JP7797233B2 - Molding condition adjustment method, computer program, molding condition adjustment device, and injection molding machine - Google Patents

Description

The present invention relates to a molding condition adjustment method, a computer program, a molding condition adjustment device, and an injection molding machine.

When a new mold is attached to an injection molding machine and injection molding is performed for the first time, the molding conditions must first be determined by adjusting the settings for various molding parameters. These molding parameters include, for example, injection cylinder temperature, injection pressure, injection speed, holding pressure switch position, and screw back position. Adjustments to these molding parameters are made based on the operator's experience, and repeated trial and error is required to obtain the appropriate molding conditions.

Patent Document 1 discloses a molding data collection method for obtaining the most suitable molding conditions. The molding data collection method described in Patent Document 1 first determines rough molding conditions. Next, for each molding condition item selected from the rough molding conditions, the molding condition value is sequentially increased or decreased by a predetermined increment within a predetermined allowable limit range to create multiple molding conditions. Next, molding data is collected for each molding condition while continuous molding operations are automatically performed a set number of times or for a set time using the multiple molding conditions created.

Japanese Patent Application Publication No. 09-216265

However, no method has been disclosed for automatically adjusting molding conditions, taking into account mold characteristics, resin properties, or the state of the molded product.

The purpose of this disclosure is to provide a molding condition adjustment method, computer program, molding condition adjustment device, and injection molding machine that can determine the molding condition items to be adjusted, the adjustment range, and the adjustment increment width, and automatically execute condition setting.

The molding condition adjustment method according to this aspect adjusts the molding conditions of an injection molding machine by acquiring mold information indicating the characteristics of the mold, resin information indicating the properties of the resin, or defect information indicating defects in the molded product; determining the molding condition items to be adjusted, the adjustment range, and the adjustment increments based on the acquired mold information, resin information, or defect information; and executing molding control while changing the molding conditions based on the determined molding condition items to be adjusted, the adjustment range, and the adjustment increments.

The computer program (program product) according to this aspect is a computer program for causing a computer to execute a process for adjusting molding conditions for an injection molding machine. It acquires mold information indicating the characteristics of the mold, resin information indicating the properties of the resin, or defect information indicating defects in the molded product, determines the molding condition items to be adjusted, the adjustment range, and adjustment increments based on the acquired mold information, resin information, or defect information, and causes the computer to execute a process for executing molding control while changing the molding conditions based on the determined molding condition items to be adjusted, the adjustment range, and adjustment increments.

The molding condition adjustment device according to this aspect is a molding condition adjustment device that adjusts the molding conditions of an injection molding machine, and includes an acquisition unit that acquires mold information indicating the characteristics of the mold, resin information indicating the properties of the resin, or defect information indicating defects in the molded product; a determination unit that determines the molding condition items to be adjusted, the adjustment range, and the adjustment increment size based on the acquired mold information, resin information, or defect information; and a control unit that executes molding control while changing the molding conditions based on the determined molding condition items to be adjusted, the adjustment range, and the adjustment increment size.

The injection molding machine according to this embodiment is equipped with the molding condition adjustment device described above.

This invention allows you to determine the molding conditions to be adjusted, the adjustment range, and the adjustment increment, and automatically execute the adjustment process.

1 is a schematic diagram illustrating an example of the configuration of an injection molding machine according to a first embodiment. FIG. 1 is a block diagram showing an example of the configuration of a molding condition adjustment device according to a first embodiment. FIG. 2 is a conceptual diagram illustrating a condition assignment learning model according to the first embodiment. 4 is a flowchart showing a processing procedure of a processor according to the first embodiment. FIG. 4 is a schematic diagram showing an example of a molding condition adjustment screen according to the first embodiment. FIG. 10 is a schematic diagram illustrating a configuration example of an injection molding machine according to a second embodiment. FIG. 10 is a conceptual diagram illustrating a condition assignment learning model according to a second embodiment.

Specific examples of molding condition adjustment methods, computer programs, molding condition adjustment devices, and injection molding machines according to embodiments of the present invention are described below with reference to the drawings. At least some of the embodiments described below may be combined in any manner. The present invention is not limited to these examples, but is defined by the claims, and is intended to include all modifications within the meaning and scope of the claims.

Figure 1 is a schematic diagram showing an example configuration of an injection molding machine 1 according to embodiment 1. The injection molding machine 1 according to embodiment 1 includes a mold clamping device 2 that clamps a mold 21, an injection device 3 that plasticizes and injects molding material, and a control device 4. The control device 4 functions as a molding condition adjustment device according to embodiment 1.

The mold clamping unit 2 comprises a fixed platen 22 fixed on the bed 20, a mold clamping housing 23 slidably mounted on the bed 20, and a movable platen 24 that also slides on the bed 20. The fixed platen 22 and the mold clamping housing 23 are connected by multiple, for example, four tie bars 25, 25, .... The movable platen 24 is configured to be slidable between the fixed platen 22 and the mold clamping housing 23. A mold clamping mechanism 26 is provided between the mold clamping housing 23 and the movable platen 24.

The mold clamping mechanism 26 is composed of, for example, a toggle mechanism. The mold clamping mechanism 26 may also be composed of a direct pressure mold clamping mechanism, i.e., a mold clamping cylinder. The fixed platen 22 and the movable platen 24 are provided with a fixed mold 21a and a movable mold 21b, respectively, and when the mold clamping mechanism 26 is driven, the mold 21 is opened and closed.

The injection device 3 is mounted on a base 30. The injection device 3 includes a heating cylinder 31 having a nozzle 31a at its tip, and a screw 32 arranged within the heating cylinder 31 so that it can rotate circumferentially and axially. The screw 32 is driven in the rotational and axial directions by a drive mechanism 33. The drive mechanism 33 is composed of a rotary motor that drives the screw 32 in the rotational direction, and a motor that drives the screw 32 in the axial direction. Note that the drive mechanism 33 shown in Figure 1 is covered by a cover, so its internal structure is not shown.

A hopper 34 into which molding material is poured is provided near the rear end of the heating cylinder 31. The injection molding machine 1 also includes a nozzle touch device 35 that moves the injection unit 3 in the front-to-rear direction (left-to-right direction in Figure 1). When the nozzle touch device 35 is driven, the injection unit 3 moves forward so that the nozzle 31a of the heating cylinder 31 touches the contact portion of the fixed platen 22.

Figure 2 is a block diagram showing an example configuration of the control device 4 according to the first embodiment. The control device 4 is a computer that controls the operation of the mold clamping unit 2 and the injection unit 3, and includes a processor (control unit) 41, a memory unit 42, a control signal output unit 43, and an operation panel 40 as its hardware configuration. The control device 4 adjusts the molding conditions of the injection molding machine 1. The control device 4 may also be a server device connected to a network. The control device 4 may also be configured to perform distributed processing using multiple computers, or may be realized by multiple virtual machines installed on a single server, or may be realized using a cloud server.

The processor 41 includes arithmetic circuits such as a CPU (Central Processing Unit), a multi-core CPU, a GPU (Graphics Processing Unit), a GPGPU (General-purpose computing on graphics processing units), a TPU (Tensor Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), and an NPU (Neural Processing Unit), internal storage devices such as a ROM (Read Only Memory) and a RAM (Random Access Memory), an I/O terminal, and a timer. The processor 41 implements the molding condition adjustment method according to this embodiment 1 by executing a computer program (program product) 42a stored in the storage unit 42 (described below). Each functional unit of the control device 4 may be implemented by software, or some or all of them may be implemented by hardware.

The storage unit 42 is a non-volatile memory such as a hard disk, an EEPROM (Electrically Erasable Programmable ROM), or a flash memory. The storage unit 42 stores a computer program 5 for causing a computer to execute the molding condition adjustment process. The storage unit 42 also stores a condition setting learning model 6 for setting molding conditions. The configuration of the condition setting learning model 6 will be described in detail later. The storage unit 42 also includes an adjustment history DB7 that stores the history of molding conditions set during the condition setting process using the molding condition adjustment method. The adjustment history DB7 stores, for example, molding condition items, setting values for those items, and date and time information when those setting values were set, in association with each other.

The computer program 5 according to the first embodiment may be recorded on a recording medium 50 in a computer-readable manner. The storage unit 42 stores the computer program 5 read from the recording medium 50 by a reading device. The recording medium 50 is a semiconductor memory such as a flash memory. The recording medium 50 may also be an optical disc such as a CD (Compact Disc)-ROM, a DVD (Digital Versatile Disc)-ROM, or a BD (Blu-ray (registered trademark) Disc). Furthermore, the recording medium 50 may also be a magnetic disc such as a flexible disk or hard disk, a magneto-optical disc, etc.

Furthermore, the computer program 5 according to this embodiment 1 may be downloaded from an external server connected to the communication network and stored in the memory unit 42.

The control signal output unit 43 outputs a control signal to the injection molding machine 1 to control the operation of the injection molding machine 1 in accordance with the control of the processor 41 based on the molding conditions.

The operation panel 40 is an interface for setting molding conditions for the injection molding machine 1 and controlling the operation of the injection molding machine 1. The operation panel 40 includes a display panel 40a, an operation unit 40b, and a communication unit 40c.

The display panel 40a is a display device such as a liquid crystal display panel or an organic EL display panel, and, under the control of the processor 41, displays a reception screen for receiving settings for the molding conditions of the injection molding machine 1, the status of the injection molding machine 1, and the implementation status of the molding condition adjustment method according to this embodiment 1.

The operation unit 40b is an input device for inputting and adjusting molding conditions for the injection molding machine 1, and includes operation buttons, a touch panel, etc. The operation unit 40b provides data indicating the received molding conditions to the processor 41.

The communication unit 40c is a communication circuit that communicates with external devices via wired or wireless connections to acquire information. For example, the communication unit 40c can communicate with an external storage medium such as a USB to acquire data. The communication unit 40c can acquire data from an external computer via wired communication via a LAN cable or wireless communication. The communication unit 40c can acquire molding conditions that are initially set in the injection molding machine 1, mold image data (mold information), and the like from an external storage medium or external computer.

The injection molding machine 1 is set with set values that define molding conditions such as the injection start position , the resin temperature in the mold, the nozzle temperature, the cylinder temperature (heater temperature), the hopper temperature, the mold clamping force, the injection speed, the injection acceleration, the injection peak pressure (injection pressure), and the injection stroke.

In addition, injection molding machine 1 is set with set values that define molding conditions such as cylinder tip resin pressure, backflow prevention ring seating state, dwell pressure, dwell pressure switching speed, dwell pressure switching position, dwell pressure completion position, cushion position, metering back pressure, and metering torque.

Furthermore, the injection molding machine 1 is set with molding conditions such as the metering completion position, screw retraction speed, cycle time, mold closing time, injection time, pressure holding time, metering time, and mold opening time. Furthermore, the injection molding machine 1 is set with molding conditions such as cooling time, screw rotation speed, mold opening/closing speed, ejection speed, and number of ejections. Once these setting values are set, the injection molding machine 1 operates in accordance with those setting values.

Figure 3 is a conceptual diagram showing the configuration of the condition-finding learning model 6 according to embodiment 1. The condition-finding learning model 6 is a trained machine learning model that, when mold image data, resin information, and defect information are input, outputs information related to the degree of adjustment necessity, adjustment range, and adjustment step size for multiple molding condition items.

Mold image data is image data obtained by capturing images of runners, gates, and other flow paths through which molten raw resin flows, as well as cavities or cores corresponding to the shape of the molded product.

Resin information includes data indicating the melting point or viscosity of the raw resin fed into the injection molding machine 1. Defect information includes data indicating the presence or absence of burrs, sink marks, flow marks, cracks, weld lines, voids, silver streaks, or shorts in the molded product.

Conditioning is performed by repeatedly taking test shots while changing the setting values of specific items in specific increments. The adjustment range includes the initial value, which is set when conditioning is performed, and the end value. The adjustment increment is the amount by which the setting value is changed each time a test shot is performed.

The conditioning learning model 6 includes a first input layer 61a and a first intermediate layer 62a to which mold image data is input, a second input layer 61b and a second intermediate layer 62b to which resin information and defect information are input, and an output layer 63.

The first input layer 61a and first intermediate layer 62a have a configuration similar to that of the input layer and intermediate layer of a convolutional neural network (CNN). The first input layer 61a has multiple nodes to which the pixel values of each pixel constituting the mold image are input. The first intermediate layer 62a has a configuration in which convolutional layers that convolve the pixel values of each pixel of the mold image data input to the first input layer 61a and pooling layers that map the pixel values convolved in the convolutional layers are alternately connected. The first intermediate layer 62a extracts feature values of the mold image while compressing the pixel information of the mold image, and outputs the extracted feature values of the mold image data to the second intermediate layer 62b.

The second input layer 61b has a node to which resin information is input and a node to which defect information is input. The second intermediate layer 62b has multiple intermediate layers each having multiple nodes, and the nodes in the input-side intermediate layer are connected to the nodes in the second input layer 61b and the nodes on the output side of the first intermediate layer 62a. The second intermediate layer 62b receives input of mold image data features, resin information, and defect information, and extracts information related to adjusting molding conditions.

The output layer 63 has multiple nodes corresponding to multiple items related to molding conditions. The node corresponding to the first item includes a node that outputs the degree of adjustment necessity of the setting value related to that item, a node that outputs the adjustment range of that setting value, and a node that outputs the adjustment increment width. Similar nodes are provided for the other items. The degree of adjustment necessity is a probability indicating whether the setting value related to that item needs to be adjusted, and is a number between 0 and 1, for example. Note that the configuration of the output layer 63 is not particularly limited, as long as it can obtain information indicating the molding condition item to be adjusted, the adjustment range, and the adjustment increment width.

The conditioning learning model 6 can be generated by machine learning using training data including mold image data, resin information, and defect information, as well as molding condition items to be adjusted under such conditions, the adjustment range, and the adjustment step size. For example, the processor 41 performs machine learning on the conditioning learning model 6 by optimizing the weight coefficients of the conditioning learning model 6 using the error backpropagation method, error gradient descent method, or the like, using the training data.

Figure 4 is a flowchart showing the processing steps of the processor 41 according to the first embodiment. The processor 41 accepts initial molding condition settings via the operation panel 40 and sets the setting values for various items related to the accepted molding conditions (step S111). The processor 41 then controls the operation of the injection molding machine 1 according to the set molding conditions (step S112). The injection molding performed in step S112 is a test shot for obtaining information necessary to implement the molding condition adjustment method according to the first embodiment. Specifically, the operator observes the condition of the molded product obtained by executing the test shot and checks for any defects in the molded product.

The processor 41 then determines whether or not the automatic condition setting mode using the molding condition adjustment method according to this embodiment 1 has been selected (step S113). The operator can use the operation panel 40 to select whether or not to set molding conditions using the automatic condition setting mode according to this embodiment 1. The processor 41 stores the setting by the operator as to whether or not to use the automatic condition setting mode.

If it is determined that the automatic condition setting mode has not been selected (step S113: NO), the processor 41 executes a manual adjustment process for the molding conditions (step S114) and ends the process. Specifically, the processor 41 accepts the adjustments of the molding conditions made by the operator via the operation panel 40 and sets the adjusted molding conditions.

If it is determined that the automatic condition setting mode has been selected (step S113: YES), the processor 41 acquires mold information via the operation panel 40 (step S115). The mold information is, for example, mold image data obtained by capturing an image of the mold 21.

Next, the processor 41 acquires resin information including data indicating the melting point or viscosity of the raw resin via the operation panel 40 (step S116). If the resin information is included in the molding conditions, the resin information may be acquired from the molding conditions initially set in step S111. Alternatively, a table may be provided that associates resin types with resin information, and the processor 41 may be configured to extract resin information from the table using information indicating the resin type entered via the operation panel 40 as a key.

Next, the processor 41 acquires defect information for the molded product via the operation panel 40 (step S117). Specifically, the operator checks the condition of the molded product obtained by the test shot in step S112 and inputs defect information indicating the presence or absence of burrs, sink marks, flow marks, cracks, weld lines, voids, silver streaks, or shorts in the molded product via the operation panel 40.

The processor 41 that executes steps S115 to S117 functions as an acquisition unit that acquires mold information, resin information, and defect phenomena.

Next, processor 41 inputs the mold information, resin information, and defect information acquired in steps S115 to S117 into condition assignment learning model 6. Processor 41 determines the molding condition items to be adjusted based on the degree of adjustment necessity for each of the multiple items output from condition assignment learning model 6 (step S118). For example, processor 41 determines the item with the greatest degree of adjustment necessity as the item to be adjusted. Note that processor 41 may also determine an item with a degree of adjustment necessity equal to or greater than a predetermined threshold as the item to be adjusted. Processor 41 may determine multiple types of items as the item to be adjusted, but for simplicity of explanation in this embodiment 1, only one item is determined as the item to be adjusted.

The processor 41 also references data indicating the adjustment range output from the node corresponding to the item determined in step S118 and determines the adjustment range of the setting value (step S119). The processor 41 also references data indicating the step size output from the node corresponding to the item determined in step S118 and determines the adjustment step size of the setting value (step S120).

The processor 41 that executes steps S117 to S119 functions as a determination unit that determines the molding condition items to be adjusted, the adjustment range, and the adjustment increment size based on the acquired mold information, resin information, or defect information.

The processor 41 adjusts the setting values of the setting items determined in step S118 using the adjustment range and increment determined in steps S119 and S120, and stores the adjusted values as adjustment history in the memory unit 42 (step S121). Specifically, the processor 41 executes a process to sequentially change the setting values of the setting items requiring adjustment, from the initial value to the end value of the adjustment range, using the increment determined in step S120. The processor 41 controls the operation of the injection molding machine 1 in accordance with the adjusted molding conditions (step S122).

After completing the processing of step S122, processor 41 determines whether the setting value has reached the end value (step S123). If it determines that the setting value of the item to be adjusted has not reached the end value (step S123: NO), processor 41 returns the processing to step S121 and repeats the adjustment processing.

If it is determined that the setting value for the item to be adjusted has reached the end value (step S123: YES), defect information for the molded product molded using the optimal value is obtained (step S124), and it is determined whether any defects remain (step S125). It is sufficient for this defect information to include at least data indicating whether the molded product is good or bad. The operator checks the condition of the multiple molded products molded in steps S121 to S123, and enters defect information for the best molded product into the operation panel 40. Based on this defect information via the operation panel 40, the processor 41 determines whether any defects remain after the automatic adjustment.

The processor 41 may also be configured to acquire molded product image data obtained by capturing an image of the molded product, and determine the quality of the molded product based on the acquired molded product image data. The quality of the molded product may be determined based on rules, or may be determined using a learning model previously obtained through machine learning.

If it is determined in step S125 that there is a defect (step S125: YES), the processor 41 returns the process to step S117 and continues the molded product adjustment process. If it is determined that there is no defect (step S125: NO), the processor 41 ends the molding condition adjustment process.

Figure 5 is a schematic diagram showing an example of the molding condition adjustment screen 8 according to embodiment 1. The processor 41 displays the molding condition adjustment screen 8 when performing the molding condition adjustment method. The molding condition adjustment screen 8 includes a setting value adjustment section 81 and a condition setting mode display section 82.

The set value adjustment unit 81 includes a set value display unit 81a, a set value input unit 81b, and an adjustment history display unit 81c. The processor 41 displays one item related to molding conditions and the current set value for that item on the set value display unit 81a. The set value input unit 81b includes soft keys such as a numeric keypad, a "CRL" key, and an "Enter" key. When the operator operates the set value input unit 81b and inputs a set value, the processor 41 stores the input set value as the set value for the item displayed on the set value display unit 81a.

Meanwhile, the processor 41 reads the history of setting values, including molding condition items and setting values, from the adjustment history DB 7, and displays the read items and setting values on the adjustment history display unit 81c. When the operator operates the pull-down button, the processor 41 displays a list of previously set items and setting values. The processor 41 may also be configured to display a list of multiple items and setting values related to previously set molding conditions.

The condition extraction mode display unit 82 includes an automatic condition extraction mode selection unit 82a, an initial value display unit 82b, an end value display unit 82c, and an increment display unit 82d. The automatic condition extraction mode selection unit 82a is, for example, a checkbox, and the operator can select whether or not to use the automatic condition extraction mode by operating the automatic condition extraction mode selection unit 82a. The processor 41 accepts operation of the automatic condition extraction mode selection unit 82a and stores whether or not to use the automatic condition extraction mode.

If the automatic condition setting mode is selected, the processor 41 displays the items determined in step S118 in the setting value display section 81a. The processor 41 also displays the initial and final values of the adjustment range determined in step S119 in the initial value display section 82b and final value display section 82c. Similarly, the processor 41 displays the step size determined in step S120 in the step size display section 82d.

When the processor 41 controls injection molding while changing the setting values through the processing of steps S121 to S123, it displays the changed setting values on the setting value display unit 81a each time.

The molding condition adjustment screen 8 allows the operator to instruct the implementation of the molding condition adjustment method according to this embodiment 1 by operating the condition setting mode display section 82. By checking the display in the condition setting mode display section 82, the operator can see the adjustment range and increments at which the setting values are being changed. By checking the display in the setting value display section 81a, the operator can see the molding condition items and setting values that are currently being automatically adjusted.

The injection molding machine 1 according to embodiment 1, configured as described above, can determine the molding condition items to be adjusted, the adjustment range, and the adjustment increment size based on mold image data, resin information, and defect information, and automatically execute condition setting.

Specifically, the processor 41 can automatically set conditions such as injection speed, injection pressure, heater temperature, dwell pressure, dwell switch position, injection start position, resin temperature in the mold, nozzle temperature, hopper temperature, mold clamping force, injection acceleration, injection stroke, cylinder tip resin pressure, backflow prevention ring seating state, dwell switch speed, dwell completion position, cushion position, metering back pressure, metering torque, metering completion position, screw retraction speed, cycle time, mold closing time, injection time, dwell time, metering time, mold opening time, cooling time, screw rotation speed, mold opening/closing speed, ejection speed, or number of ejections, based on mold image data, resin information indicating the melting point and viscosity of the raw resin, and defect information indicating the presence or absence of flash, sink marks, flow marks, cracks, weld lines, voids, silver streaks, or shorts in the molded product.

In addition, the processor 41 and adjustment history display unit 81c can display the history of items and setting values set in the automatic condition setting mode.

In this first embodiment, a conditioning learning model 6 using a neural network has been described, but conditioning information may also be determined using other well-known machine learning models such as an SVM (Support Vector Machine) or a Bayesian network.

In addition, in this embodiment 1, an example was described in which conditions were determined using all of the mold information, resin information, and defect information, but conditions may also be determined using any one of the mold information, resin information, and defect information.

Furthermore, in this embodiment 1, a conditioning learning model 6 learned by supervised learning was described, but conditioning information may also be determined using a model learned by reinforcement learning.

The configuration of the control device 4, which uses reinforcement learning, is as follows. The processor 41 of the control device 4 has, as functional units, an observation unit, a reward calculation unit, and an agent. Note that each functional unit of the control device 4 may be implemented in software, or some or all of them may be implemented in hardware.

The observation unit acquires mold information, resin information, and defect information, and outputs the acquired information as observation data to the agent and reward calculation unit.

The reward calculation unit calculates reward data based on the acquired observation data, particularly defect information, and outputs the calculated reward data to the agent. For example, if the molded product has a high degree of defect, the reward value will be small or negative. Furthermore, the reward value will be larger the fewer times it takes to obtain a defect-free molded product.

The agent may be, for example, a reinforcement learning model with a deep neural network such as DQN, A3C, or D4PG, or a model-based reinforcement learning model such as PlaNet or SLAC.

In the case of a reinforcement learning model with a deep neural network, the agent is equipped with a DQN (Deep Q-Network) and determines an action a corresponding to the state s of the injection molding machine 1 indicated by the observation data. The state s includes mold information, resin information, and defect information.

DQN is a neural network model that outputs the value of each of multiple actions a when a state s indicated by observed data is input. The multiple actions a are the items to be adjusted, the adjustment range, and the adjustment increment. High-value actions a represent appropriate conditioning conditions. The agent selects a high-value action a, and the selected action a causes the injection molding machine 1 to transition to another state. After the state transition, the agent receives the reward calculated by the reward calculation unit, and is trained to maximize profits, i.e., cumulative rewards.

More specifically, a DQN has an input layer, a hidden layer, and an output layer. The input layer has multiple nodes to which the state s, i.e., the observed data, is input. The output layer has multiple nodes that correspond to multiple actions a, respectively, and output the value Q(s, a) of that action a in the input state s.

Based on the state s, action a, and the reward r obtained from the action, the value Q expressed by the following formula (1) is used as training data, and various weighting coefficients that characterize the DQN can be adjusted to enable reinforcement learning of the agent's DQN.
Q(s, a)←Q(s,a)+α(r+γmaxQ(snext,anext)−Q(s,a))...(1)
however,
s: State a: Action α: Learning coefficient r: Reward γ: Discount rate maxQ(snext, next): Maximum Q value for the next possible action

(Embodiment 2)
The injection molding machine 1 according to the second embodiment differs from the first embodiment in that it automatically acquires information about defects in molded products and automatically adjusts molding conditions. Since the other configurations of the injection molding machine 1 are the same as those of the injection molding machine 1 according to the first embodiment, the same parts are designated by the same reference numerals and detailed descriptions are omitted.

Figure 6 is a schematic diagram illustrating an example configuration of an injection molding machine 1 according to embodiment 2. The injection molding machine 1 according to embodiment 2 is equipped with an imaging device 9 that captures an image of a molded product. The imaging device 9 captures an image of the molded product and outputs the resulting molded product image data to the control device 4.

Figure 7 is a conceptual diagram showing the condition-setting learning model 6 according to embodiment 2. The condition-setting learning model 6 according to embodiment 2 includes a first input layer 61a and a first intermediate layer 62a, a second input layer 61b and a second intermediate layer 62b, an output layer 63, and a third input layer 61c and a third intermediate layer 62c to which molded product image data is input.

The third input layer 61c and third intermediate layer 62c have a configuration similar to that of the input layer and intermediate layer of a convolutional neural network, for example. The third input layer 61c has multiple nodes to which the pixel values of each pixel constituting the molded product image are input. The third intermediate layer 62c has a configuration in which convolutional layers that convolve the pixel values of each pixel of the molded product image data input to the third input layer 61c are alternately connected, and pooling layers that map the pixel values convolved in the convolutional layers. The third intermediate layer 62c extracts feature values of the molded product image while compressing the pixel information of the molded product image, and outputs the extracted feature values of the molded product image data to the second intermediate layer 62b.

The second intermediate layer 62b comprises multiple intermediate layers each having multiple nodes, and the input-side intermediate layer nodes are connected to the nodes of the second input layer 61b, the output-side nodes of the first intermediate layer 62a, and the output-side nodes of the third intermediate layer 62c. Feature quantities of mold image data, resin information, and feature quantities of molded product image data are input to the second intermediate layer 62b, and information related to adjusting molding conditions is extracted. Feature quantities of molded product image data are information equivalent to defect information. The configuration of the output layer 63 is the same as in embodiment 1.

In step S117, the processor 41 according to the second embodiment acquires molded product image data from the imaging device 9. Then, in step S118, the processor 41 inputs the acquired mold image data, resin information, and molded product image data into the condition setting learning model 6. The processor 41 determines the molding condition items to be adjusted based on the degree of adjustment necessity for each of the multiple items output from the condition setting learning model 6. The following processing is the same as in the first embodiment.

However, the processing of steps S124 and S125 can be performed using the condition-setting learning model 6. Specifically, the processor 41 acquires molded product image data obtained by imaging the molded product molded in the processing of step S122. The processor 41 then inputs the mold image data, resin information, and molded product image data into the condition-setting learning model 6. If the degree of adjustment necessity output from the condition-setting learning model 6 is equal to or greater than a predetermined value, the processor 41 determines that there is a defect in step S125, and if the degree of adjustment necessity is less than the predetermined value, it determines that there is no defect in step S125.

With the injection molding machine 1 according to embodiment 2 configured in this way, molding conditions can be automatically determined simply by inputting mold image data and resin information.

1 Injection molding machine 2 Mold clamping device 3 Injection device 4 Control device 5 Computer program 50 Recording medium 6 Conditioning learning model 7 Operation history DB

Claims (6)

A molding condition adjusting method for adjusting molding conditions of an injection molding machine, comprising:
Obtaining mold information including image data relating to the mold cavity and indicating the characteristics of the mold , resin information including data indicating the melting point or viscosity of the raw material resin and indicating the properties of the raw material resin, and defect information including data indicating the presence or absence of burrs, sink marks, flow marks, cracks, weld lines, voids, silver streaks or shorts in the molded product and indicating defective phenomena in the molded product under the current molding conditions ,
When the mold information, resin information, and defect information are input, the acquired mold information, resin information, and defect information are input into a learning model that outputs information related to the necessity of adjusting a plurality of items that are molding conditions, the adjustment range, and the adjustment step width, thereby determining the molding condition items to be adjusted, the adjustment range, and the adjustment step width;
A molding condition adjusting method for automatically changing molding conditions and controlling molding based on the determined molding condition items to be adjusted, the adjustment range, and the adjustment step width. The molding condition items include injection speed, injection pressure, heater temperature, dwell pressure, dwell switching position, injection start position, resin temperature in the mold, nozzle temperature, hopper temperature, mold clamping force, injection acceleration, injection stroke, cylinder tip resin pressure, backflow prevention ring seating state, dwell switching speed, dwell completion position, cushion position, metering back pressure, metering torque, metering completion position, screw retreat speed, cycle time, mold closing time, injection time, dwell time, metering time, mold opening time, cooling time, screw rotation speed, mold opening/closing speed, ejection speed, or number of ejections.
The molding condition adjusting method according to claim 1 . 3. The molding condition adjusting method according to claim 1 , further comprising displaying a history of changes to the molding conditions. A computer program for causing a computer to execute a process for adjusting molding conditions of an injection molding machine,
Obtaining mold information including image data relating to the mold cavity and indicating the characteristics of the mold , resin information including data indicating the melting point or viscosity of the raw material resin and indicating the properties of the raw material resin, and defect information including data indicating the presence or absence of burrs, sink marks, flow marks, cracks, weld lines, voids, silver streaks or shorts in the molded product and indicating defective phenomena in the molded product under the current molding conditions ,
When the mold information, resin information, and defect information are input, the acquired mold information, resin information, and defect information are input into a learning model that outputs information related to the necessity of adjusting a plurality of items that are molding conditions, the adjustment range, and the adjustment step width, thereby determining the molding condition items to be adjusted, the adjustment range, and the adjustment step width;
A computer program for causing the computer to execute a process of automatically changing molding conditions and executing molding control based on the determined molding condition items to be adjusted, adjustment ranges, and adjustment step widths. A molding condition adjusting device that adjusts molding conditions of an injection molding machine,
an acquisition unit that acquires mold information including image data related to the mold cavity and indicating the characteristics of the mold , resin information including data indicating the melting point or viscosity of the raw material resin and indicating the properties of the raw material resin, and defect information including data indicating the presence or absence of burrs, sink marks, flow marks, cracks, weld lines, voids, silver streaks or shorts in the molded product and indicating defects in the molded product under the current molding conditions ;
a determination unit that determines molding condition items to be adjusted, adjustment ranges and adjustment increments by inputting the acquired mold information, resin information and defect information into a learning model that outputs information related to the necessity of adjusting a plurality of molding condition items, the adjustment ranges and adjustment increments when the mold information, resin information and defect information are input ;
and a control unit that executes molding control while automatically changing the molding conditions based on the determined molding condition items to be adjusted, the adjustment range, and the adjustment step width. An injection molding machine comprising the molding condition adjusting device according to claim 5 .