Adjustment device, injection molding machine, and program

Abstract

To provide a system in which an adjustment method with higher utilization effect is applied in adjustment of a molding condition in an injection molding apparatus.SOLUTION: An adjustment device that adjusts a molding condition of an injection molding machine that performs injection molding on the basis of a molding condition comprises: an adjustment quantity calculation unit 130 that uses a model for an object molding condition being a molding condition to be adjusted and an object molding quality being a molding quality to be adjusted, and a coefficient according to the model, to calculate an adjustment quantity for the object molding condition on the basis of a target molding quality being a molding quality to be targeted and the object molding quality; and a molding condition adjustment unit for adjusting the object molding condition on the basis of the adjustment quantity calculated by the adjustment quantity calculation unit 130.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] This invention relates to an adjustment device, an injection molding machine, and a program. [Background technology]

[0002] Patent Document 1 discloses a molding condition determination support device that uses a first learning model to acquire a molding condition adjustment amount, which is a value corresponding to the difference between molding condition data detected by a sensor and a target value of molding condition data, and uses a second learning model to acquire an adjustment amount for a molding condition element corresponding to the molding condition adjustment amount, and adjusts the molding condition element based on the acquired adjustment amount.

[0003] Patent Document 2 discloses a molding optimization method in which a data processing unit is configured to set, for molding data relating to input parameters and output parameters, the molding data during the molding process and evaluation information relating to molding quality as constraint conditions, a prediction function based on the learning of a neural network as an objective function, and an optimization processing program that finds optimized molding conditions relating to input parameters that satisfy the said constraint conditions and objective function. During production operation, the data processing unit detects molding data relating to output parameters during the molding process, the optimization processing program finds optimized molding conditions based on the molding data relating to said output parameters, and the existing molding conditions are changed according to the found molding conditions. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2020-49929 [Patent Document 2] Japanese Patent Publication No. 2017-119425 [Overview of the project] [Problems that the invention aims to solve]

[0005] In injection molding, machine learning models can be used to calculate adjustments to molding conditions based on variations in molding quality, in response to multivariate inputs and outputs. However, there was still room for further improvement in adjusting molding conditions based on adjustments calculated using machine learning models.

[0006] The present invention aims to provide a system that applies a more efficient adjustment method to the molding conditions in an injection molding apparatus. [Means for solving the problem]

[0007] One aspect of the present invention is an adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, characterized in that it comprises: an adjustment amount calculation unit that calculates an adjustment amount for the target molding conditions based on the target molding quality and the target molding quality, using a model relating to the target molding conditions, which are the molding conditions to be adjusted, and the target molding quality, which is the molding quality to be adjusted, and coefficients corresponding to this model; and a molding condition adjustment unit that adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit. [Effects of the Invention]

[0008] According to one aspect of the present invention, the burden required for calculating the adjustment amount when adjusting the molding conditions in an injection molding apparatus can be reduced. [Brief explanation of the drawing]

[0009] [Figure 1] This diagram shows the configuration of an injection molding machine to which this embodiment is applied. [Figure 2] This is a diagram showing the configuration of the control device. [Figure 3] This diagram shows the configuration of a data processing device. [Figure 4] This figure shows an example of the hardware configuration of a control device and a data processing device. [Figure 5] This diagram shows the functional configuration of the adjustment amount calculation unit. [Figure 6] It is a diagram showing the functional configuration of an adjustment amount calculation unit when correcting an adjustment amount. [Figure 7] It is a diagram explaining the reliability when adjusting molding conditions using a target condition difference. FIG. 7(A) is a diagram showing the predicted distribution of molding quality when the reliability of the target condition difference is high, and FIG. 7(B) is a diagram showing the predicted distribution of molding quality when the reliability of the target condition difference is low. [Figure 8] It is a diagram showing an example of a function for converting an RMSE model residual into a relaxation coefficient α. [Figure 9] It is a diagram showing a probability distribution β0 and a probability distribution β. [Figure 10] It is a diagram showing an example of a function used for converting an improvement degree γ. DESCRIPTION OF EMBODIMENTS

[0010] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. <APPARATUS CONFIGURATION> FIG. 1 is a diagram showing the configuration of an injection molding machine to which the present embodiment is applied. The injection molding machine 10 includes an injection device 20, a mold clamping device 30, a control device 100, a data processing device 200, and a display device 300. In the following description, the direction from the injection device 20 toward the mold clamping device 30 may be referred to as the front direction.

[0011] The injection device 20 includes a cylinder that heats a molding material, a screw provided rotatably in the cylinder and capable of advancing and retracting in the axial direction, a rotary motor that drives the screw in a rotational direction, a motor that drives the screw in the axial direction, and the like. The molding material is, for example, resin or the like. The injection device 20 injects the molding material that has been heated and liquefied in the cylinder by advancing the screw forward while rotating the screw, and fills the molding material into the mold of the mold clamping device 30 disposed in front of the injection device 20. The injection device 20 performs, for example, a metering step, a filling step, a pressure holding step, and the like in the manufacturing process of a molded product. The filling step and the pressure holding step are collectively also referred to as an injection step.

[0012] The mold clamping device 30 comprises a mold, a clamping mechanism for clamping the mold, and a motor for driving the clamping mechanism. The mold clamping device 30 closes the mold and receives the molding material injected from the injection device 20 into the mold. At this time, the mold clamping device 30 clamps the mold with the clamping mechanism to prevent the mold from opening as the molding material fills it (mold clamping). A molded product is produced when the molding material filled into the mold solidifies. After this, the mold clamping device 30 opens the mold and sends out the produced molded product. In the manufacturing process of a molded product, the mold clamping device 30 performs, for example, a mold closing process, a pressurization process, a mold clamping process, a depressurization process, and a mold opening process.

[0013] The control device 100 is a device that controls the operation of the injection device 20 and the clamping device 30. The data processing device 200 is a device that processes data obtained as the injection device 20 and the clamping device 30 operate. The display device 300 displays information related to the control of the injection device 20 and the clamping device 30 by the control device 100, data acquired by the data processing device 200, and the processing results of the data processing device 200. The display device 300 also displays an operation screen for inputting commands and data to the control device 100 and the data processing device 200.

[0014] <Configuration of control device 100> Figure 2 shows the configuration of the control device 100. The control device 100 controls the operation of the injection device 20 and the clamping device 30. The control device 100 is implemented, for example, by a computer. The control device 100 comprises a control unit 110, a molding condition adjustment unit 120, an adjustment amount calculation unit 130, and a storage unit 140. The control device 100 controls the injection device 20 and the clamping device 30 to repeatedly manufacture molded products by repeatedly performing the processes related to the manufacture of molded products. The processes related to the manufacture of molded products include a metering process, a mold closing process, a pressurizing process, a mold clamping process, a filling process, a holding pressure process, a cooling process, a depressurizing process, a mold opening process, and an ejection process. Hereinafter, these manufacturing processes may be collectively referred to as the "manufacturing process." Also, a series of operations to obtain a molded product, for example, the operations from the start of one metering process to the start of the next metering process in the above manufacturing process, may be referred to as a "shot," "molding cycle," etc. Note that the above processes for manufacturing molded products are merely examples. For example, the process performed in a single shot may include other processes not listed above.

[0015] The control unit 110 controls the injection unit 20 and the clamping unit 30 based on control information. The control information consists of conditions set by the user, which are input by the user using, for example, an input device (not shown). The control information includes molding conditions such as resin temperature, mold temperature (cylinder temperature), injection holding pressure time, metering value, VP switching position, holding pressure, injection speed (filling speed), screw rotation speed, screw back pressure, and clamping force. Multiple combinations of these molding conditions are determined depending on the molded product and mold. This combination data of molding conditions will be referred to below as the molding condition dataset. The molding condition dataset is prepared according to the type of molded product and mold, etc., and stored in the storage unit 140.

[0016] The control unit 110 controls the injection unit 20 and the clamping unit 30 using the above-mentioned molding condition dataset, and carries out the processes related to the manufacture (shot) of molded products, including the above-mentioned steps. At the start of manufacturing of a molded product, the control unit 110 reads the molding condition dataset corresponding to the molded product to be manufactured from the storage unit 140. Then, the control unit 110 controls the operation of the injection unit 20 and the clamping unit 30 based on the control information read. Specifically, the control unit 110 controls the injection unit 20 and the clamping unit 30 so that the data obtained from the injection unit 20 and the clamping unit 30 during the manufacturing process matches the set values ​​in the molding condition dataset. The control unit 110 may also display the molding condition dataset read from the storage unit 140 on the display device 300. The user may refer to the molding condition data displayed on the display device 300 and perform operations such as correcting the values ​​as necessary.

[0017] The molding condition adjustment unit 120 adjusts the molding conditions used in the control of the injection device 20 and the clamping device 30 by the control unit 110. As the manufacturing process of molded products is repeated, the state of the injection device 20 and the clamping device 30 changes, and this change in the state of the devices 20 and 30 affects the quality of the molded product (hereinafter referred to as "molding quality"). Therefore, in order to maintain molding quality during the operation of mass production of molded products, the molding condition adjustment unit 120 automatically adjusts the molding conditions.

[0018] The molding condition adjustment unit 120 adjusts the molding conditions used by the control unit 110 to control the injection device 20 and the clamping device 30, based on the information of the adjustment amount of the molding conditions calculated by the adjustment amount calculation unit 130. The molding condition adjustment by the molding condition adjustment unit 120 may be performed after every shot, every few shots, after a predetermined period of time has elapsed, or when the molding quality satisfies the adjustment conditions. The molding quality is indicated by data (quality information) that is identified as representing the molding quality from the manufacturing data of the molded product acquired by the data processing device 200 from the injection device 20 and the clamping device 30. When the molding quality satisfies the adjustment conditions, for example, it may be when the data representing the molding quality acquired by the data processing device 200 exceeds a predetermined threshold.

[0019] The adjustment amount calculation unit 130 calculates the amount of adjustment to the molding conditions performed by the molding condition adjustment unit 120. Machine learning and other mathematical optimization methods are used to calculate the adjustment amount. Here, the adjustment amount calculation unit 130 uses a model based on the difference in molding conditions and the difference in molding quality, rather than the values ​​of the molding conditions and molding quality themselves, as a model for obtaining the adjustment amount of the molding conditions. The difference in molding conditions is the difference between the molding conditions to be adjusted (hereinafter referred to as "target molding conditions") and the standard molding conditions (hereinafter referred to as "standard molding conditions"). In this embodiment, the target molding conditions are, for example, the molding conditions used to manufacture the current molded product (in other words, the last molded product manufactured) (hereinafter referred to as "current molding conditions"). Here, as the current molding conditions, the molding conditions corresponding to the last molded product from which quality information was obtained may be used, or if the molding conditions have been changed or adjusted after the last molded product was manufactured, the changed or adjusted molding conditions may be used. The difference in molding quality is the difference between the quality information of the standard molding quality (hereinafter referred to as "standard molding quality") and the molding quality to be adjusted. Here, the molding quality to be adjusted is quality information of the molding quality manufactured based on the molding conditions to be adjusted, and may be the quality information of the current molding quality of the molded product (hereinafter referred to as "current molding quality"). In this embodiment, the difference is not only the difference between numerical values ​​obtained by subtraction in arithmetic operations, but can be anything that represents the difference or gap between two values, and broadly includes values ​​obtained by division or functions, for example. Details of the calculation method of the molding condition adjustment amount by the adjustment amount calculation unit 130 will be described later.

[0020] Here, the reference molding conditions are the molding conditions identified as a comparison point to the current molding conditions, and specifically, are the molding conditions used in a shot prior to the shot in which the current molding conditions were used (the last shot performed). The choice of which shot's molding conditions to use as the reference molding conditions can be set individually in the actual equipment and is not particularly limited. For example, the molding conditions used in the shot immediately preceding the shot in which the current molding conditions were used may be used as the reference molding conditions, or the molding conditions used in a predetermined number of shots prior may be used as the reference molding conditions. Reference molding quality is information that represents the quality of the molded product produced in a shot in which the reference molding conditions were used. Hereinafter, the reference molding conditions and reference molding quality may be collectively referred to as reference information, reference values, etc.

[0021] The model used to calculate the adjustment amount for molding conditions is a regression model that represents the relationship between the difference in molding conditions and the difference in molding quality. Various regression models, including neural networks, can be used for this model. For example, a simple linear regression model may be used, or a probabilistic model such as Bayesian multiple regression or a Gaussian process may be used. The model may be subjected to preprocessing such as standardization, which is common when creating regression models. The model is updated according to predetermined update conditions. The update conditions may be, for example, that the molding condition adjustment unit 120 has made adjustments a predetermined number of times (one or more times), or that the model be updated periodically. The model may be updated by sequential learning, or by accumulating information on the differences over several times and using the accumulated difference data as training data for learning.

[0022] By using a multivariate input / output model and multi-objective optimization to calculate the amount of adjustment to the molding conditions, it is possible to handle one or more items in each case. Furthermore, by using a model that represents the relationship between the difference in molding conditions and the difference in molding quality, unlike when using a model that represents the relationship between molding conditions and molding quality, it is only necessary to create a model that represents the relationship between the molding conditions and molding quality items set according to the molded product and mold. This reduces the processing load and realizes a highly versatile system that can be applied to various molded products and molds. In principle, during mass production when molding conditions are constant, adjustments to the molding conditions are made to suppress changes in molding quality caused by disturbances such as changes in room temperature. For this reason, it is not appropriate to use the relationship between molding conditions and molding quality in the model to calculate the amount of adjustment to the molding conditions.

[0023] The memory unit 140 stores control information 141 used to control the injection unit 20 and the clamping unit 30. The molding condition dataset included in the control information 141 is prepared in association with the molded product to be manufactured and the mold. The memory unit 140 stores molding condition datasets for each molded product to be manufactured and the mold. The memory unit 140 also stores reference information 142 used to calculate the adjustment amount of the molding conditions. The reference information 142 includes reference molding conditions and reference molding quality. The memory unit 140 also stores quality information 143 for the target molding quality (hereinafter referred to as "target molding quality"). The target molding quality is the molding quality that is not evaluated as defective. If there is a range of molding quality that is not evaluated as defective (in other words, if a range is allowed for the value of the target molding quality), a representative value such as the average value may be used as the target molding quality. The target molding quality is set by the user, for example, and input using the input means described later, and stored in the memory unit 140.

[0024] Although not shown in the diagram, the memory unit 140 holds programs for the control unit 110 to control the injection unit 20 and the clamping device 30, programs for the molding condition adjustment unit 120 to adjust the molding conditions, and programs and models for the adjustment amount calculation unit 130 to calculate the adjustment amount of the molding conditions. As will be described in more detail later, the processor in the control device 100 reads and executes the programs held in the memory unit 140, thereby realizing the functions of the control unit 110, the molding condition adjustment unit 120, and the adjustment amount calculation unit 130. The molding condition adjustment unit 120 and the adjustment amount calculation unit 130 are examples of adjustment devices that adjust the molding conditions in the control of the injection unit 20 and the clamping device 30 by the control unit 110.

[0025] <Configuration of data processing device 200> Figure 3 shows the configuration of the data processing device 200. The data processing device 200 acquires and processes data obtained as the injection device 20 and the clamping device 30 perform operations in the process of manufacturing the molded product described above. The data processing device 200 is implemented, for example, by a computer. The data processing device 200 comprises a data acquisition unit 210, a processing unit 220, a storage unit 230, a display control unit 240, and a receiving unit 250.

[0026] The data acquisition unit 210 acquires data to be processed from the injection molding device 20 and the clamping device 30. Various sensors and detectors are attached to the injection molding device 20 and the clamping device 30. Measuring instruments may also be connected to the injection molding device. The data acquired by these sensors, detectors and measuring instruments (hereinafter referred to as "acquired data") is information representing the molding results by the injection molding device 20 and the clamping device 30, and is used for quality control of molded products. Specifically, it includes, for example, the weight of the molded product, the dimensions of the molded product, the in-mold pressure, the minimum cushion position, and characteristic waveforms of the filling pressure. This acquired data is actual values ​​obtained in the manufacturing process of molded products. The data acquisition unit 210 receives the acquired data transmitted from the sensors, detectors and measuring instruments and stores it in the storage unit 230.

[0027] The processing unit 220 processes the acquired data stored in the storage unit 230. Specifically, the processing unit 220 performs processes such as extracting representative values ​​of the acquired data in each process and generating time-series data by sequencing the acquired data in each process. In extracting representative values, the processing unit 220 performs statistical processing on the acquired data, such as calculating the average value, identifying the range of possible values, and identifying the maximum and minimum values.

[0028] The storage unit 230 stores the acquired data obtained by the data acquisition unit 210. The data format of the acquired data stored in the storage unit 230 may include, for example, binary, text, CSV (Comma Separated Values), INI, YAML ("YAML Ain't Markup Language"), JSON (JavaScript Object Notation), etc. Using these general-purpose data formats for data files makes it possible to exchange adjustment condition data files with other information processing devices or to edit adjustment condition data files acquired from external devices.

[0029] Although not shown in the figures, the storage unit 230 also holds a program for the processing unit 220 to perform data processing, a program for the display control unit 240 to display a screen on the display device 300, and a program for the reception unit 250 to receive user operations performed on the operation screen. As will be described in more detail later, the functions of the processing unit 220, the display control unit 240, and the reception unit 250 are realized when the processor in the data processing device 200 reads and executes the program held in the storage unit 230.

[0030] The display control unit 240 generates an operation screen for the user to perform various operations and displays it on the display device 300. The reception unit 250 receives the operations performed by the user on the operation screen displayed on the display device 300. Specifically, it receives, for example, data input into input fields on the operation screen and screen switching operations.

[0031] <Hardware configuration of control device 100 and data processing device 200> Figure 4 shows an example of the hardware configuration of a computer 400 that implements a control device 100 and a data processing device 200. The computer shown in Figure 4 includes a processor 401 as an arithmetic means, and a main memory 402 and an auxiliary memory 403 as storage means. Various arithmetic circuits such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), ASIC (Application Specific Integrated Circuit), and FPGA (Field-Programmable Gate Array) can be used as the processor 401. The processor 401 reads programs stored in the auxiliary memory 403 into the main memory 402 and executes them. For example, RAM (Random Access Memory) can be used as the main memory 402. For example, a magnetic disk drive or SSD (Solid State Drive) can be used as the auxiliary memory 403. The computer also includes a display mechanism 404 for displaying output to a display device (display) 300, and an input device 405 as an input means for input operations performed by the computer user. For example, a keyboard or mouse can be used as the input device 405. Note that the computer configuration shown in Figure 4 is merely an example, and the computer used in this embodiment is not limited to the configuration shown in Figure 4. For example, a configuration including non-volatile memory such as flash memory or ROM (Read Only Memory) as a storage device is also possible.

[0032] When the control device 100 is implemented by the computer shown in Figure 4, the functions of the control unit 110, the molding condition adjustment unit 120, and the adjustment amount calculation unit 130 are implemented, for example, by the processor 401 reading and executing a program. The storage unit 140 is implemented, for example, by the auxiliary storage device 403.

[0033] When the data processing device 200 is implemented by the computer shown in Figure 4, the functions of the data acquisition unit 210 and the processing unit 220 are implemented, for example, by the processor 401 reading and executing a program. The storage unit 230 is implemented, for example, by the auxiliary storage device 403. The display control unit 240 is implemented, for example, by the processor 401 that reads and executes a program and the display mechanism 404. The receiving unit 250 is implemented, for example, by the processor 401 that reads and executes a program and the input device 405.

[0034] <Calculation of adjustment amount> Next, the calculation of the adjustment amount for molding conditions by the adjustment amount calculation unit 130 of the control device 100 will be explained. As mentioned, the adjustment amount calculation unit 130 calculates the adjustment amount using a model based on the difference in molding conditions and the difference in molding quality. More specifically, the adjustment amount calculation unit 130 uses a model that represents the relationship between the difference in molding conditions and the difference in molding quality, based on the difference in molding conditions and the difference in molding quality, and applies the difference between the current molding quality and the target molding quality to this model to identify the molding conditions necessary to obtain the target molding quality. The functions of the adjustment amount calculation unit 130 will be explained in more detail below.

[0035] Figure 5 shows the functional configuration of the adjustment amount calculation unit 130. The adjustment amount calculation unit 130 comprises a molding condition difference calculation unit 131, a molding quality difference calculation unit 132, a model update unit 133, a target quality difference calculation unit 134, and a change amount calculation unit 135. The adjustment amount calculation unit 130 receives input from the standard molding conditions (labeled "Standard Value (Molding Conditions)" in the figure), the current molding conditions (labeled "Current Value (Molding Conditions)" in the figure), the standard molding quality (labeled "Standard Value (Molding Quality)" in the figure), and the current molding quality (labeled "Current Value (Molding Quality)" in the figure). The standard molding conditions and standard molding quality are read from the storage unit 140. The current molding conditions and current molding quality are obtained from the data processing device 200. The adjustment amount used to adjust the molding conditions is output from the adjustment amount calculation unit 130. The adjustment amount output from the adjustment amount calculation unit 130 is sent to the molding condition adjustment unit 120.

[0036] The molding condition difference calculation unit 131 acquires the reference molding conditions and the current molding conditions and calculates the difference between them. Hereinafter, the difference between the reference molding conditions and the current molding conditions will be referred to as the "molding condition difference". The reference molding conditions are read from the storage unit 140. The current molding conditions are the molding conditions used in the last shot and are acquired from the control unit 110. The control unit 110 sends the molding conditions used to control the injection unit 20 and the clamping unit 30 in response to a request from the adjustment amount calculation unit 130. The current molding conditions are used by the adjustment amount calculation unit 130 to calculate the adjustment amount of the molding conditions, and are then stored in the storage unit 140 as reference molding conditions to be used in subsequent calculations of the adjustment amount of the molding conditions.

[0037] The molding quality difference calculation unit 132 acquires the reference molding quality and the current molding quality and calculates the difference between them. Hereinafter, the difference between the reference molding quality and the current molding quality will be referred to as the "molding quality difference". The reference molding quality is read from the storage unit 140. The current molding quality is information representing the quality of the molded product manufactured in the last shot, and is acquired from the data processing unit 200. The data processing unit 200 extracts quality information from the acquired data obtained from the injection unit 20 and the clamping unit 30 and sends it to the control device 100. The current molding quality is used in the calculation of the adjustment amount of the molding conditions by the adjustment amount calculation unit 130, and is then stored in the storage unit 140 as the reference molding quality used in the calculation of the adjustment amount of the molding conditions later.

[0038] The model update unit 133 acquires the difference in molding conditions calculated by the molding condition difference calculation unit 131 and the difference in molding quality calculated by the molding quality difference calculation unit 132, and updates the model used to calculate the adjustment amount for molding conditions based on this difference information. By updating the model using the difference in molding conditions and the difference in molding quality, the model update unit 133 can learn the model in response to changes in the state of the injection device 20 and the clamping device 30, even during mass production of molded products.

[0039] The target quality difference calculation unit 134 obtains the target molding quality and the current molding quality and calculates the difference between them. Hereinafter, the difference between the target molding quality and the current molding quality will be referred to as the "target quality difference". The target molding quality is read from the storage unit 140. The current molding quality is obtained from the data processing device 200 as described above.

[0040] The change amount calculation unit 135 obtains the target quality difference calculated by the target quality difference calculation unit 134, applies a model to this target quality difference, and calculates the amount of change in the molding conditions. In this embodiment, the amount of change in the molding conditions calculated by the change amount calculation unit 135 is used as the adjustment amount in the molding condition adjustment by the molding condition adjustment unit 120. Therefore, this change amount is sent as an adjustment amount from the adjustment amount calculation unit 130 to the molding condition adjustment unit 120. The model used in calculating the change amount is prepared according to the molded product and mold to be manufactured, and is updated by the model update unit 133 in accordance with the state changes of the injection device 20 and the clamping device 30 during mass production of molded products.

[0041] The model used in the change amount calculation unit 135 to obtain the change amount of the molding conditions is a model that represents the relationship between the difference in molding conditions and the difference in molding quality. Therefore, by applying the model to the difference between the target molding quality and the current molding quality obtained by the target quality difference calculation unit 134, the difference between the current molding conditions and the target molding conditions (hereinafter referred to as "target molding conditions") can be obtained. The target molding conditions are molding conditions that satisfy the target molding quality for the molded product manufactured using those conditions. This difference between the current molding conditions and the target molding conditions becomes the change amount of the molding conditions. Hereinafter, the difference between the current molding conditions and the target molding conditions will be referred to as the "target condition difference". The change amount calculation unit 135 may also obtain the target condition difference that is expected to achieve the desired target quality difference by analytically solving the model to be applied. Alternatively, the change amount calculation unit 135 may obtain the target condition difference by applying the model to a numerical optimization algorithm such as the gradient method or Bayesian optimization.

[0042] The target condition difference information calculated by the change amount calculation unit 135 is sent to the molding condition adjustment unit 120 as the molding condition adjustment amount calculated by the adjustment amount calculation unit 130. The molding condition adjustment unit 120 adjusts the molding conditions by applying (adding or subtracting) the molding condition adjustment amount (i.e., target condition difference) shown in the target condition difference calculated by the adjustment amount calculation unit 130 to the current molding conditions (current molding conditions) of the control unit 110. In the adjustment amount calculation unit 130 shown in Figure 5, the change amount calculated by the change amount calculation unit 135 is an example of the molding condition adjustment amount.

[0043] The molding condition adjustment unit 120 changes the molding conditions, which affect the quality of the molded product manufactured. Therefore, the purpose of adjusting the molding conditions is to suppress changes in molding quality due to changes in the state of the injection unit 20 and the clamping unit 30, etc. In other words, if there is no change in molding quality, there is no need to adjust the molding conditions. However, in order to efficiently train the model used to calculate the amount of adjustment to the molding conditions, the molding conditions may be changed by a small amount even if the current molding quality matches the target molding quality. In this case, the change in molding conditions is made so that the quality of the molded product manufactured under the changed molding conditions falls within the range of the target molding quality.

[0044] If the values ​​obtained in a particular shot, such as the current molding conditions obtained by the molding condition difference calculation unit 131, the current molding quality obtained by the molding quality difference calculation unit 132, the difference in molding conditions calculated by the molding condition difference calculation unit 131, and the difference in molding quality calculated by the molding quality difference calculation unit 132, differ from the values ​​obtained in previous shots or the trend of their changes, or if they are other values ​​that are considered to be predetermined abnormal values, then the values ​​obtained in that shot may be noise or reflect abnormalities in sensors, detectors, or measuring instruments. In such cases, to avoid the model used to calculate the amount of adjustment for molding conditions collapsing, the model update may be postponed. Also, to prevent deterioration of molding quality by adjusting the molding conditions according to the amount of adjustment for molding conditions obtained based on such abnormal values, the adjustment of molding conditions may be postponed.

[0045] <Adjustment amount correction> In the above configuration example explained with reference to Figure 5, the adjustment amount calculation unit 130 sent the target condition difference (amount of change in molding conditions) calculated by the change amount calculation unit 135 directly to the molding condition adjustment unit 120 as the molding condition adjustment amount. In contrast, in order to prevent hunting from occurring during the molding condition adjustment by the molding condition adjustment unit 120, it is conceivable to send a corrected value (hereinafter referred to as the "corrected value") obtained by multiplying the target condition difference value by a relaxation coefficient α that takes a value of 0 ≤ α ≤ 1 to the molding condition adjustment unit 120 as the molding condition adjustment amount. The function of the adjustment amount calculation unit 130 when this corrected value is used as the molding condition adjustment amount will be explained below. In the following explanation, the value of the target condition difference will be Δx, and the corrected value obtained by multiplying by the relaxation coefficient α will be αΔx.

[0046] When modeling the relationship between molding conditions and molding quality using machine learning or the like, and calculating the molding conditions necessary to achieve the target molding quality, it is desirable to manage the accuracy of the model. In machine learning, the model is not always generated at a level that satisfies the function of calculating the molding conditions necessary to achieve the target molding quality. Therefore, if the model used is an inaccurate model, adjusting the molding conditions based on such a model may worsen the production status of the molded product. In addition, data acquired by sensors, detectors, and measuring instruments generally contains noise. If the molding conditions are adjusted to counteract this noise, the production status of the molded product may also worsen. Therefore, it is conceivable that the adjustment amount calculation unit 130 performs a correction on the calculated target condition difference based on the reliability of the target condition difference (described later). If the reliability is low, the robustness of the system as a system for adjusting molding conditions is improved by suppressing the amount of adjustment of the molding conditions based on the target condition difference.

[0047] Figure 6 shows the functional configuration of the adjustment amount calculation unit 130 when correcting the adjustment amount. The adjustment amount calculation unit 130 shown in Figure 6 comprises a molding condition difference calculation unit 131, a molding quality difference calculation unit 132, a model update unit 133, a target quality difference calculation unit 134, a change amount calculation unit 135, and a change amount correction unit 136. In the configuration shown in Figure 6, the molding condition difference calculation unit 131, the molding quality difference calculation unit 132, the model update unit 133, the target quality difference calculation unit 134, and the change amount calculation unit 135 are the same as the corresponding configurations shown in Figure 5, so their explanation is omitted.

[0048] The change amount correction unit 136 calculates a correction value αΔx by multiplying the target condition difference Δx calculated by the change amount calculation unit 135 by the relaxation coefficient α. This correction value αΔx is sent to the molding condition adjustment unit 120 as the adjustment amount of the molding conditions calculated by the adjustment amount calculation unit 130 shown in Figure 6. The molding condition adjustment unit 120 adjusts the molding conditions by applying (adding or subtracting) the adjustment amount of the molding conditions (i.e., the correction value αΔx) calculated by the adjustment amount calculation unit 130 to the current molding conditions (current molding conditions) of the control unit 110. In the adjustment amount calculation unit 130 shown in Figure 6, the correction value αΔx calculated by the change amount correction unit 136 is an example of the adjustment amount of the molding conditions.

[0049] <Example of relaxation coefficient α> The relaxation coefficient α can take values ​​of 0 ≤ α ≤ 1 and can be any coefficient that represents some degree of confidence in the target condition difference Δx. It is also possible to set the relaxation coefficient α using a probability distribution. The relaxation coefficient α used in the change amount calculation unit 135 is a coefficient corresponding to the model used to obtain the change amount of the molding conditions. A model-dependent coefficient is a coefficient predetermined for the model, or a coefficient determined by the molding conditions, molding quality, etc., based on a predetermined function corresponding to the model. In this embodiment, the relaxation coefficient α (confidence of the target condition difference Δx) is also based on the reliability of the model. The reliability of the model varies from shot to shot, and it is thought that reliability improves as more shots are repeated and more data is acquired. Furthermore, it is thought that reliability improves if the molding quality improves as a result of the adjustments to the molding conditions, and decreases if the molding quality does not improve or deteriorates as a result of the adjustments to the molding conditions. Therefore, the relaxation coefficient α may be updated periodically based on the reliability of the model. The reliability of the model changes based on the increase in data or the quality of the adjustment results. A specific example of the relaxation coefficient α is given below for further explanation.

[0050] (1) Various models can be used as the model for calculating the target condition difference Δx in the adjustment amount calculation unit 130. For example, a linear regression model and its coefficient of determination (R-squared value) or a value obtained by transforming the coefficient of determination with an increasing function can be used. Here, the value of the coefficient of determination is 1 when the relationship between the input and output is perfectly matched, and in most cases it is 0 or greater. Therefore, it is conceivable to use this coefficient of determination as is as the relaxation coefficient α. In this case, if a value of 0 or less is obtained as the value of the coefficient of determination, the value of the relaxation coefficient α may be set to 0. Alternatively, the coefficient of determination may be adjusted for degrees of freedom.

[0051] (2) As the relaxation coefficient α, it is conceivable to use a value based on the Root Mean Squared Error (RMSE) of the model used to calculate the target condition difference Δx. The possible values ​​of RMSE (residuals) are greater than or equal to 0, and there is no upper limit. Furthermore, a larger value indicates lower model accuracy. Therefore, in order to use a value based on RMSE as the relaxation coefficient α, it is necessary to transform it using an appropriate function so that the value decreases as the RMSE increases, and so that 0 ≤ α ≤ 1.

[0052] Figure 8 shows an example of a function that converts the RMSE model residuals to the relaxation coefficient α. In the graph shown in Figure 8, the vertical axis is the relaxation coefficient α and the horizontal axis is the model residuals. The function represented by the graph shown in Figure 8 is 0 < α ≤ 1, and the value of α decreases as the value of the model residuals increases. Note that existing calculation methods may be used to convert the model residuals to the relaxation coefficient α.

[0053] (3) A probability distribution may be used to calculate the relaxation coefficient α. In the following explanation, molded quality is y, and target molded quality is y. target , current molding quality y current Therefore, the target quality difference Δy is Δy = |y current -y target It is represented by |. Also, the possible range of the target quality difference Δy is Δy range Then, Δy range This is expressed as (UCL - LCL) / 2, where UCL is the Upper Control Limit and LCL is the Lower Control Limit. Also, in the following examples, for the sake of simplicity, y is treated as a scalar, but it is also possible to extend y to a vector.

[0054] FIG. 7 is a diagram for explaining reliability when molding conditions are adjusted using a target condition difference. FIG. 7(A) is a diagram showing a predicted distribution of molding quality when the reliability of the target condition difference is high, and FIG. 7(B) is a diagram showing a predicted distribution of molding quality when the reliability of the target condition difference is low. In FIGS. 7(A) and 7(B), the graph is a predicted distribution of molding quality obtained when molding conditions are adjusted using a certain target condition difference Δx.

[0055] In the illustrated graph, the reliability of the target condition difference Δx is determined by how much of the predicted distribution falls between UCL and LCL. In the example of FIG. 7(A), almost the entire predicted distribution is included between UCL and LCL. When such a predicted distribution is obtained, the relaxation coefficient α becomes a value close to 1, and thus the reliability of the target condition difference Δx in this case is high. The correction value αΔx becomes a value close to the target condition difference Δx. In contrast, in the example of FIG. 7(B), the predicted distribution extends beyond the range from UCL to LCL (the portion indicated by hatching). When such a predicted distribution is obtained, the relaxation coefficient α becomes a smaller value than in the example of FIG. 7(A), and thus the reliability of the target condition difference Δx in this case is low. The correction value αΔx becomes a small value that is away from the target condition difference Δx.

[0056] For example, after adjustment, the molding quality is y pred the predicted distribution of molding quality y with respect to the molding condition x predicted to be can be expressed as follows, where RMSE of the model is σ: p(y)=N(y|y pred ,σ 2 ) can be assumed. Here, N(y|y pred ,σ 2 ) represents a normal distribution with a mean of y pred and a variance of σ 2 . Therefore, it is conceivable to use the probability distribution that is the predicted distribution of this molding quality y as the relaxation coefficient α. This probability distribution β is calculated by the following Equation 1. Further, f(β) may be used as the relaxation coefficient α by using a monotonically increasing function f that outputs a value of 0 or more.

Formula

[0057] In the above equation 1, the integral interval Y when calculating the probability distribution β is min , Y max y target You may also use ±Δy. Additionally, the integration interval Y min , Y max y target ±Δy range You can also use y target ±ηΔy range You may also use the following. Here, η is a value greater than 0 and less than or equal to 1. When a value related to Δy is used as the integration interval in this way, the change amount correction unit 136 calculates Δy using the current molding quality y current and target molding quality y target Obtain the current molding quality y. current This is obtained from the data processing device 200. Target molding quality y target This is read from memory unit 140.

[0058] (4) The relaxation coefficient α may include an element representing the certainty of the molding quality. In this case, the change amount correction unit 136 acquires information on the reliability of the molding quality (hereinafter referred to as "quality reliability") as an element representing the certainty of the molding quality. The quality reliability information is specified according to the type of molding quality indicated by the quality information. As an example, the quality reliability may be determined by considering the accuracy of the means for acquiring the molding quality. Specifically, if the quality information is the weight of the molded product, the accuracy of the measuring instrument (weighing scale) that measures the weight of the molded product may be used as the quality reliability. The quality reliability may be set to be constant for the quality information of all shots, or it may be set as a value that changes from shot to shot.

[0059] As an example of including the element of certainty in molding quality measurement in the relaxation coefficient α, we show an example where the value of quality information obtained from the data processing device 200 is y0, and the accuracy of the measuring instrument is σ0. In this case, the true value of molding quality is N(y|y0,σ0 2 We can assume that it is distributed as follows. This probability distribution β0 is calculated by the following equation 2.

number

[0060] Figure 9 shows probability distributions β0 and β. In Figure 9, probability distribution β0 is shown as a graph representing a distribution with the quality information y0 obtained from the data processing device 200 as its peak, and probability distribution β is the predicted value y of the molding quality. pred The graph shows the distribution with the peak at [point]. The shaded areas in the graphs of probability distribution β0 and probability distribution β represent the integration interval Y, respectively. min , Y max This shows the distribution within the range. In probability distributions β0 and β, which have the relationship shown in Figure 9, the degree of improvement γ, which represents the extent to which molding quality improves by adjusting the molding conditions, can be defined, for example, as γ = β / β0. This degree of improvement γ is a value of 0 or greater, and a value of 1 or less indicates that the adjustment is counterproductive, while a value of 1 or greater indicates that the effect of the adjustment is significant as the value increases. In order to use this degree of improvement γ as the relaxation coefficient α, the degree of improvement γ is transformed by an increasing function with an upper limit of 1 as the calculated value.

[0061] Figure 10 shows an example of a function used to convert the improvement degree γ. According to the function shown in Figure 10, when the improvement degree γ is 1 or less, the calculated relaxation coefficient α is 0. When the improvement degree γ is 1 or greater, the calculated relaxation coefficient α approaches 1 as the value increases.

[0062] (5) When a probabilistic model is used in the adjustment amount calculation unit 130 to calculate the target condition difference Δx, the posterior predictive distribution may be used as the predictive distribution p(y) of the molding quality y used in (3) and (4) above, and the probability distribution β and the degree of improvement γ may be calculated.

[0063] Although embodiments of the present invention have been described above, the technical scope of the present invention is not limited to the above embodiments. For example, in the above embodiments, the injection molding machine 10 was described as having a configuration comprising a control device 100, a data processing device 200, and a display device 300, but the invention is not limited to such a configuration. For example, the functions of the molding condition adjustment unit 120 and the adjustment amount calculation unit 130 of the control device 100 may be implemented by an information processing device such as a personal computer or a server on a network, and information such as acquired data and adjustment conditions may be acquired from the data processing device 200 to adjust the molding conditions used for control by the control unit 110 of the control device 100.

[0064] Alternatively, the functions of the display control unit 240 of the data processing device 200 and the display device 300 may be implemented in an information processing device such as a personal computer, smartphone, or tablet terminal, and this information processing device may acquire data from the data processing device 200 and display an operation screen. In this case, the operation screen may be created as web content and displayed by the information processing device via a web browser.

[0065] Furthermore, the control device 100 may be provided with a function that prevents the control condition values ​​from becoming excessively large or small by allowing the user to pre-set a range of configurable molding conditions. In this case, the control device 100 may also be provided with a function that outputs an alarm to the user or stops the operation of the injection device 20 and the clamping device 30 when the adjusted molding conditions reach a threshold determined based on the configurable range.

[0066] In the above embodiment, the amount of adjustment to the molding conditions was calculated using a model that represents the relationship between the difference in molding conditions and the difference in molding quality. In addition, in the above embodiment, in addition to adjusting the molding conditions using the adjustment amount calculated using the model, a configuration can also be adopted in which the adjustment amount is corrected based on the reliability of the obtained adjustment amount, and the molding conditions are adjusted using the corrected adjustment amount. Here, calculating the amount of adjustment to the molding conditions using a model based on the difference in molding conditions and the difference in molding quality, and correcting the obtained adjustment amount based on the reliability can be carried out independently. For example, even in a control device that calculates the amount of adjustment to the molding conditions using a model that represents the relationship between molding conditions and molding quality itself, rather than the difference in molding conditions and the difference in molding quality, the obtained adjustment amount can be corrected based on the reliability described in the above embodiment, and the molding conditions can be adjusted using the corrected adjustment amount. Furthermore, various changes and substitutions of configurations that do not depart from the scope of the technical concept of the present invention are included in the present invention. [Explanation of Symbols]

[0067] 10…Injection molding machine, 20…Injection device, 30…Clamping device, 100…Control device, 110…Control unit, 120…Molding condition adjustment unit, 130…Adjustment amount calculation unit, 131…Molding condition difference calculation unit, 132…Molding quality difference calculation unit, 133…Model update unit, 134…Target quality difference calculation unit, 135…Change amount calculation unit, 136…Change amount correction unit, 140…Storage unit, 141…Control information, 142…Reference information, 143…Target molding quality, 200…Data processing unit, 210…Data acquisition unit, 220…Processing unit, 230…Storage unit, 240…Display control unit, 250…Reception unit, 300…Display device

Claims

1. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, An adjustment amount calculation unit calculates the adjustment amount for the target molding conditions based on the target molding conditions, the predicted target molding quality, and the reliability of the predicted target molding quality, using a model relating to the target molding conditions and the target molding quality, which are the molding conditions to be adjusted. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, An adjustment device characterized by comprising:

2. The model used by the adjustment amount calculation unit to calculate the adjustment amount is a linear regression model, The adjustment device according to claim 1, characterized in that the reliability of the predicted molding quality is expressed by the coefficient of determination of the linear regression model.

3. The model used by the adjustment amount calculation unit to calculate the adjustment amount is a linear regression model, The adjustment device according to claim 1, characterized in that the reliability of the predicted molding quality is expressed by a coefficient transformed using the mean square root error of the linear regression model so that it is a value between 0 and 1.

4. The adjustment device according to claim 1, characterized in that the reliability of the predicted target molding quality is identified based on the predicted distribution of molding quality that is expected to be obtained after adjusting the molding conditions based on the model.

5. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, An adjustment amount calculation unit calculates the adjustment amount for the target molding condition based on the target molding quality and the target molding quality, using a model relating to the target molding condition (the molding condition to be adjusted) and the target molding quality (the molding quality to be adjusted), and coefficients corresponding to the model. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, Equipped with, The adjustment device is characterized in that the coefficient used by the molding condition adjustment unit is determined based on the ratio of the predicted distribution of molding quality expected to be obtained after adjustment of the molding conditions to the predicted distribution of molding quality obtained considering the reliability of the molding quality, which is specified according to the type of molding quality.

6. The adjustment device according to claim 5, characterized in that the reliability of the molding quality is determined based on the accuracy of the means for obtaining the molding quality.

7. The model used by the adjustment amount calculation unit to calculate the adjustment amount is a probabilistic model. The adjustment device according to claim 4 or claim 5, characterized in that the molding condition adjustment unit uses a post-prediction distribution as the prediction distribution.

8. The adjustment device according to any one of claims 1 to 5, characterized in that the adjustment amount calculation unit applies a model representing the relationship between the difference between a reference molding condition, which is a standard molding condition, and the target molding condition, and the difference between a reference molding quality, which is a standard molding quality, and the target molding quality, which is the molding quality to be adjusted, to the difference between the target molding quality and the target molding quality, which is the target molding quality, and calculates the difference between the target molding condition and the target molding condition.

9. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, A model representing the relationship between the difference between the standard molding conditions and the target molding conditions, and the difference between the standard molding quality and the target molding quality, which is the quality information of the molded product manufactured under the target molding conditions, is applied to the difference between the target molding quality and the target molding quality, and an adjustment amount calculation unit calculates the difference between the target molding conditions and the target molding conditions, as an adjustment amount for the molding conditions. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, Equipped with, The adjustment amount calculation unit calculates the adjustment amount by applying a coefficient, whose value is determined according to the model and is between 0 and 1, to the difference between the target molding conditions and the target molding conditions. The molding condition adjustment unit is characterized by adjusting the target molding conditions based on the adjustment amount to which the coefficient is applied.

10. The adjustment device according to claim 9, characterized in that the coefficient is determined based on a model used to calculate the difference in the target molding conditions.

11. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, A model representing the relationship between the difference between the standard molding conditions and the target molding conditions, and the difference between the standard molding quality and the target molding quality, which is the quality information of the molded product manufactured under the target molding conditions, is applied to the difference between the target molding quality and the target molding quality, and an adjustment amount calculation unit calculates the difference between the target molding conditions and the target molding conditions, as an adjustment amount for the molding conditions. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, Equipped with, The molding condition adjustment unit is characterized in that, when the target molding quality and the target molding quality match, the molding condition is adjusted so that the quality of the molded product obtained in the manufacturing process of the molded product by the injection molding machine after adjusting the molding condition falls within a range recognized for the target molding quality.

12. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, A model representing the relationship between the difference between the standard molding conditions and the target molding conditions, and the difference between the standard molding quality and the target molding quality, which is the quality information of the molded product manufactured under the target molding conditions, is applied to the difference between the target molding quality and the target molding quality, and an adjustment amount calculation unit calculates the difference between the target molding conditions and the target molding conditions, as an adjustment amount for the molding conditions. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, A molding condition difference calculation unit obtains the current molding conditions from a control device that controls the injection molding machine, uses the current molding conditions as the target molding conditions, and calculates the difference between the current molding conditions and the reference molding conditions. A molding quality difference calculation unit obtains the current molding quality, which is quality information of the molded product manufactured based on the current molding conditions, obtained in the manufacturing process of the molded product by the injection molding machine, uses the current molding quality as the target molding quality, and calculates the difference between the current molding quality and the reference molding quality. A model update unit updates the model using the difference between the current molding conditions and the standard molding conditions, and the difference between the current molding quality and the standard molding quality. Equipped with, An adjustment device characterized in that, if one or more of the values ​​of the current molding conditions, the current molding quality, the difference calculated by the molding condition difference calculation unit, and the difference calculated by the molding quality difference calculation unit differ from the values ​​obtained in previous shots or the trend of the changes in such values, the model update unit does not update the model.

13. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, A model representing the relationship between the difference between the standard molding conditions and the target molding conditions, and the difference between the standard molding quality and the target molding quality, which is the quality information of the molded product manufactured under the target molding conditions, is applied to the difference between the target molding quality and the target molding quality, and an adjustment amount calculation unit calculates the difference between the target molding conditions and the target molding conditions, as an adjustment amount for the molding conditions. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, A molding condition difference calculation unit obtains the current molding conditions from a control device that controls the injection molding machine, uses the current molding conditions as the target molding conditions, and calculates the difference between the current molding conditions and the reference molding conditions. A molding quality difference calculation unit obtains the current molding quality, which is quality information of the molded product manufactured based on the current molding conditions, obtained in the manufacturing process of the molded product by the injection molding machine, uses the current molding quality as the target molding quality, and calculates the difference between the current molding quality and the reference molding quality. Equipped with, An adjustment device characterized in that, if one or more of the values ​​of the current molding conditions, the current molding quality, the difference calculated by the molding condition difference calculation unit, and the difference calculated by the molding quality difference calculation unit differ from the values ​​obtained in previous shots or the trend of the changes in such values, the molding condition adjustment unit does not adjust the molding conditions.

14. An adjustment device for adjusting the molding conditions of an injection molding machine that performs injection molding based on molding conditions, A model representing the relationship between the difference between the reference molding conditions and the target molding conditions, and the difference between the reference molding quality and the target molding quality, which is the quality information of a molded product manufactured under the target molding conditions, is applied to the difference between the target molding quality and the target molding quality, which is the target molding quality. Based on the ratio of the predicted distribution of molding quality expected to be obtained after adjusting the molding conditions and the predicted distribution of molding quality obtained considering the reliability of the molding quality specified according to the type of molding quality, a coefficient corresponding to the model is identified, and an adjustment amount calculation unit calculates the difference between the target molding conditions and the target molding conditions, which is the target molding conditions, based on the model and the coefficient. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, An adjustment device characterized by comprising:

15. A molding condition difference calculation unit obtains the current molding conditions from a control device that controls the injection molding machine, uses the current molding conditions as the target molding conditions, and calculates the difference between the current molding conditions and the reference molding conditions. A molding quality difference calculation unit obtains the current molding quality, which is quality information of the molded product manufactured based on the current molding conditions, obtained in the manufacturing process of the molded product by the injection molding machine, uses the current molding quality as the target molding quality, and calculates the difference between the current molding quality and the reference molding quality. A model update unit updates the model using the difference between the current molding conditions and the standard molding conditions, and the difference between the current molding quality and the standard molding quality. The adjustment device according to any one of claims 9, 11, 13, and 14, further comprising the above.

16. The adjustment device according to claim 15, characterized in that the standard molding conditions are molding conditions used in a shot prior to the shot in which the molded product was manufactured based on the current molding conditions, and the standard molding quality is quality information of the molded product manufactured based on the standard molding conditions.

17. Injection device and Clamping device and The system includes a control device that has an automatic adjustment function for molding conditions and controls the injection device and the clamping device, The control device is An adjustment amount calculation unit calculates the adjustment amount for the target molding conditions based on the target molding quality and the target molding quality, using a model relating to the target molding conditions and the target molding quality, respectively. A molding condition adjustment unit adjusts the target molding conditions based on a coefficient that is either 0 or more and 1 or less, corresponding to the model used in the calculation of the adjustment amount by the adjustment amount calculation unit, and the adjustment amount. An injection molding machine characterized by being equipped with the following.

18. Injection device and Clamping device and The system includes a control device that has an automatic adjustment function for molding conditions and controls the injection device and the clamping device, The control device is A model representing the relationship between the difference between the reference molding conditions and the target molding conditions, and the difference between the reference molding quality and the target molding quality, which is the quality information of a molded product manufactured under the target molding conditions, is applied to the difference between the target molding quality and the target molding quality, which is the target molding quality. Based on the ratio of the predicted distribution of molding quality expected to be obtained after adjusting the molding conditions and the predicted distribution of molding quality obtained considering the reliability of the molding quality specified according to the type of molding quality, a coefficient corresponding to the model is identified, and an adjustment amount calculation unit calculates the difference between the target molding conditions and the target molding conditions, which is the target molding conditions, based on the model and the coefficient. A molding condition adjustment unit adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation unit, An injection molding machine characterized by being equipped with the following.

19. Computers, An adjustment amount calculation means calculates the amount of adjustment to the target molding conditions based on the target molding quality and the target molding quality, using a model relating to the target molding conditions and the target molding quality, respectively. A molding condition adjustment means that adjusts the target molding conditions based on a coefficient that is either 0 or more and 1 or less, corresponding to the model used in calculating the adjustment amount by the adjustment amount calculation means, A program characterized by its ability to function.

20. Computers, A model representing the relationship between the difference between a standard molding condition and the target molding condition, and the difference between a standard molding quality and the target molding quality, which is the quality information of a molded product manufactured under the target molding condition, is applied to the difference between the target molding quality and the target molding quality, which is the target molding quality. Based on the ratio of the predicted distribution of the molding quality expected to be obtained after adjusting the molding conditions and the predicted distribution of the molding quality obtained considering the reliability of the molding quality specified according to the type of molding quality, a coefficient corresponding to the model is identified, and an adjustment amount calculation means calculates the difference between the target molding condition and the target molding condition, which is the target molding condition, based on the model and the coefficient. A molding condition adjustment means that adjusts the target molding conditions based on the adjustment amount calculated by the adjustment amount calculation means, A program characterized by its ability to function.

Images