula (2), the weight vector w' 0~199 is a weight vector for outputting values ​​to the output variable y from each node in the intermediate layer. Note that the output variable vector y 1~19 can also be calculated using a similar formula, so the explanation will be omitted. Although the bias b' is omitted in Fig. 8, it is assumed that it is a bias used in the neural network. The activation function φ' may be, for example, a ReLU function, but other functions may also be used.

[0174] y0 = φ'(w' 0, 0~199 z 0~199 +b')……(2)

[0175] In addition, when performing learning or inference with the trained model LM, the cross-entropy error used for multi-value classification problems is used. This allows each element of the row vector output from the trained model LM to be output as a numerical value between 0 and 1 (hereinafter referred to as confidence) instead of 0 or 1.

[0176] In this embodiment, the output variable vector y 0~19 Here, the number of elements included in the row vector is not limited to 20, but is determined according to the number of intervals into which the range to be adjusted is divided. The number of intervals may be 2 or more.

[0177] In addition, in this embodiment, an example will be described in which the settable range is determined according to the rated clamping force that can be set in the injection molding machine 10, but the present invention is not limited to a method of determining the settable range based on the rating of the injection molding machine 10. Of the settable ranges, a range limited by a predetermined method may be used.

[0178] In addition, the items to be set in this embodiment are not qualitative data (categorical data, nominal scale) but quantitative data (continuous scale). Quantitative data includes both discrete values ​​such as integers and continuous values ​​such as real numbers.

[0179] In other words, in this embodiment, the setting value of the item to be set can be calculated even when a regression model is used, but it is not possible to recognize how confident the trained model is of the setting value calculated when the regression model is used. Therefore, it is difficult for the user to guess whether the setting of the setting value is acceptable or not.

[0180] In contrast, the trained model LM according to this embodiment outputs a confidence level for each interval, allowing the control device 700 or a user to estimate whether an adjustment value is appropriate for each interval.

[0181] The learning unit 1315 stores the generated trained model LM in the trained model memory unit 1322.

[0182] Returning to FIG. 4, the communication control unit 1316 transmits and receives information to and from an external device such as the injection molding machine 10 using the communication I / F 1303. For example, the communication control unit 1316 may transmit the trained model LM stored in the trained model storage unit 1322 to the control device 700 of the injection molding machine 10. The communication control unit 1316 may also extract a structure indicating parameters (e.g., weights and biases) set in each layer constituting the trained model LM stored in the trained model storage unit 1322, and transmit the structure to the control device 700.

[0183] FIG. 4 shows components of the control device 700 of the injection molding machine 10 in the form of functional blocks. Each functional block shown in FIG. 4 is conceptual, and does not necessarily have to be physically configured as shown in the figure. All or a part of each functional block can be functionally or physically distributed and integrated in any unit. All or any part of the processing functions performed by each functional block are realized by a program executed by the CPU 701. Alternatively, each functional block may be realized as hardware using wired logic. As shown in FIG. 4, the CPU 701 of the control device 700 includes a communication control unit 711, an update unit 712, an injection molding processing unit 713, a setting unit 714, and a display control unit 716. In addition, the control device 700 includes a learned model storage unit 721 in the storage medium 702.

[0184] The trained model storage unit 721 stores the trained model LM. The configuration of each layer of the trained model LM is the same as that of the trained model LM stored in the trained model storage unit 1322 of the learning device 1300.

[0185] The communication control unit 711 transmits and receives information to and from an external device such as the learning device 1300 using the communication I / F 705. For example, the communication control unit 711 may receive the trained model LM from the learning device 1300. The communication control unit 711 may also receive information indicating a structure indicating parameters (e.g., weights and biases) set in each layer constituting the trained model LM.

[0186] In the present embodiment, an example of receiving the trained model LM or the information indicating the structure from the learning device 1300 has been described as a method of acquiring the trained model LM or the information indicating the structure, but the method of acquiring the trained model LM or the information indicating the structure is not limited. For example, the trained model LM or the information indicating the structure may be acquired via an external storage medium.

[0187] The update unit 712 updates the trained model LM stored in the trained model storage unit 721 with the received trained model LM or information indicating the structure. In this way, the update unit 712 can match the trained model LM stored in the trained model storage unit 721 with the trained model LM of the learning device 1300.

[0188] When updating is performed using information indicating a structure indicating parameters (e.g., weights and biases) set for each layer constituting the trained model LM, the update unit 712 may be configured as a program for updating the trained model LM. The update unit 712 may determine whether or not the number of layers and the number of nodes per layer defined between the trained model LM stored in the trained model storage unit 721 and the received information indicating the structure match. Then, when it is determined that there is a match, the update unit 712 updates the trained model LM. When it is determined that there is a mismatch, the update unit 712 may not update the trained model LM, and may output an alert screen indicating that the structure is different to the display device 760 or the like.

[0189] The injection molding processing unit 713 executes processing for producing a molded product by the injection molding machine 10. For example, when producing a molded product, the injection molding processing unit 713 may perform injection molding after setting each item constituting the molding conditions.

[0190] The setting unit 714 includes an inference unit 715 and sets a mold clamping force setting value based on information acquired from the inference unit 715 .

[0191] The inference unit 715, for example, inputs waveform data (an example of second data) indicating the results detected by a clamping force detector while a molded product is being produced by the injection molding processing unit 713 to the input layer of the trained model LM, and receives a row vector indicating the confidence level for each interval dividing the range that can be set as the clamping force setting value from the output layer of the trained model LM.

[0192] Furthermore, the inference unit 715 may perform preprocessing on the waveform data before inputting the waveform data to the input layer.

[0193] Then, the setting unit 714 sets the mold clamping force setting value based on the row vector received from the inference unit 715.

[0194] Fig. 9 is a diagram illustrating information based on row vectors output from the inference unit 715 according to this embodiment. In the example shown in Fig. 9, for the range from "1000" to "-1000", the interval label, the center value of the interval, the confidence level, and the cumulative sum of the confidence levels are shown in association with each other for every "100" (a half-open interval).

[0195] The confidence level can be interpreted as the probability that the clamping force adjustment value is included in the corresponding section. In the example shown in Fig. 9, the half-open section [-200, -100) with a confidence level of "0.21" can be interpreted as having the highest probability of including the clamping force adjustment value.

[0196] Then, the setting unit 714 calculates a mold clamping force adjustment value from the row vector output from the inference unit 715, and sets a mold clamping force set value based on the calculated mold clamping force adjustment value.

[0197] For example, the setting unit 714 identifies the section label "[-200,-100)" with the highest degree of certainty based on the degree of certainty for each section. When using the central value, the setting unit 714 sets "-150" as the mold clamping force adjustment value and sets it using the mold clamping force adjustment value and the current mold clamping force setting value or the detected mold clamping force detection value. Note that this embodiment is not limited to the method of setting the central value as the mold clamping force adjustment value, and for example, the setting unit 714 may set the upper limit value or the lower limit value of the section as the mold clamping force adjustment value. Similarly, the subsequent mold clamping force adjustment values ​​are not limited to the central value.

[0198] Furthermore, this embodiment is not limited to a method of determining the mold clamping force adjustment value based on the section with the highest degree of certainty, but may be determined based on a cumulative sum of the degrees of certainty.

[0199] For example, the threshold value of the cumulative sum of the confidence levels may be set to 0.7. In this case, the setting unit 714 identifies the section label "[-100,0)" in which the cumulative sum of the confidence levels for each section is 0.7 or more. When the central value is used, the setting unit 714 sets "-50" as the mold clamping force adjustment value and sets it using the mold clamping force adjustment value and the current mold clamping force setting value or the detected mold clamping force detection value. In this way, by increasing the threshold value of the cumulative sum, the mold clamping force setting value becomes larger. When the mold clamping force setting value becomes larger, it contributes to the safety side with respect to the occurrence of burrs, and therefore the occurrence of burrs can be suppressed.

[0200] In this embodiment, the threshold value of the cumulative sum may be adjustable. A user or a maintenance person can adjust the accuracy of the mold clamping force adjustment value according to the state of the injection molding machine 10 (including the mold device 800). Therefore, by adjusting the threshold value of the cumulative sum, the occurrence of an overload on the mold device 800 due to an inference error in the mold clamping force adjustment value can be suppressed. Or, the occurrence of burrs can be suppressed.

[0201] In addition, in this embodiment, when the setting unit 714 determines the clamping force adjustment value based on the cumulative sum of the confidence levels, the clamping force adjustment value at which the cumulative sum of the confidence levels is exactly the threshold value may be obtained by linear interpolation, rather than the central value of the interval.

[0202] In the row data shown in FIG. 9, when linearly interpolating a mold clamping force adjustment value that is exactly 50% of the threshold value, the setting unit 714 calculates it from the following formula (3).

[0203] -250+{(0.50-0.48) / (0.69-0.48)}×{-150-(-250)}=-240.48[kN]…(3)

[0204] The setting unit 714 is not limited to the above-mentioned method of determining the mold clamping force adjustment value and setting the mold clamping force adjustment value based on the mold clamping force adjustment value. For example, the setting unit 714 may display information output from the trained model LM to the user and determine the mold clamping force adjustment value based on an operation by the user who visually recognizes the information.

[0205] The display control unit 716 displays information on the display device 760. For example, the display control unit 716 may display image information showing the confidence level (an example of the degree to which the answer is correct) for each section.

[0206] Fig. 10 is a diagram showing a first example of a graph showing the confidence level for each section output from the inference unit 715 according to this embodiment. The display control unit 716 according to this embodiment may display image information including the graph shown in Fig. 10 on the display device 760. By referring to the graph shown in Fig. 10, the user can recognize the probability (confidence level) that each section can be used as a mold clamping force adjustment value. Then, the user can adjust the mold clamping force adjustment value based on the confidence level.

[0207] For example, the user may input a mold clamping force adjustment value via the operation device 750 with reference to the image information. Furthermore, the user may set a threshold value for the cumulative sum of confidence with reference to the image information. Furthermore, the user may select a method for setting the mold clamping force adjustment value with reference to the image information. For example, the user may select to set the mold clamping force adjustment value of a section with the highest confidence, or to set the mold clamping force adjustment value of a section with a cumulative sum of confidence equal to or greater than a predetermined threshold.

[0208] The setting unit 714 according to this embodiment may switch whether or not to set a mold clamping force set value (an example of a predetermined item) based on the confidence level for each section. Specifically, the setting unit 714 may switch the setting method for the mold clamping force set value according to the kurtosis representing the confidence level for each section. The kurtosis is an index indicating the sharpness of the distribution of the confidence levels for each section included in the row vector. The calculation method for the kurtosis may be any method, regardless of whether it is a well-known method.

[0209] Fig. 11 is a diagram showing a second example of a graph showing the confidence level for each interval, output from the inference unit 715 according to this embodiment. The example shown in Fig. 11 shows the confidence level for each interval and a cumulative value 2101 of the confidence levels. In the example shown in Fig. 11, it can be seen that the confidence levels are not distributed across multiple intervals and the kurtosis is high, compared to the example shown in Fig. 10.

[0210] In the example shown in Fig. 11, the confidence level is 0.8 or more in the half-open interval "[-100, 0)". In other words, the example shown in Fig. 11 indicates that there is a very high probability that the mold clamping force adjustment value is included in "[-100, 0)".

[0211] Therefore, when the kurtosis is higher than the first standard, the setting unit 714 may identify the section label with the highest confidence level, determine the mold clamping force adjustment value based on the section label, and set the mold clamping force adjustment value based on the current mold clamping force setting value or the detected mold clamping force detection value. Instead of using the kurtosis as the judgment criterion, it may be possible to use whether the highest confidence level is equal to or higher than a predetermined value. The same applies to the following.

[0212] In addition, when the kurtosis is lower than the first standard and higher than the second standard, the setting unit 714 may identify a section label in which the cumulative sum of the confidence levels for each section is greater than or equal to a predetermined value, determine the clamping force adjustment value based on the section label, and set the clamping force adjustment value based on the current clamping force setting value or the detected clamping force detection value.

[0213] In addition, if the kurtosis is lower than the second standard, the setting unit 714 may determine that the values ​​are too dispersed and therefore that the mold clamping force adjustment value should not be set based on the row vector, and may terminate the process of setting the mold clamping force adjustment value.

[0214] In this embodiment, a case where waveform data is input to the trained model LM has been described. However, this embodiment is not limited to the case where waveform data is input to the trained model LM, and collective data showing the results of measuring a molded product after the molded product is injection molded may be input. The collective data may include, for example, measurement results of the amount of burrs at a plurality of predetermined positions on the molded product after injection molding.

[0215] In this embodiment, the clamping force adjustment value is determined in consideration of the confidence level for each section, and the clamping force set value can be set using the clamping force adjustment value. In this embodiment, the quality accuracy of the molded product produced after the setting in the injection molding machine 10 can be improved. In addition, since the clamping force set value can be automatically set, the operation burden on the user and the processing burden on the control device 700 can be reduced.

[0216] In this embodiment, an example of setting the mold clamping force setting value has been described. However, in this embodiment, the setting target is not limited to the mold clamping force setting value. The setting target may be a setting for performing injection molding, and may be quantitative data that needs to be adjusted according to the conditions and environment for performing injection molding. The setting target may be a detection value by a sensor provided in the injection molding machine 10, or may be a time required for a predetermined process of the injection molding machine 10. For example, the setting may be applied to a dwell time, a VP switching position, a dwell pressure setting, a filling speed setting, a filling pressure, or a back pressure setting. Furthermore, the setting information related to metering may be applied to a metering rotation speed, a metering delay, or a metering completion position. Furthermore, the setting information related to mold clamping may be applied to a pressure increase timing, a mold opening / closing speed, or a mold opening position other than the mold clamping force setting value. Furthermore, the setting information related to the ejector device 200 may be applied to an ejector ejection position, an ejector ejection pressure, an ejector speed, or an ejector compression timing. Furthermore, the cylinder temperature setting, nozzle temperature setting, water cooling temperature, and mold temperature may be applied as settings related to the temperature inside the injection molding machine 10. Furthermore, the dimensions of the mold device 800, the characteristics of the molding material, and the like may be applied.

[0217] In this embodiment, data obtained from a sensor of the injection molding machine 10 or an external measuring device (including an imaging device) is pre-processed and then input to the trained model LM, thereby achieving end-to-end setting of predetermined items. Therefore, the processing can be simplified compared to the conventional method. In the above-described embodiment, the processing load can be reduced compared to the conventional method, and since a high-performance CPU or the like is not required, a cost reduction can be achieved.

[0218] In addition, in the above-described embodiment and modified example, when the adjustment value is determined based on the cumulative sum of the confidence levels, the adjustment value can be changed by simply adjusting the threshold value, even though the trained model LM is used. Therefore, the settings can be easily changed according to the situation of the injection molding machine 10 (including the mold device 800).

[0219] (Modification) In the above-described embodiment, an example has been described in which the trained model LM outputs a row vector indicating the degree of certainty for each interval into which the settable range for the item to be set is divided. However, the above-described embodiment does not limit the information output by the trained model LM to information on the item to be set. For example, the trained model LM may output a row vector indicating the degree of certainty that the evaluation is included in each interval into which the range of the evaluation of the molded product is divided. In this modified example, a case in which the weight of the molded product is estimated as the evaluation of the molded product will be described.

[0220] FIG. 12 is a conceptual diagram showing inference performed by a trained model LM mounted on a control device 700 according to this modification.

[0221] The trained model LM for this modified example was generated by machine learning based on waveform data included in a teacher dataset, which indicates the results detected by a clamping force detector, and row vectors indicating, for each section dividing the range that can be detected as the weight of the injection molding machine 10, whether the weight of the molded product produced when the waveform data was detected is included.

[0222] When waveform data 2201 is input, the trained model LM of the control device 700 outputs a row vector 2202 indicating the confidence level for each interval obtained by dividing the weight of the molded product by 1 g.

[0223] The waveform data 2201 may be, for example, data showing a time series of one or more of the position of the screw 330, the detected holding pressure value, and the detected mold clamping force value.

[0224] 12, the confidence level is 0.79, which is the highest value, in the half-open interval [31.0, 32.0). Therefore, the control device 700 may determine the weight of the molded part based on the half-open interval [31.0, 32.0). For example, the control device 700 may determine the center value 31.5 of the half-open interval [31.0, 32.0) as the weight of the molded part.

[0225] In the above embodiment, the elements constituting the row vector output from the trained model LM are the confidence levels (0 to 1) for each interval. However, in this embodiment, the elements constituting the row vector output from the trained model LM are not limited to the confidence levels (0 to 1) for each interval, and may be any information indicating whether the interval is correct or not.

[0226] In this modified example, the evaluation of the molded product is not limited to the weight. For example, the dimensions of the molded product may be used, or any other quantitative data related to the quality characteristics of the molded product may be used.

[0227] Second Embodiment In the above-described embodiment, an example has been described in which the control device 700 of the injection molding machine 10 performs setting using a trained model as a management device of the injection molding machine 10. However, the above-described embodiment is not limited to a method in which the control device 700 of the injection molding machine 10 performs setting using a trained model. Therefore, in the second embodiment, an example will be described in which the group management device 2300 that controls the injection molding machine 10 performs setting using a trained model LM.

[0228] Fig. 13 is a diagram illustrating the configuration of a learning device 1300, a group management device 2300, and an injection molding machine 10 according to this embodiment. As shown in Fig. 13, the group management device 2300 manages, for example, eight injection molding machines 10. Note that the number of injection molding machines to be managed is shown as an example, and may be any number.

[0229] In this embodiment, the setting using the learned model LM by the control device 700 shown in the above-mentioned embodiment is applied to a group management device 2300 having a group management function for multiple injection molding machines 10.

[0230] The learning device 1300 according to this embodiment has the same configuration as that of Embodiment 1. The learning device 1300 transmits information about the trained model LM to the group management device 2300 via the communication line NW.

[0231] The communication line NW is, for example, an Internet communication line. When communication is performed between the learning device 1300 and the group management device 2300, it is preferable to connect them using a VPN (Virtual Private Network). By connecting them using a VPN, the security of communication can be improved.

[0232] The group management device (an example of a management device) 2300 is a device that manages multiple injection molding machines 10 in terms of productivity, and is connected to each injection molding machine 10. Like the control device 700 described above, it receives molding conditions and detection results obtained from various sensors, and assists in managing and planning the production status.

[0233] The group management device 2300 may be realized by, for example, a personal computer. However, although the group management device 2300 does not normally have a control function for the injection molding operation of each injection molding machine 10, the control function can be provided by expanding the function of the personal computer.

[0234] The group management device 2300, like the control device 700, has a storage device (not shown), and the storage device stores the learned model LM.

[0235] The group management device 2300 has a CPU (not shown) like the control device 700, and by having the CPU execute a program stored in a storage device, a communication control unit 711, an update unit 712, an injection molding processing unit 713, a setting unit 714 including an inference unit 715, and a display control unit 716 are realized like the control device 700. The processing executed by each component is the same as in the above-mentioned embodiment, so description thereof will be omitted.

[0236] The group management device 2300 can be connected to the learning device 1300 via a communication line NW.

[0237] The group management device 2300 receives information about the trained model LM from the learning device 1300. Then, the group management device 2300 updates the trained model LM based on the received information.

[0238] As an aid in managing and planning the production status, for example, the group management device 2300 may calculate adjustment values ​​for predetermined items based on waveform data indicated by the received detection results, and transmit a request to set the adjustment values ​​for each of the injection molding machines 10. Also, in the group management device 2300, the display control unit 716 may display image information indicating the confidence level for each section. Then, the user may refer to the displayed image information and perform settings for each of the injection molding machines 10.

[0239] <effect> In the above-described embodiment and modified example, when waveform data is input to the input layer of the trained model LM, a column vector representing a confidence level indicating the degree to which the value included in each interval is correct is output from the output layer. Therefore, by referring to the confidence level, it is possible to recognize how much the value included in the interval can be trusted. For example, the control device 700 or the user can understand whether or not it is acceptable to perform settings based on the values ​​included in the interval by referring to the confidence level. Therefore, it is possible to improve safety when performing settings, processing, or numerical calculations based on values.

[0240] In the above-described embodiment and modified example, the quantitative data is divided into intervals, and the confidence level in each interval is output. Therefore, the classification problem that was handled in the trained model to output the confidence level can be converted back into quantitative data. Therefore, the quantitative data can be used for evaluation of molded products, settings related to injection molding, control, or numerical calculations.

[0241] Although the embodiments of the trained model and the management device for the injection molding machine according to the present invention have been described above, the present invention is not limited to the above-mentioned embodiments. Various changes, modifications, substitutions, additions, deletions, and combinations are possible within the scope of the claims. Naturally, these also fall within the technical scope of the present invention. [Explanation of symbols]

[0242] 10 injection molding machine 700 Control device 701 CPU 702 Storage medium 711 Communication Control Unit 712 Update Department 713 Injection Molding Processing Department 714 Setting Department 715 Reasoning Department 716 Display control unit 721 Trained Model Memory 1300 Learning Device 1301 CPU 1302 Storage medium 1303 Communication Interface 1311 Acquisition Department 1312 Evaluation Department 1313 Correct Answer Information Calculation Unit 1314 Teacher Data Generation Unit 1315 Learning Department 1316 Communication Control Unit 1321 Teacher Data Storage Unit 1322 Trained model memory 2300 Group Management Device

Claims

1. An input layer, one or more intermediate layers connected to the input layer, and an output layer connected to the intermediate layers, machine learning is performed based on first data indicating a result detected while a product is being produced by the injection molding machine or a result measured after the product is produced, and correct answer information indicating whether a value included in each of the first and second intervals is a correct answer for each of the first and second intervals; when second data indicating a result detected during production or a result measured after production is input from the input layer, a degree to which a value included in each of the intervals is correct is output from the output layer, A trained model for making a computer work.

2. The correct answer information classifies a range that can be set as a predetermined item into two or more intervals, and indicates, for each interval, whether or not a value included in the interval is a correct answer as a value to be set in the predetermined item; the output layer outputs, for each of the intervals, a confidence level indicating a degree to which a value included in the interval is correct as a value to be set in the predetermined item. The trained model according to claim 1.

3. The correct answer information classifies a range that can be set as an evaluation of the product into two or more intervals, and indicates, for each interval, whether a value included in the interval is a correct answer as an evaluation of the product; the output layer outputs, for each of the intervals, a confidence level indicating the degree to which a value included in the interval is correct as an evaluation of the product. The trained model according to claim 1.

4. a trained model having an input layer, one or more intermediate layers connected to the input layer, and an output layer connected to the intermediate layers, the trained model being machine-learned based on first data indicating results detected while a product is being produced by an injection molding machine or results measured after the product is produced, and correct answer information dividing a quantity related to the production of the product into two or more intervals and indicating whether or not a value included in each interval is correct; an inference unit that, when second data indicating a result detected during production or a result measured after production is input from the input layer, acquires from the output layer a degree to which a value included in each of the intervals is correct; An injection molding machine management device comprising:

5. A display control unit that displays image information representing a degree to which a value included in each of the intervals is a correct answer, The management device for an injection molding machine according to claim 4 .

6. The correct answer information classifies a range that can be set as a predetermined item into two or more intervals, and indicates, for each interval, whether or not a value included in the interval is a correct answer as a value to be set in the predetermined item; a setting unit that sets the predetermined item based on a confidence level indicating a degree to which a value included in each of the intervals output from the output layer is a correct answer as a value to be set for the predetermined item; The management device for an injection molding machine according to claim 4 .

7. The setting unit sets a value included in the interval to the predetermined item based on a cumulative value of the confidence for each interval. The control device for an injection molding machine according to claim 6.

8. The setting unit switches whether or not to set the predetermined item based on the confidence level for each section. The control device for an injection molding machine according to claim 6.

Images