Molding condition correction apparatus and injection molding machine

Abstract

To provide a molding condition correction apparatus capable of appropriately correcting molding conditions.SOLUTION: A molding condition correction apparatus that corrects molding conditions of an injection molding machine for molding a molded product by injecting molten resin into a mold executes acquisition processing for acquiring a molten state of the molten resin, prediction processing for causing a trained model to predict a molding defect that may occur in the molded product based on the molten state acquired by the acquisition processing, and correction processing for causing the trained model to correct the molding conditions so that the molding defect predicted by the prediction processing is eliminated.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a molding condition correction device that corrects molding conditions of an injection molding machine, and to an injection molding machine equipped with the same. [Background technology]

[0002] In order to stabilize the quality of molded products molded by injection molding machines, it is necessary to properly manage the molten state of the molten resin. Therefore, there is a technique for optimizing molding conditions to bring the molten state of the molten resin closer to the ideal state (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7100011 Summary of the Invention [Problem to be solved by the invention]

[0004] Depending on the combination of the melt state of the molten resin and the injection conditions, various types of molding defects (such as short shots, sink marks, burrs, silver streaks, and cold slugs) can occur in molded products. Depending on which molding defect you are trying to eliminate, the direction of how you should correct the molding conditions (such as increasing the heater temperature or slowing down the screw rotation speed) will be completely different.

[0005] The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a molding condition correction device that can appropriately correct molding conditions. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems, the present invention provides a molding condition correction device that corrects the molding conditions of an injection molding machine that injects molten resin into a mold to form a molded product, and is characterized by performing an acquisition process that acquires the molten state of the molten resin, a prediction process that causes a trained model that has been trained in advance to predict molding defects that may occur in the molded product based on the molten state acquired in the acquisition process, and a correction process that causes the trained model to correct the molding conditions so that the molding defects predicted in the prediction process are eliminated. [Effects of the Invention]

[0007] According to the present invention, molding conditions can be appropriately corrected. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a side view of an injection molding machine according to an embodiment of the present invention. [Figure 2] FIG. 2 is a hardware configuration diagram of an injection molding machine. [Figure 3] 10 is a flowchart of an injection control process. [Figure 4] 10 is a flowchart of a molding condition correction process. [Figure 5] 10 is a screen example of a defect notification screen. [Figure 6] FIG. 1 is a diagram illustrating a neural network included in a trained model. [Figure 7] FIG. 1 is a diagram illustrating the relationship between input data and output data of a trained model. DETAILED DESCRIPTION OF THE INVENTION

[0009] An injection molding machine 10 according to the present invention will be described below with reference to the drawings. The injection molding machine 10 is a device that injects a measured amount of molding material into a mold to form a molded product (hereinafter referred to as "injection molding").

[0010] [Configuration of injection molding machine 10] Fig. 1 is a side view of an injection molding machine 10 according to this embodiment. Fig. 2 is a hardware configuration diagram of the injection molding machine 10. As shown in Figs. 1 and 2, the injection molding machine 10 mainly includes a mold clamping unit 20, an injection unit 30, and a control unit 60.

[0011] The mold clamping device 20 opens, closes, and clamps the mold 21. Specifically, the mold clamping device 20 mainly includes a fixed die plate 23 that supports a fixed-side mold 22, and a movable die plate 25 that supports a movable-side mold 24. The fixed-side mold 22 and the movable-side mold 24 are supported so as to face each other in the left-right direction (horizontal direction) of the injection molding machine 10.

[0012] The movable die plate 25 moves left and right along the tie bars 27 as the driving force of the die opening / closing motor 28 is transmitted through the toggle link mechanism 26. When the movable die plate 25 moves leftward, the fixed die 22 and the movable die 24 move apart. On the other hand, when the movable die plate 25 moves rightward, the fixed die 22 and the movable die 24 come into contact with each other, forming a cavity (internal space) inside the die 21. Then, when pressure is further applied in a direction that moves the movable die plate 25 rightward, the fixed die 22 and the movable die 24 are clamped.

[0013] The injection unit 30 plasticizes, measures, and injects the molding material. The injection unit 30 according to this embodiment is disposed facing the mold clamping unit 20 in the horizontal direction (to the right of the mold clamping unit 20). The injection unit 30 mainly includes a heating cylinder 31, a screw 32, a hopper 33, and a hopper block 34.

[0014] The heating cylinder 31 is a cylindrical member extending in the left-right direction of the injection molding machine 10. The heating cylinder 31 mainly includes a resin passage 35 and a nozzle 36. In addition, a band heater 39 that heats the heating cylinder 31 is attached to the outer circumferential surface of the heating cylinder 31.

[0015] The resin passage 35 is a cylindrical space extending in the axial direction (longitudinal direction) inside the heating cylinder 31. The resin passage 35 communicates with the outside of the heating cylinder 31 (the cavity of the mold 21) through a nozzle 36 provided at the tip (front end) of the heating cylinder 31. In other words, the resin passage 35 is a space extending from the nozzle 36 along the axial direction.

[0016] The screw 32 is a cylindrical member. A groove (hereinafter referred to as a "spiral groove") extending spirally along the longitudinal direction of the screw 32 is formed on the outer circumferential surface of the screw 32. The screw 32 is housed in the internal space of the heating cylinder 31 in a state in which it can move left and right (hereinafter referred to as "forward and backward") and rotate in the injection molding machine 10. The screw 32 in the heating cylinder 31 is configured to be replaceable. In other words, screws 32 with different specifications (for example, material, shape of the spiral groove, volume of the spiral groove) can be inserted into the heating cylinder 31.

[0017] The screw 32 advances and retreats when the driving force of the injection motor 37 is transmitted thereto, and rotates when the driving force of the metering motor 38 is transmitted thereto. More specifically, when the injection motor 37 is rotated forward, the screw 32 moves (advances) toward the tip end of the heating cylinder 31 (i.e., the nozzle 36). On the other hand, when the injection motor 37 is rotated reversely, the screw 32 moves (retreats) toward the base end of the heating cylinder 31 (i.e., the side opposite the nozzle 36).

[0018] Hereinafter, within the range that the tip position of the screw 32 can reach within the heating cylinder 31, the position closest to the nozzle 36 will be referred to as the "forward limit," and the position farthest from the nozzle 36 will be referred to as the "rear limit." Furthermore, the terms "forward rotation" and "reverse rotation" of the injection motor 37 do not specify an absolute direction of rotation, but merely specify a relative relationship (i.e., forward rotation and reverse rotation are rotations in opposite directions).

[0019] The hopper 33 is a funnel-shaped member that stores granular resin as a raw material. The hopper block 34 is a member that supports the heating cylinder 31 and the hopper 33. The hopper 33 is connected to a resin passage 35 through the hopper block 34 on the base end side of the tip of the heating cylinder 31. The granular resin stored in the hopper 33 is supplied to the resin passage 35 of the heating cylinder 31 through an opening provided at the bottom end. The granular resin used in this injection molding machine 10 is, for example, so-called "pellets" molded into a cylindrical shape.

[0020] In the injection device 30, the screw 32 moves backward while rotating by rotating the injection motor 37 in the reverse direction and rotating the metering motor 38. As a result, pellets supplied through the hopper 33 are plasticized and filled (metered) into the resin passage 35 ahead of the screw 32. In addition, in the injection device 30, the injection motor 37 rotates forward to move the screw 32 forward. As a result, the plasticized resin ahead of the screw 32 is injected into the cavity of the mold 21 through the nozzle 36.

[0021] Resins of different types (e.g., different degrees of plasticization) are supplied to the hopper 33 depending on the molded product. The particle size (size of particles) of the pellets supplied to the hopper 33 varies depending on the raw material supply device (not shown) that supplies raw material to the hopper 33. In addition to pellets, recycled resin may also be supplied to the hopper 33. Recycled resin refers to unnecessary parts (runners) separated from the molded product, resin discharged (purged) from the heating cylinder 31, etc. The ratio of pellets and recycled resin supplied to the hopper 33 gradually changes during the injection control process, which will be described later with reference to FIG. 3.

[0022] [Configuration of control device 60] 2, the control device 60 includes a CPU (Central Processing Unit) 61 and a memory 62. The memory 62 is configured, for example, with a ROM (Read Only Memory), a RAM (Random Access Memory), an HDD (Hard Disk Drive), or a combination of these. The control device 60 realizes the processing described below by having the CPU 61 read and execute program code stored in the ROM or HDD. The RAM is used as a work area when the CPU 61 executes the program.

[0023] However, the specific configuration of the control device 60 is not limited to this, and may be realized by hardware such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array).

[0024] The control device 60 controls the overall operation of the injection molding machine 10. More specifically, the control device 60 controls the mold opening / closing motor 28, the injection motor 37, the metering motor 38, the band heater 39, and a communication IF (Interface) 68 based on various signals output from a rotary encoder 64, a load cell 65 (pressure sensor), a state quantity sensor 66, and a display / input device 67.

[0025] The mold opening / closing motor 28, the injection motor 37, and the metering motor 38 are servo motors that generate driving forces to open and close the mold 21, to move the screw 32 back and forth, and to rotate the screw 32, for example, under the control of a servo amplifier (not shown).

[0026] The rotary encoder 64 is a sensor that detects the speed and tip position of the screw 32. More specifically, the rotary encoder 64 outputs pulse signals corresponding to the rotation of the injection motor 37 to the control device 60. The control device 60 then determines the speed of the screw 32 based on the number of pulse signals output per unit time. The control device 60 also determines the tip position of the screw 32 based on the cumulative value of the pulse signals.

[0027] The load cell 65 is a pressure sensor that detects the pressure applied to the screw 32. More specifically, the load cell 65 outputs a pressure signal (voltage value) corresponding to the pressure applied to the screw 32 to the control device 60. Then, the control device 60 identifies the pressure applied to the screw 32 based on the pressure signal output from the load cell 65.

[0028] The state quantity sensor 66 detects the melting state of the molten resin plasticized in the heating cylinder 31 and outputs a state quantity signal indicating the detected melting state (state quantity) to the control device 60. The state quantity sensor 66 may include, for example, at least one of a viscosity sensor that detects the viscosity of the molten resin, a temperature sensor that detects the temperature of the molten resin, and a density sensor that detects the density of the molten resin. The viscosity, temperature, and density of the molten resin are examples of the melting state. The state quantity sensor 66 may also detect the absolute value of the melting state, or may detect the distribution (degree of uniformity) of the melting state within the heating cylinder 31.

[0029] The display input device 67 is a user interface that includes a display (display device) that displays various information to be notified to the operator, and buttons, switches, dials, etc. (input devices) that accept input operations by the operator. The display input device 67 may include a touch panel superimposed on the display. The display input device 67 accepts input operations by the operator and outputs an input signal corresponding to the accepted input operation to the control device 60.

[0030] The control device 60 receives input operations from the operator to input the type of screw 32 inserted into the heating cylinder 31 and molding conditions (described later) via the display input device 67. For example, the control device 60 receives input of a new type of screw 32 every time the screw 32 is replaced. In addition, the control device 60 receives input of new molding conditions every time the molded product to be molded is switched, for example.

[0031] The communication IF 68 is an interface for communicating with the AI ​​server 40 via a communication network. The communication network may be, for example, the Internet, a public line, a wired LAN, a wireless LAN, or a combination of these. The control device 60 can instruct the AI ​​server 40 to perform at least a portion of the processing shown in FIG. 4 via the communication IF 68, and can receive the results of the processing performed by the AI ​​server 40 via the communication IF 68. Details of the AI ​​server 40 will be described later with reference to FIGS. 6 and 7. The functions of the AI ​​server 40 may be implemented in the control device 60 (molding condition correction device).

[0032] The control device 60 may also communicate with the raw material supply device through the communication IF 68. More specifically, the control device 60 may receive the type of resin, the particle size of the pellets, and the proportion of recycled resin supplied to the hopper 33 from the raw material supply device through the communication IF 68. As another example, the control device 60 may accept, through the display / input device 67, an input operation by an operator to input the type of resin, the particle size of the pellets, and the proportion of recycled resin supplied to the hopper 33.

[0033] Furthermore, the control device 60 may include a molding condition correction device. The molding condition correction device is a device that corrects the molding conditions of the injection molding machine 10 by executing the molding condition correction process shown in Fig. 4 based on the melting state of the molten resin detected by the state quantity sensor 66. That is, the molding condition correction device according to this embodiment realizes the molding condition correction process shown in Fig. 4 by reading and executing a correction program stored in the memory 62 by the CPU 61.

[0034] As another example, the molding condition correction device may be realized by a computer independent of the injection molding machine 10. Furthermore, the molding condition correction device may be connected to at least one of the injection molding machine 10 and the AI ​​server 40 via a communication network. Furthermore, the molding condition correction device may realize the molding condition correction process shown in FIG. 4 by reading and executing a correction program stored in the memory of the computer using the CPU of the computer. Furthermore, the molding condition correction device may be implemented in the AI ​​server 40.

[0035] [Information set in memory 62] The memory 62 stores identifiers of a plurality of screws 32 that can be inserted into the heating cylinder 31 in association with the specifications of the screws 32. The specifications of the screws 32 include at least the volume of the spiral groove of the screw 32. In addition, the specifications of the screw 32 may include the material of the screw 32, the shape of the spiral groove, etc.

[0036] The memory 62 also stores molding conditions for injection molding. The molding conditions are operating conditions of the injection molding machine 10 that molds the molded product. In the injection control process of FIG. 3, the control device 60 operates the injection molding machine 10 in accordance with the molding conditions stored in the memory 62. The molding conditions according to this embodiment include a plurality of parameters (e.g., mold opening / closing speed, cooling time, removal time, injection speed, injection stroke, heater temperature, and screw rotation speed). However, the parameters included in the molding conditions are not limited to these.

[0037] The parameter "mold opening / closing speed" is the speed (mm / s) at which the mold 21 switches from one of the mold open state and mold clamped state to the other. The parameter "cooling time" is the waiting time (sec) until the mold 21 is opened after the molten resin is injected into the mold 21. The molten resin injected into the cavity of the mold 21 solidifies during this cooling time to become a molded product. The parameter "removal time" is the time (sec) required for a robot arm (not shown) to remove the molded product from the open mold 21.

[0038] The parameter "injection speed" is the forward speed (mm / s) of the screw 32 during the injection process. The parameter "injection stroke" is the forward distance (mm) of the screw 32 during the injection process. The parameter "heater temperature" is the temperature (°C) of the band heater 39 that heats the heating cylinder 31. The parameter "screw rotation speed" is the rotation speed (rpm) of the screw 32 during the metering process.

[0039] The parameter "screw rotation speed" is a parameter that affects the time (metering time) required to meter a predetermined amount of molten resin. More specifically, the metering time (sec) required to meter the same amount of molten resin becomes shorter as the set value of the parameter "screw rotation speed" becomes larger (the rotation speed of the screw 32 becomes faster), and becomes longer as the set value of the parameter "screw rotation speed" becomes smaller (the rotation speed of the screw 32 becomes slower). Furthermore, the metering time refers to the time from the start of metering to the completion of metering.

[0040] [Injection control processing] 3 is a flowchart of the injection control process. The injection control process is a process for molding a molded product by injecting molten resin filled in the heating cylinder 31 into the cavity of the clamped mold 21. The control device 60 starts the injection control process in response to, for example, a molding instruction being input through the display input device 67. The molding instruction includes, for example, the number of molded products to be molded (hereinafter referred to as the "molding number").

[0041] At the start of the injection control process, the mold 21 is opened, the molten resin to be injected next is measured into the space ahead of the screw 32 of the heating cylinder 31, and the spiral groove of the screw 32 on the tip side of the hopper 33 is filled with resin (pellets, resin in the process of plasticization, molten resin).

[0042] First, the control device 60 closes and clamps the mold 21 by rotating the mold opening / closing motor 28 in accordance with the molding condition "mold opening / closing speed" (S11). This forms a cavity in the mold 21. The processing of step S11 is an example of mold clamping processing.

[0043] Next, after the mold clamping process (S11) is completed, the control device 60 rotates the injection motor 37 in the forward direction in accordance with the molding conditions "injection speed" and "injection stroke" to move the screw 32 forward (S12). As a result, the molten resin measured in the region ahead of the screw 32 in the heating cylinder 31 is injected into the cavity of the mold 21. The process of step S12 is an example of the injection process.

[0044] Next, after the injection process (S12) is completed, the control device 60 rotates and retracts the screw 32, thereby plasticizing the granular resin supplied to the heating cylinder 31 through the hopper 33 and measuring the molten resin to be injected next into the space ahead of the screw 32 of the heating cylinder 31. The process of step S13 is an example of a measuring process.

[0045] Furthermore, after the injection process (S12) is completed, the control device 60 cools the molten resin in the cavity in accordance with the molding condition "cooling time" (S14) in parallel with the measurement process (S13). As a result, the molten resin is cooled in the mold 21, and a molded product is formed. At this time, the control device 60 may circulate cooling water through cooling water channels provided in the mold 21. The process of step S14 is an example of a cooling process.

[0046] Next, when both the weighing process (S13) and the cooling process (S14) are completed (S14: Yes), the control device 60 opens the mold 21 by rotating the mold opening / closing motor 28 in accordance with the molding condition "mold opening / closing speed," and causes the robot arm to remove the molded product from the opened mold 21 in accordance with the molding condition "removal time" (S15). The process of step S15 is an example of the removal process.

[0047] Next, after the measurement process (S15) is completed, the control device 60 determines whether the number of molded articles molded in the injection control process has reached the molding quantity specified in the molding instruction (S16). If the number of molded articles has not reached the molding quantity (S19: No), the control device 60 executes the molding condition correction process (S17) shown in Fig. 4 and executes the processes from step S11 onwards again. That is, the control device 60 repeatedly executes the processes of steps S11 to S15 and S17 until the number of molded articles reaches the molding quantity.

[0048] Furthermore, the control device 60 ends the injection control process when the number of molded articles reaches the molding quantity (S16: Yes). As another example, the control device 60 may end the injection control process when an operator's operation to instruct the end of the injection control process is received through the display input device 67 (S16: Yes).

[0049] [Molding condition correction processing] Fig. 4 is a flowchart of the molding condition correction process. Fig. 5 is an example of a defect notification screen. The molding condition correction process predicts molding defects in the molded product from the melting state of the molten resin in the heating cylinder 31, and corrects the molding conditions so as to eliminate the predicted molding defects.

[0050] First, the control device 60 acquires the melting state of the molten resin (for example, at least one of the viscosity, temperature, and density of the molten resin) from the state quantity sensor 66 (S21). The process of step S11 is an example of the acquisition process.

[0051] Furthermore, in step S21, the control device 60 may further acquire some or all of the machine state of the injection molding machine 10 (e.g., specifications of the screw 32), the mold state of the mold 21 attached to the mold clamping device 20 (e.g., cavity volume), and the resin state of the resin supplied to the hopper 33 (e.g., type of resin, pellet particle size, proportion of recycled resin). The machine state and mold state are stored in, for example, the memory 62. Furthermore, the resin state is acquired from the raw material supply device via, for example, the communication IF 68.

[0052] Next, the control device 60 causes the predictively trained model 41 to predict molding defects of the molded product (S22) based on the melting state acquired in the acquisition process (S21). The predictively trained model 41 predicts the types of molding defects (e.g., short shots, sink marks, flash, silver streaks, cold slugs) that may occur in the molded product molded in the mold 21 when the molten resin in the molten state acquired in the acquisition process is injected into the mold 21 under the current molding conditions. The control device 60 may also cause the predictively trained model 41 to predict the probability of occurrence of molding defects (e.g., 0 to 100%, high, medium, low). The process of step S22 is an example of a prediction process. Details of the process of step S22 will be described later with reference to FIGS. 6 and 7(A).

[0053] The number of molding defects predicted in step S22 may be one, two or more, or 0. In the prediction process according to this embodiment, the following process will be explained assuming that the occurrence of short shots (probability of occurrence: 70%), burrs (probability of occurrence: 60%), and cold slugs (probability of occurrence: 25%) is predicted as molding defects.

[0054] Next, if one or more molding defects are predicted in the prediction process (S22) (S23: Yes), the control device 60 causes the display / input device 67 to display a defect notification screen shown in Fig. 5(A) (S24). The defect notification screen is a screen that notifies the operator of molding defects predicted in the prediction process and allows the operator to specify molding defects to be resolved and their priority. The process of step S24 is an example of display processing.

[0055] The defect notification screen displays a list of molding defects predicted in the prediction process, corresponding to their occurrence probabilities, as shown in Fig. 5(A), for example. The defect notification screen also includes check boxes associated with each of the listed molding defects, an "Up" icon, a "Down" icon, a "Correct" icon, and a "Do Not Correct" icon. The predicted molding defects are displayed on the defect notification screen in descending order of their occurrence probability. Among molding defects with checked check boxes, the higher the molding defect is displayed, the higher the priority for rectifying the defect.

[0056] The check boxes accept operator operation to specify the corresponding molding defect. The [Up] icon accepts operator operation to raise the priority of the corresponding molding defect. The [Down] icon accepts operator operation to lower the priority of the corresponding molding defect. The [Correct] icon accepts operator operation to instruct the molding conditions to be corrected so that the molding defect whose check box is checked is resolved. The [Do not correct] icon accepts operator operation to instruct the molding conditions not to be corrected.

[0057] Next, the control device 60 accepts the operator's operation on the defect notification screen through the display input device 67 (S25). That is, the control device 60 allows the operator to specify some or all of the molding defects listed on the defect notification screen using check boxes. The control device 60 also allows the operator to specify the priority for eliminating the specified molding defects using the [Up] icon and the [Down] icon. The processing of step S25 is an example of the specification processing.

[0058] The operator checks the checkbox corresponding to the molding defect to be resolved. The operator also taps the [Up] icon or the [Down] icon to change the priority of the molding defect to be resolved. In this embodiment, for example, as shown in FIG. 5(B), two molding defects are specified with the order of priority: "burr" and "short shot." The operator then taps the [Correct] icon shown in FIG. 5(B).

[0059] Next, when the control device 60 receives the operator's operation of tapping the "Correct" icon through the display input device 67 (S25: Yes), it causes the corrected trained model 42 to correct the molding conditions so as to eliminate the specified molding defects (S26). The corrected trained model 42 corrects the current molding conditions so as to eliminate the molding defects in the order of priority specified by the operator. The processing of step S26 is an example of a correction process. Details of the processing of step S26 will be described later with reference to FIGS. 6 and 7(B).

[0060] On the other hand, when the control device 60 receives the operator's operation of tapping the [Do not correct] icon through the display input device 67 (S25: No), it skips the processing of step S26 (i.e., does not correct the molding conditions) and ends the molding condition correction processing. Also, when no molding defect is predicted in step S22 (S23: No), the control device 60 skips the processing from step S24 onwards and ends the molding condition correction processing.

[0061] 3, the control device 60 executes the processes from step S11 onward again using the molding conditions corrected in step S17. Then, by repeatedly executing the processes of steps S11 to S17, the molding conditions are gradually corrected so as to eliminate molding defects, and the molten state of the molten resin in the heating cylinder 31 approaches an ideal state. Note that the execution timing of step S17 is not limited to the example in FIG. 3.

[0062] [AI Server 40 Processing] Fig. 6 is a diagram showing the neural networks included in the trained models 41 and 42. Fig. 7 is a diagram showing the relationship between input data and output data of the predictive trained model 41 (A) and the corrected trained model 42 (B).

[0063] The AI ​​server 40 is realized by, for example, a workstation or a general-purpose computer such as a personal computer. The AI ​​server 40 realizes AI (Artificial Intelligence) including trained models 41 and 42. The AI ​​installed in the AI ​​server 40 is a so-called "generative AI" that processes input data and generates output data. The AI ​​installed in the AI ​​server 40 generates output data from input data using, for example, a neural network shown in FIG. 6.

[0064] The AI ​​server 40 according to this embodiment includes, for example, a predictive trained model 41 and a corrected trained model 42 (hereinafter, these may be collectively referred to as "trained models 41, 42"). The predictive trained model 41 is a trained model that executes the prediction process (S22). The corrected trained model 42 is a trained model that executes the correction process (S26). However, the predictive trained model 41 and the corrected trained model 42 may be integrated into one. Furthermore, it is not necessary for the AI ​​server 40 to execute all of the processes of steps S22 and S26. In other words, at least a portion of steps S22 and S26 may be executed by the AI ​​server 40, and the other portions may be executed by the control device 60.

[0065] As shown in Figure 6, the neural network is composed of an input layer L1 consisting of multiple nodes I1, I2, and I3, a hidden layer L2 consisting of multiple nodes H1, H2, H3, and H4, and an output layer L3 consisting of multiple nodes O1, O2, and O3. In the example of Figure 6, the number of nodes in the input layer L1 and the output layer L3 is the same, but the number of nodes in the input layer L1 and the output layer L3 may be different. The neural network may also have multiple hidden layers L2. Furthermore, Figure 6 shows a fully connected neural network in which multiple nodes in each layer L1, L2, and L3 are connected to all nodes in adjacent layers, but the structure of the neural network is not limited to this.

[0066] The trained models 41 and 42 are generated by inputting multiple pieces of training data, including input data and correct answer data, into a pre-training model (hereinafter referred to as the "pre-training model"). The input data refers to the data input to the pre-training model. The correct answer data refers to the data that should be output when the input data is input. Then, by inputting multiple pieces of training data into the pre-training model, the neural network is optimized to become the trained models 41 and 42. The training data is generated based on, for example, the results of experiments, simulations, or injection molding performed by the injection molding machine 10. This process is an example of a training process in which the weight coefficients and biases of each node are adjusted so that correct answer data is output from the output layer L3 when input data is input to the input layer L1.

[0067] As an example, the molten state of the molten resin is used as input data, and training data, in which molding defects that may occur when the molten resin is injected in this molten state, are used as correct data, are input to a pre-training model, thereby generating a predictive trained model 41. The input data may further include some or all of the machine state, mold state, resin state, and molding conditions. Furthermore, the output data may further include the probability of occurrence of molding defects.

[0068] As another example, one or more molding defects and molding conditions are used as input data, and training data in which molding conditions corrected to eliminate the input molding defects are used as corrected data is input to the pre-training model, thereby generating the corrected trained model 42. The input data may further include some or all of the priority of the molding defects to be eliminated, the machine condition, the mold condition, and the resin condition.

[0069] Furthermore, the learning process may be performed not only on the pre-learning model but also on the trained models 41 and 42. Furthermore, the AI ​​does not need to learn using the input data and correct answer data actually used in the present invention, but may learn using general-purpose training data. Furthermore, the learning process performed by the AI ​​is not limited to "supervised learning" in which input data and correct answer data are input, but may also be "unsupervised learning" in which correct answer data is not input, or may be reinforcement learning, transfer learning, or the like.

[0070] Furthermore, the AI ​​server 40 generates and outputs output data by inputting input data into the neural networks of the trained models 41 and 42. This process is an example of a generation process in which input data input to the input layer L1 is processed using weighting coefficients and biases adjusted in advance for each node to generate output data and output it from the output layer L3.

[0071] As an example, in step S22, the control device 60 inputs the melting state acquired in step S21 as input data to the predictive trained model 41, and outputs the types of molding defects that may occur as output data. Note that the control device 60 may further include some or all of the machine state, mold state, resin state, and molding conditions in the input data. Furthermore, the predictive trained model 41 may further output the probability of molding defects occurring as output data.

[0072] As another example, in step S26, the control device 60 inputs the molding defect and molding conditions specified by the operator in step S25 as input data to the predictive trained model 41, and outputs the corrected molding conditions as output data. Note that the control device 60 may further include some or all of the priority of the molding defect to be resolved, the machine condition, the mold condition, and the resin condition in the input data.

[0073] [Effects of the embodiment] According to the above embodiment, molding defects that may occur in a molded product are predicted based on the molten state of the molten resin, and molding conditions are corrected to eliminate the predicted molding defects. In this way, by clarifying the direction in which the molding conditions are corrected, the molding conditions can be appropriately corrected. As a result, molding conditions in which multiple parameters interact in a complex manner can be optimized.

[0074] Furthermore, according to the above embodiment, by having the operator specify the molding defects to be eliminated, it is possible to clarify the direction for correcting molding conditions depending on the type of molded product, the quality required for the molded product, etc. Furthermore, by having the operator further specify the priority of the molding defects to be eliminated, the direction for correcting molding conditions becomes even clearer.

[0075] However, it is not limited to the operator to specify the molding defects to be eliminated from the predicted molding conditions and their priorities. As another example, the control device 60 may specify the molding defect with the highest probability of occurrence. As yet another example, the control device 60 may specify the molding defect with an occurrence probability equal to or greater than a threshold value (e.g., 50%). Note that when multiple molding defects are specified, the control device 60 may prioritize them in descending order of occurrence probability.

[0076] Furthermore, according to the above embodiment, by providing the predictive trained model 41 and the corrective trained model 42, appropriate training data can be provided to each of them for training. As a result, the accuracy of the predictive process (S22) and the corrective process (S26) is improved.

[0077] In addition, when the molding condition correction device is realized by a computer independent of the control device 60, the molding condition correction device may acquire the melt state from the injection molding machine 10 through the communication IF in step S21. As another example, the molding condition correction device may predict (acquire) the melt state through a simulation using some or all of the input molding conditions, machine state, mold state, and resin state. Furthermore, in step S26, the molding condition correction device may transmit the corrected molding conditions to the injection molding machine 10 through the communication IF.

[0078] The above-described embodiments are merely illustrative examples of the present invention, and are not intended to limit the scope of the present invention to these embodiments. Those skilled in the art can implement the present invention in various other forms without departing from the spirit of the present invention. [Explanation of symbols]

[0079] 10...injection molding machine, 20...mold clamping device, 21...mold, 22...fixed side mold, 23...fixed die plate, 24...movable side mold, 25...movable die plate, 26...toggle link mechanism, 27...tie bar, 28...mold opening / closing motor, 30...injection unit, 31...heating cylinder, 32...screw, 33...hopper, 34...hopper block, 35...resin passage, 36...nozzle, 37...injection motor, 38...metering motor, 39...band heater, 40...AI server, 41...prediction trained model, 42...correction trained model, 60...control device, 61...CPU, 62...memory, 64...rotary encoder, 65...load cell, 66...state quantity sensor, 67...display input device, 68...communication IF

Claims

1. A molding condition correction device that corrects molding conditions of an injection molding machine that injects molten resin into a mold to form a molded product, an acquisition process for acquiring a melting state of the molten resin; a prediction process in which a trained model that has been trained in advance predicts molding defects that may occur in the molded product based on the melting state acquired in the acquisition process; A molding condition correction device characterized by executing a correction process in which the trained model corrects the molding conditions so as to eliminate the molding defects predicted in the prediction process.

2. 2. The molding condition correction device according to claim 1, When a plurality of molding defects are predicted in the prediction process, a display process for displaying a list of the plurality of molding defects predicted by the prediction process; and executing a designation process for allowing an operator to designate some of the plurality of molding defects listed in the display process; A molding condition correction device characterized in that, in the correction process, the trained model corrects the molding conditions so that the molding defects specified in the specification process are eliminated.

3. 3. The molding condition correction device according to claim 2, In the prediction process, the trained model is caused to further predict the probability of occurrence of the molding defect; The molding condition correcting device is characterized in that, in the display processing, the molding defects predicted in the prediction processing and the occurrence probabilities are displayed in association with each other in a list.

4. 3. The molding condition correction device according to claim 2, In the designation process, the operator is further prompted to designate priorities of the designated plurality of molding defects; A molding condition correction device characterized in that, in the correction process, the trained model corrects the molding conditions so that the molding defects are eliminated in the priority order specified in the designation process.

5. 2. The molding condition correction device according to claim 1, The trained model is A predictive trained model that executes the prediction process; A molding condition correction device characterized by including a correction trained model that executes the correction process.

6. a mold clamping device that opens, closes, and clamps the mold; an injection device that injects molten resin into the cavity of the clamped mold; 10. An injection molding machine comprising: the molding condition correction device according to claim 1; and a control device that controls the mold clamping device and the injection device in accordance with the molding conditions corrected by the molding condition correction device.

Images