CONDITION DETERMINATION DEVICE AND CONDITION DETERMINATION METHOD

Abstract

State determining device (1) which determines an operating state of an injection molding machine (2), wherein the state determining device a data acquisition unit (30) that acquires data related to an injection molding machine (2); a learning model storage unit (130) that stores a learning model obtained by learning an operating state of an injection molding machine (2) with respect to data related to the injection molding machine (2); an estimation unit (120) that estimates an abnormality degree of the operating state based on data acquired by the data acquisition unit (30) using the learning model stored in the learning model storage unit (130); a correction coefficient storage unit (52) that stores a correction coefficient in association with at least one of a type of injection molding machine (2) and a machine device attached to the injection molding machine (2); and a numerical conversion unit (34) that acquires a correction coefficient stored in the correction coefficient storage unit (52) based on at least one of a type of injection molding machine (2) whose data was acquired by the data acquisition unit (30) and a machine device attached to the injection molding machine (2), and numerically converts and corrects an estimation result of the degree of anomaly obtained by the estimation unit (120) using a predetermined correction function to which the acquired correction coefficient is applied, includes.

Description

BACKGROUND OF THE INVENTION 1. Field of the invention

[0001] The present invention relates to a condition determining device and a condition determining method, and more particularly relates to a condition determining device and a condition determining method for supporting the maintenance of injection molding machines. 2. Description of the state of the art

[0002] Industrial machinery, such as injection molding machines, requires maintenance either regularly or when an abnormality occurs. During industrial machinery maintenance, maintenance personnel determine whether an operating condition of the industrial machine is normal or abnormal using a physical quantity that represents operating conditions previously recorded during the operation of the industrial machine, and then perform maintenance activities such as replacing a part that exhibits an abnormality.

[0003] For example, as a maintenance activity for a check ring of an injection cylinder in an injection molding machine, a method is known in which a screw is periodically removed from the injection cylinder to directly measure the dimension of the check ring. However, performing the measurement activity in this process requires temporarily stopping production, which adversely reduces productivity.

[0004] In addition, there are a wide variety of injection molding machines, such as an injection molding device comprising an injection cylinder, a mold clamping device, and a molded article ejection device, which have different specifications from each other. Therefore, it is necessary to prepare as many condition determination devices and criteria for determining the presence of an abnormality as there are types of injection molding machines.

[0005] Conventional techniques for solving this problem include techniques for detecting a rotational torque acting on a screw and a backflow phenomenon in which resin flows to the rear of the screw to indirectly measure the wear amount of a check ring of an injection cylinder and diagnose an abnormality without temporarily stopping production, such as when a screw is removed from the injection cylinder. For example, JP H01-168421 A discloses a technique for measuring a rotational torque acting on a screw rotation direction, and determining an abnormality when the rotational torque is not within an allowable range. Furthermore, JP 2007-30221 A and JP 2017-202632 A disclose techniques for diagnosing an abnormality by supervised learning of a load on a drive unit, a resin pressure, and the like.

[0006] However, the above-described technique disclosed in JP H01-168421 A has a problem that, for machines having different specifications such as a rated torque and inertia of a motor constituting a drive unit of an injection molding machine and a reduction ratio of a reduction gear, an operation for setting allowable ranges used for determining an abnormality is required.

[0007] Furthermore, the above-described techniques disclosed in JP 2007-30221 A and JP 2017-202632 A have a problem that a deviation between measured values ​​obtained from machines with different specifications of components constituting drive units of injection molding machines and numerical values ​​of learning data input during machine learning is too large to make a correct determination by machine learning. In particular, there is a problem that a learning model obtained by machine learning cannot be generally applied to a wide variety of injection molding machines.

[0008] In addition, there is a problem that when the type of resin that is the raw material of a molded article to be manufactured by an injection molding machine and the types of auxiliary equipment of the injection molding machine such as a mold, a mold temperature adjusting machine, and a resin dryer are different from those in machine learning, a deviation is generated between the measurement values ​​obtained by the machine and measurement values ​​used in constructing the learning model due to the influence of the differences between the types, and accordingly, a determination as to whether or not an abnormality exists cannot be correctly made by machine learning.

[0009] It is well known that machine learning is performed in the creation of machine learning models under as many different types of learning conditions as there are combinations of machine devices—such as a motor, a gear reducer, and a moving unit—that constitute an injection molding machine, in order to improve the determination accuracy of machine learning. However, performing machine learning with various types of injection molding machines, auxiliary devices, and components is expensive. Furthermore, raw materials such as resins and workpieces must be prepared when the machines are running, which requires high costs for the raw materials used in obtaining learning data. Furthermore, the activities of obtaining learning data are time-consuming. This poses the problem of inefficient collection of learning data.

[0010] Furthermore, reference is made to the disclosure in US 2017 / 0 028 593 A1 and in DE 43 01 130 A1. SUMMARY OF THE INVENTION

[0011] An object of the present invention is to provide a condition determination apparatus and a condition determination method, whereby the maintenance of various injection molding machines can be supported without requiring high costs.

[0012] According to the present invention, the above-described problems are solved by deriving an abnormality degree correction value, which is obtained by performing numerical conversion such as adding a predetermined correction amount to an abnormality degree estimated value of a drive unit of an injection molding machine, for an abnormality degree estimated by machine learning. The abnormality degree estimate is derived by performing machine learning on a physical quantity (e.g., current and speed) acquired as learning data from a control unit in a time series.

[0013] More specifically, an abnormality degree correction value is derived by numerically converting an abnormality degree estimation value such that differences between types of machines and auxiliary equipment are absorbed even when the types of injection molding machines are different from each other, and when the auxiliary equipment of the injection molding machines and the types of resins that are manufacturing materials are different from each other, and particularly when the sizes of large-size and small-size machines are different, and when components of the injection molding machines such as injectors, mold clamps, injection cylinders, screws, and motors are different, thereby realizing a means for determining the presence of an abnormality by generally and efficiently applying a learning model to various types of injection molding machines.

[0014] Further, based on an abnormality degree obtained as an output of the machine learning, means are provided for displaying a message or an icon expressing a state of an abnormality on a display device, stopping an operation of a movable unit of a machine to ensure safety for an operator when an abnormality degree is equal to or greater than a predetermined value, decelerating a motor driving the movable unit to allow safe operation of the movable unit, and limiting a drive torque of the motor to a low value.

[0015] A state determination device according to one aspect of the present invention determines an operating state of an injection molding machine and includes a data acquisition unit that acquires data associated with an injection molding machine; a learning model storage unit that stores a learning model obtained by learning an operating state of an injection molding machine with respect to data associated with the injection molding machine; an estimation unit that estimates based on data acquired by the data acquisition unit using the learning model stored in the learning model storage unit; a correction coefficient storage unit that stores a correction coefficient associated with at least one of a type of injection molding machine and a machine device attached to the injection molding machine;and a numerical conversion unit that acquires a correction coefficient stored in the correction coefficient storage unit based on at least one of a type of an injection molding machine whose data was acquired by the data acquisition unit and a machine device attached to the injection molding machine, and numerically converts and corrects an estimation result obtained by the estimation unit using a predetermined correction function to which the acquired correction coefficient is applied;

[0016] The learning model can be learned by at least one of supervised learning, unsupervised learning and reinforcement learning.

[0017] The correction function may be at least one of a polynomial function and a rational function.

[0018] The data acquisition unit may acquire data related to each of the multiple injection molding machines connected to each other via a wired / wireless network.

[0019] A state determination method according to another aspect of the present invention is a method for determining an operating state of an injection molding machine, and includes a data acquisition step for acquiring data related to an injection molding machine; an estimation step for performing estimation based on data acquired in the data acquisition step using a learning model obtained by learning an operating state of an injection molding machine with respect to data related to the injection molding machine;and a numerical conversion step for numerically converting and correcting an estimation result obtained in the estimation step using a predetermined correction function to which a correction coefficient has been applied, and which is connected to at least one of a type of injection molding machine whose data was acquired in the data acquisition step and a machine device attached to the injection molding machine;

[0020] According to the present invention, an abnormality degree representing a state of an injection molding machine output during estimation is converted depending on a type of injection molding machine that is the object of determination and a machine device attached to the injection molding machine, and an abnormality determination for the injection molding machine is made based on the result obtained by the conversion, without performing learning by collecting learning data from various types of machines. Accordingly, various states of injection molding machines can be estimated without incurring high learning costs. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic hardware configuration diagram illustrating a state determination device according to an embodiment. Fig. Figure 2 is a schematic functional block diagram illustrating the state determination device during learning. Fig. 3 is a schematic functional block diagram illustrating the state determination device according to an embodiment. Fig. 4 illustrates an example of a correction coefficient stored in a correction coefficient storage unit. Fig. 5 illustrates another example of a correction coefficient stored in the correction coefficient storage unit. Fig. Figure 6 illustrates an example of an interface for setting correction coefficients. Fig. 7 illustrates a display example of an anomaly condition. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

[0021] Fig. 1 is a schematic hardware configuration diagram showing main portions of a state determination apparatus including a machine learning apparatus according to an embodiment.

[0022] A state determination device 1 according to the present embodiment can be installed, for example, on a control unit that controls an injection molding machine. Furthermore, the state determination unit 1 can also be installed as a personal computer provided with a control unit that controls an injection molding machine, a management device 3 connected to the control unit via a wired / wireless network, or a computer such as an edge computer, a cell computer, a host computer, and a cloud server. The present embodiment provides the description of an example in which the state determination device 1 is installed as a personal computer provided with a control unit that controls an injection molding machine.

[0023] A CPU 11 included in the state determination device 1 according to the present embodiment is a processor for controlling the state determination device 1 as a whole. The CPU 11 reads a system program stored in a ROM 12 via a bus 20 and controls the entire state determination device 1 according to the system program. A RAM 13 temporarily stores transient calculation data, various types of data input by an operator via an input device 71, and the like.

[0024] A non-volatile memory 14 consists of a memory, a solid-state hard disk (SSD), or the like, which is supported, for example, by a battery (not shown) so that its storage state is maintained even when the state determination device 1 is turned off.The non-volatile memory 14 stores a setting area in which setting information related to an operation of the state determination device 1 is stored, data input from the input device 71, various types of data acquired from an injection molding machine 2 (for example, a machine type, the mass and material of a mold, and the type of a resin), time series data of various types of physical quantities (for example, a temperature of a nozzle; a position, a speed, a current, a voltage, and a torque of a prime mover that drives the nozzle; a temperature of a mold; and a flow rate, a flow rate, and a pressure of the resin) acquired in a molding operation by the injection molding machine 2, data read via an external storage device (not shown) and a network, and the like.Programs and various types of data stored in the non-volatile memory 14 can be loaded into the RAM 13 when the programs and data are executed or used.

[0025] In the ROM 12, a known analysis program for analyzing various types of data and a system program including, for example, a program for controlling communication with a machine learning device 100 described later are written in advance.

[0026] The condition determination device 1 is connected to a wired / wireless network 7 via an interface 16. At least one injection molding machine 2, the management device 3 that manages a manufacturing activity performed by the injection molding machine 2, and the like are connected to the network 7 and exchange data with the condition determination device 1.

[0027] The injection molding machine 2 is a machine for manufacturing a product molded using a resin such as plastic, and melts the resin, which is the material, and fills a mold with the molten resin (injects the molten resin into the mold) to mold the resin. The injection molding machine 2 is composed of various machine devices such as a nozzle, a prime mover (for example, a motor), a transmission mechanism, reduction gears, and a movable unit. A state of each of the components is detected by a sensor or the like, and an operation of each of the components is controlled by a control unit. Examples of the prime mover used in the injection molding machine 2 include an electric motor, a hydraulic cylinder, a hydraulic motor, and an air motor.Further, examples of the transmission mechanism used in the injection molding machine 25 include a ball screw, gears, pulleys, and belts.

[0028] Each piece of data read into a memory, data obtained as results of program execution or the like, data output from the machine learning device 100 described later, and the like are output via an interface 17 for display on a display device 70. Furthermore, the input device 71, which consists of a keyboard, a pointing device, or the like, transmits a command, data, and the like based on an operator's operation to the CPU 11 via an interface 18.

[0029] An interface 21 is an interface for connecting the state determination device 1 to the machine learning device 100. The machine learning device 100 includes a processor 101 for controlling the entire machine learning device 100, a ROM 102 for storing a system program and the like, a RAM 103 for performing temporary storage during each processing related to machine learning, and a non-volatile memory 104 for storing learning models and the like.The machine learning device 100 is capable of observing each piece of information (for example, various types of data such as a type of injection molding machine 2, the mass and material of a mold, and the type of resin; and time series data of various types of physical quantities (a temperature of a nozzle; a position, a speed, an acceleration, a current, a voltage, and a torque of a prime mover that drives the nozzle; a temperature of a mold; and a flow rate, a flow rate, and a pressure of the resin)) that can be obtained by the state determination device 1 via the interface 21.In addition, the state determination device 1 acquires processing results output from the machine learning device 100 via the interface 21, and stores and displays the acquired results, and transmits the acquired results to other devices via a network not shown or the like.

[0030] Fig. 2 is a schematic functional block diagram illustrating the state determination device 1 and the machine learning device 10 during learning.

[0031] The Fig. The state determination device 1 shown in Figure 2 comprises components required for the learning (learning mode) performed by the machine learning device 100. The Fig. 2 are implemented when the CPU 11 contained in the state determination device 1 and the processor 101 of the machine learning device 100, which is shown in Fig. 1, execute respective system programs, and respectively control an operation of each unit of the state determination device 1 and each unit of the machine learning device 100.

[0032] The Fig. The state determination device 1 shown in Figure 2 includes a data acquisition unit 30 and a preprocessing unit 32, and the machine learning device 100 included in the state determination device 1 includes a learning unit 110. Furthermore, an acquisition data storage unit 50 is provided in the nonvolatile memory 14, which stores data acquired from external machines and the like. A learning model storage unit 130 is provided in the nonvolatile memory 14 of the machine learning device 100, which stores learning models constructed by the machine learning performed by the learning unit 110.

[0033] The data acquisition unit 30 acquires various kinds of data input from the injection molding machine 2, the input device 71, and the like.

[0034] The data acquisition unit 30 acquires various types of data such as the type of the injection molding machine 2, the mass and material of a mold, and the type of resin; time series data of various types of physical quantities (such as a temperature of a nozzle; a position, a speed, an acceleration, a current, a voltage, and a torque of a prime mover that drives the nozzle; a temperature of a mold; and a flow rate, a flow rate, and a pressure of the resin); and various types of data such as, for example, information input by an operator related to a maintenance job for the injection molding machine, and stores these pieces of data in the acquisition data storage unit 50.When acquiring time series data, the data acquiring unit 30 sets time series data acquired in a predetermined time range (for example, the range of one step) as a time series data set based on an output of signal data acquired from the injection molding machine 2 and a change in other time series data, and stores the time series data in the acquiring data storage unit 50. The data acquiring unit 30 may acquire data from other computers via an external storage device (not shown) or the wired / wireless network 7.

[0035] The preprocessing unit 32 generates data for use in the learning performed by the machine learning device 100 based on the acquisition data stored in the acquisition data storage unit 50. The preprocessing unit 32 generates data obtained by converting (e.g., quantifying, sampling) the acquisition data stored in the acquisition data storage unit 50 into data with a uniform format for use in the machine learning device 100.When the machine learning device 100 performs unsupervised learning, the preprocessing unit 32 generates state data S having a predetermined format for the unsupervised learning; when the machine learning device 100 performs supervised learning, the preprocessing unit 32 generates a set of state data S and label data L having a predetermined format for the supervised learning; and when the machine learning device 100 performs reinforcement learning, the preprocessing unit 32 generates a set of state data S and determination data D having a predetermined format for the reinforcement learning.

[0036] The learning unit 100 performs machine learning using data generated by the preprocessing unit 32. The learning unit 110 generates learning models by performing machine learning using data acquired from the injection molding machine 2 using a known machine learning method such as unsupervised learning, supervised learning, and reinforcement learning, and stores the generated learning models in the learning model storage unit 130. Examples of unsupervised learning performed by the learning unit 110 include the autoencoder method and the k-means method. Examples of supervised learning include the multilayer perceptron method, the recurrent neural network method, the long-short-time memory method, and the convolutional neural network method. Examples of reinforcement learning include Q-learning.

[0037] The learning unit 110 performs unsupervised learning based on data obtained by converting acquisition data acquired from the injection molding machine 2 in a normal operating state and stored in the acquisition data storage unit 50 by the preprocessing unit 32. Therefore, the learning unit 110 is capable of generating, for example, a distribution of data acquired in a normal state as a learning model. With the thus generated learning model, an estimation unit 120 described later can estimate how much the data obtained by converting acquisition data acquired from the injection molding machine 2 by the preprocessing unit 32 deviates from the data obtained in a normal operating state, and thereby calculate an abnormality degree as an estimation result.

[0038] Or, for example, the learning unit 110 performs supervised learning using data obtained by assigning a normal label to acquisition data acquired from the injection molding machine 2 in a normal operating state and assigning an abnormal label to acquisition data acquired from the injection molding machine 2 before and after the occurrence of an abnormality (and converting the acquisition data by the preprocessing unit 32), and can generate a discrimination boundary between normal data and abnormal data as a learning model.By the learning model thus generated, the estimation unit 120 described later can estimate whether the data obtained by converting acquisition data acquired from the injection molding machine 2 by the preprocessing unit 32 belongs to normal data or abnormal data, and calculate a label value (normal / abnormal) as an estimation result and the reliability of the label value.

[0039] In the state determination device 1 having the above-described configuration, the learning unit 110 performs learning using data acquired from the injection molding machine 2. Data used for learning by the learning unit 110 may be, for example, data acquired from one injection molding machine 2; specifically, data acquired from multiple injection molding machines including different machine devices need not be used. A learning model generated by the learning unit 110 is used to estimate a state of an injection molding machine, which is performed by the estimation unit 120 described later.However, there is no limitation to the estimation for one injection molding machine used for learning, the learning model may be used for estimation of the state of another injection molding machine based on data obtained from the other injection molding machine, wherein numerical conversion of the estimation results is performed by a numerical conversion unit 34 described later.

[0040] Fig. 3 is a schematic functional block diagram illustrating the state determination device 1 and the machine learning device 100 according to a first embodiment.

[0041] The state determination device 1 according to the present embodiment has the structure required for the estimation (estimation mode) performed by the machine learning device 100. The Fig. 3 are implemented when the CPU 11 contained in the state determination device 1 and the processor 101 of the machine learning device 100, which is shown in Fig. 1, execute respective system programs, and respectively control an operation of each unit of the state determination device 1 and each unit of the machine learning device 100.

[0042] The state determination device 1 according to the present embodiment has, as in the case of the state determination device 1 shown in Fig. 2, the machine learning device 100 includes the data acquisition unit 30 and the preprocessing unit 32, and further includes the numerical conversion unit 34, and the machine learning device 100 included in the state determination device 1 includes the estimation unit 120. Furthermore, the nonvolatile memory 14 includes the acquisition data storage unit 50 that stores data used for the state estimation performed by the machine learning device 100, and a correction coefficient storage unit 52 that stores correction amounts used for the numerical conversion performed by the numerical conversion unit 34. The nonvolatile memory 104 of the machine learning device 100 includes the learning model storage unit 130 that stores learning models constructed by the machine learning performed by the learning unit 100.

[0043] The data acquisition unit 30 according to the present embodiment has the same functions as those in Fig. 2 shown data acquisition unit 30.

[0044] The preprocessing unit 32 according to the present embodiment generates, from the acquisition data stored in the acquisition data storage unit 50, state data S having a predetermined format to be used for the estimation performed by the machine learning device 100. The preprocessing unit 32 generates state data obtained by converting (e.g., quantifying, sampling) the acquisition data stored in the acquisition data storage unit 50 into data having a uniform format for use in the machine learning device 100.

[0045] The estimation unit 120 estimates a state of an injection molding machine using a learning model stored in the learning model storage unit 130 based on the state data S generated by the preprocessing unit 32. The estimation unit 120 according to the present embodiment estimates and calculates an abnormality degree related to a state of the injection molding machine by inputting the state data S input from the preprocessing unit 32 into the learning model (whose parameters have been determined) generated by the learning unit 110, thereby estimating and calculating an abnormality degree related to a state of the injection molding machine.

[0046] The numerical conversion unit 34 numerically converts a result estimated by the estimation unit 120. The numerical conversion unit 34 may numerically convert a result estimated by the estimation unit 120, for example, using a predetermined correction function. In this case, correction coefficients of the correction function are stored in advance in the correction coefficient storage unit 52 such that the correction coefficients are associated with respective types of injection molding machines and respective machine devices mounted on the injection molding machines.Then, the numerical conversion unit 34 acquires correction coefficients associated with a type and a machine device of the injection molding machine 2, which is the object of determination, from the correction coefficient storage unit 52, and numerically converts a result estimated by the estimation unit 120 with a correction function to which the acquired correction coefficients are applied.

[0047] Examples of a correction function set in the numerical conversion unit 34 include a polynomial function and a rational function. The following formula (1) is an example of a polynomial function used as a correction function. In formula (1), x denotes an anomaly degree as the result estimated by the estimation unit 120, a and b denote correction coefficients, and y denotes an anomaly degree after correction. y=ax+b

[0048] When the correction function exemplified by the above formula (1) is used, the correction coefficient a and the correction coefficient b are calculated in advance as shown in Fig. 4 and Fig. 5 are stored in the correction coefficient storage unit 52 in such a way that they are associated with types and machine devices of injection molding machines. When the correction coefficients shown as examples in Fig. 4 and Fig. 5, the numerical conversion unit 34 determines and acquires the correction coefficients a and b based on the type of injection molding machine that is the object of the condition determination and a screw diameter, and numerically converts the abnormality degree estimated by the estimation unit 120 by the formula (1) to which the acquired correction coefficients are applied, and outputs the converted abnormality degree.Here, the correction coefficients stored in the correction coefficient storage unit 52 may be set in such a manner as to be associated only with one type of the injection molding machine 2, set in such a manner as to be associated only with the machine equipment attached to the injection molding machine 2, or set in such a manner as to be associated with circumstances affecting operation of other injection molding machines 2.

[0049] Regarding the correction functions used by the numerical conversion unit 34 and the correction coefficients stored in the correction coefficient storage unit 52, several types of injection molding machines are operated in a normal state and in an abnormal state while their machine devices are exchanged, and respective physical quantities are observed to plot respective abnormality degrees calculated based on the physical quantities, and each correction coefficient can be calculated based on a relationship of abnormality degrees calculated under the same abnormal conditions when respective types of machines and respective machine devices were used.so that, for any type of injection molding machine and for injection molding machines equipped with any type of machine device, the same degree of abnormality is obtained with respect to the same abnormal condition. Furthermore, in this case, a structure may be employed in which correction coefficients stored in the correction coefficient storage unit 52 are stored, for example, via a memory device shown in , Fig. 6, which is used to set the correction coefficients. Once obtained, correction coefficients express a trend of the anomaly levels calculated with respect to respective combinations of an injection molding machine, a machine setup, and the like, and can be used not only for a single learning model but for various learning models.

[0050] Results (e.g., abnormality levels related to states of injection molding machines) obtained through estimation by the estimation unit 120 and numerical conversion by the numerical conversion unit 34 may be output to the display device 70 for display, and may be sent to a host computer, a cloud computer, and the like via a wired / wireless network (not shown) and output for use. And when a result estimated by the estimation unit 120 is in a predetermined state (e.g., when an abnormality level obtained through numerical conversion performed by the numerical conversion unit 34 exceeds a predetermined threshold), the state determination unit 1 may, for example, Fig.7, perform a display output by a warning message or an icon on the display device 70, or issue a command to stop or slow down an operation, a command to limit the torque of a motor, or the like to the injection molding machine.

[0051] In the condition determination device 1 having the above-described configuration, the estimation unit 120 performs an estimation of a condition of the injection molding machine 2 based on acquisition data acquired from the injection molding machine 2. Then, the numerical conversion unit 34 numerically converts an abnormality degree of the injection molding machine 2 estimated by the estimation unit 120 using a correction function to which correction coefficients set in such a manner as to be associated with a type and machine device of the injection molding machine 2 are applied, and an abnormality degree of the injection molding machine 2 is determined based on the result obtained by the conversion.Learning models stored in the learning model storage unit 130 are learned based on data acquired from an injection molding machine 2 of a reference type to which a reference machine device is attached. However, an abnormality degree as an estimation result obtained by the estimation unit 120 based on data acquired from an injection molding machine of a different type and having different machine devices than that of that injection molding machine 2 is converted using a correction function to which correction coefficients set in such a way as to be associated with the type and machine device of the injection molding machine are applied; and using a predetermined threshold value, it can be determined whether that injection molding machine is in a normal state or an abnormal state.

[0052] So far, one embodiment of the present invention has been described. However, the present invention is not limited to the examples of the above-described embodiment, but can be embodied in various forms by appropriately adding modifications.

[0053] For example, the state determination device 1 and the machine learning device 100 according to the above embodiment are devices each having different CPUs (processors), but the machine learning device 100 can be implemented by the CPU 11 included in the state determination device 1 and the system program stored in the ROM 12.

[0054] And when a plurality of injection molding machines 2 are connected to each other via a network, the operating states of the injection molding machines 2 may be determined by a single state determining device 1, or state determining devices 1 may be set up in respective control units provided in the injection molding machines 2, and the operating states of respective injection molding machines 2 may be determined by respective state determining devices 1 provided in the injection molding machines 2.

Claims

[1] State determination device (1) which determines an operating state of an injection molding machine (2), wherein the state determination device a data acquisition unit (30) that acquires data related to an injection molding machine (2); a learning model storage unit (130) which stores a learning model obtained by learning an operating state of an injection molding machine (2) with respect to data related to the injection molding machine (2); an estimation unit (120) which, based on data obtained by the data acquisition unit (30), makes an estimate of an anomaly level of the operating state using the learning model stored in the learning model storage unit (130); a correction coefficient storage unit (52) which stores a correction coefficient in conjunction with at least one from a type of injection molding machine (2) and a machine device attached to the injection molding machine (2); and a numerical conversion unit (34) which, on the basis of at least one type of injection molding machine (2), whose data were obtained by the data acquisition unit (30), and a machine device attached to the injection molding machine (2), obtains a correction coefficient stored in the correction coefficient storage unit (52), and numerically converts and corrects an estimation result of the degree of anomaly obtained by the estimation unit (120) using a predetermined correction function to which the obtained correction coefficient is applied, includes. [2] State determination device (1) according to claim 1, wherein the learning model is learned by at least one learning method consisting of supervised learning, unsupervised learning and reinforcement learning. [3] State determination device (1) according to claim 1, wherein the correction function is at least one of a polynomial function and a rational function. [4] State determination device (1) according to claim 1, wherein the data acquisition unit (30) of several injection molding machines (2) which are interconnected via a wired / wireless network (7) acquires data relating to each of the several injection molding machines (2). [5] State determination method for determining an operating state of an injection molding machine (2), wherein the state determination method a data acquisition step to obtain data related to an injection molding machine; an estimation step to perform an estimation of an anomaly degree of the operating state based on data obtained in the data acquisition step, using a learning model obtained by learning an operating state of an injection molding machine (2) with respect to data related to the injection molding machine (2); and a numerical conversion step for numerically converting and correcting an estimation result of the degree of anomaly obtained in the estimation step, using a predetermined correction function to which a correction coefficient has been applied, and which is connected with at least one of a type of injection molding machine (2), the data of which were obtained in the data acquisition step, and a machine device attached to the injection molding machine (2), includes.

Images