Neutron quality estimation system, neutron quality estimation method, program

Abstract

To provide a core quality estimation system estimating the quality of a core at high precision while considering an influence of a circumferential environment.SOLUTION: A core quality estimation system is used for estimating the quality of a core molded by filling a mold with kneaded sand kneaded in a kneading tank and comprises: a mold temperature information acquiring part acquiring the mold temperature information of the mold; an environment information acquiring part acquiring environment information related to a circumferential environment in which the core is molded; and an estimation part estimating the quality of the core based on the mold temperature information and the environment information.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a core quality estimation system, a core quality estimation method, a program, a trained model, and a machine learning machine. [Background technology]

[0002] Generally, in a core-making device used to make cores for casting, a mixed sand made of the raw materials for the core - core sand, water glass, water, surfactants, etc. - is filled into a mold and heated to make the core.

[0003] Patent document 1 discloses a technology in which, in such a core-making device, the temperature of the mold is measured when mixed sand is filled into the mold, and if the difference between this temperature and the firing temperature, which is the appropriate temperature for making a core, is greater than or equal to a predetermined value, an abnormality is signaled, thereby indicating a possible deterioration in the quality of the core. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2020-110811 A Summary of the Invention [Problem to be solved by the invention]

[0005] Here, the inventors have found the following problems in quality control of the core molding device. In the core making machine, core sand, water glass, water, surfactant, etc. are mixed in a pre-calculated optimal ratio, but since water evaporates during mixing, the amount of water contained in the mixed sand when it is actually filled into a mold is affected by the surrounding environment, such as temperature and humidity. Furthermore, the kinetic viscosity of the mixed sand changes depending on the amount of water contained in it, so as a result, the behavior of the mixed sand when it is filled into a mold is affected by the surrounding environment.

[0006] However, the quality control of the core formation apparatus described in Patent Document 1 does not take into account such influences of the surrounding environment, and there is room for improvement in the quality control of the core formation apparatus.

[0007] The present invention has been made in consideration of these circumstances, and provides a core quality estimation system, a core quality estimation method, a program, and a trained model that can estimate core quality with greater accuracy, as well as a machine learning device that can learn core quality. [Means for solving the problem]

[0008] A core quality estimation system according to one aspect of the present invention comprises: This is a core quality estimation system that estimates the quality of a core that is formed by filling a mold with mixed sand that has been mixed in a mixing tank and heating it, and includes a mold temperature information acquisition unit that acquires mold temperature information of the mold, an environmental information acquisition unit that acquires environmental information related to the surrounding environment in which the core is formed, and an estimation unit that estimates the quality of the core based on the mold temperature information and the environmental information.

[0009] With this configuration, the quality of the core to be formed is estimated taking into account environmental information regarding the surrounding environment in which the core is formed, so that the quality of the core can be estimated more accurately by taking into account changes in the moisture content of the mixed sand.

[0010] In the core quality estimation system of the above aspect, the environmental information may include at least one of humidity information and atmospheric pressure information of the surrounding environment. With this configuration, by taking into account at least one of the humidity information and the atmospheric pressure information, the quality of the core can be estimated more accurately based on changes in the moisture content of the mixed sand.

[0011] In the core quality estimation system of the above aspect, the environmental information acquisition unit may acquire the environmental information based on position information for molding the core. With this configuration, environmental information is acquired based on position information, making it unnecessary to provide a sensor or the like for acquiring the environmental information, thereby enabling costs to be reduced.

[0012] In the core quality estimation system of the above aspect, the estimation unit may estimate the quality of the core to be molded from the mixed sand before filling of the mixed sand into the mold is completed. With this configuration, estimation is performed before the filling of the mixed sand is completed, so that the mold temperature of the mold can be controlled based on the estimation result, thereby preventing defects from occurring in the core.

[0013] In the core quality estimation system of the above aspect, the estimation unit may estimate the quality of the core to be formed from the mixed sand before mixing of the mixed sand in the mixing tank is completed. With this configuration, the estimation is performed before the mixing of the sand in the mixing tank is completed, so that the amount of water input during mixing in the mixing tank can be controlled and defects in the core can be prevented.

[0014] In the core quality estimation system of the above aspect, the estimation unit may estimate the quality of the core by using CAE to analyze the behavior of mixed sand in a mold. With this configuration, CAE can be used to analyze the behavior of the mixed sand inside the mold, making it possible to estimate the quality of the core based on the behavior inside the mold.

[0015] In the core quality estimation system of the above aspect, the estimation unit may estimate the quality of the core based on mold temperature information of the mold when the mixed sand is filled or heated, obtained using the behavior analysis. With this configuration, it is possible to estimate the quality of the core based on mold temperature information from an analysis of the behavior of the mixed sand inside the mold, and thereby estimate the quality of the core that is affected by the mold temperature.

[0016] In the core quality estimation system of the above aspect, the estimation unit may estimate the quality of the core based on behavior information of the mixed sand when it is filled, obtained using the behavior analysis. With this configuration, it is possible to estimate the quality of the core based on behavior information of the mixed sand when it is filled from an analysis of the behavior of the mixed sand in the mold, and thereby estimate the quality of the core that is affected by the behavior of the mixed sand.

[0017] In the core quality estimation system of the above aspect, the estimation unit may estimate the quality of the core based on the mold temperature information and the environmental information by using a trained model that has been machine-learned using teacher data that inputs mold temperature information and environmental information and outputs quality information related to the quality of the core. With this configuration, the quality of the core to be molded is estimated using a trained model that has previously learned the relationship between mold temperature information, environmental information, and core quality information, so that the quality of the core can be estimated more accurately taking into account changes in the moisture content of the mixed sand.

[0018] The core quality estimation system of the above aspect may further include a correction receiving unit that accepts corrections to the estimation result by the estimation unit, and a model update unit that updates the trained model based on the corrected content when the correction receiving unit accepts the correction. With this configuration, the trained model can be suitably retrained, and the quality of the core can be estimated with greater accuracy.

[0019] The core quality estimation system of the above aspect may further include a control unit that controls the molding conditions for molding the core based on the estimation result by the estimation unit. With this configuration, defects occurring in the core can be prevented by controlling the molding conditions based on the estimation results by the estimation unit.

[0020] In the core quality estimating system of the above aspect, the control unit may control at least one of the mold temperature of the mold and the amount of water input during kneading in the kneading tank. With this configuration, defects occurring in the core can be prevented by adjusting the mold temperature and the amount of water added.

[0021] Further, a method for estimating core quality according to one aspect of the present invention includes the steps of: This is a core quality estimation method that uses a computer to estimate the quality of a core that is formed by filling a mold with mixed sand that has been mixed in a mixing tank and heating it, and includes a mold temperature information acquisition step of acquiring mold temperature information of the mold, an environmental information acquisition step of acquiring environmental information related to the surrounding environment in which the core is formed, and an estimation step of estimating the quality of the core based on the mold temperature information and the environmental information.

[0022] In addition, a program according to one aspect of the present invention includes: This is a program for executing a core quality estimation method for estimating the quality of a core that is formed by filling a mold with mixed sand that has been mixed in a mixing tank and heating it, and causes a computer to execute the following steps: a mold temperature information acquisition step for acquiring mold temperature information of the mold; an environmental information acquisition step for acquiring environmental information related to the surrounding environment in which the core is formed; and an estimation step for estimating the quality of the core based on the mold temperature information and the environmental information.

[0023] In addition, the trained model according to one aspect of the present invention is This is a trained model for causing a computer to function to output quality information regarding the quality of a core based on mold temperature information of a mold into which mixed sand is filled and environmental information regarding the surrounding environment in which the core is formed, and has been machine-learned using training data in which the mold temperature information and the environmental information are input and the quality information is output.

[0024] With this configuration, the quality of the core to be formed is estimated taking into account environmental information regarding the surrounding environment in which the core is formed, so that the quality of the core can be estimated more accurately by taking into account changes in the moisture content of the mixed sand.

[0025] Furthermore, a machine learning machine according to one aspect of the present invention includes: This is a machine learning machine that learns the quality of a core that is formed by filling a mold with mixed sand that has been mixed in a mixing tank and heating it, and includes a mold temperature information acquisition unit that acquires mold temperature information of the mold, an environmental information acquisition unit that acquires environmental information related to the surrounding environment in which the core is formed, a quality information acquisition unit that acquires quality information related to the quality of the core, and a learning unit that learns the quality of the core to be formed using the mold temperature information and teacher data that inputs the environmental information and outputs the quality information.

[0026] With this configuration, it is possible to generate a trained model that can learn the quality of the core to be produced and estimate the quality of the core to be produced while taking into account environmental information regarding the surrounding environment. Effect of the Invention

[0027] The present invention makes it possible to provide a core quality estimation system, a core quality estimation method, a program, and a trained model that can estimate core quality with greater accuracy, as well as a machine learning device that can learn core quality. [Brief description of the drawings]

[0028] [Figure 1] FIG. 2 is an overall cross-sectional view showing an outline of the mixing of mixed sand in the core-making device. [Diagram 2] FIG. 2 is an overall cross-sectional view showing an outline of the core-making device when mixed sand is filled. [Diagram 3] FIG. 1 is a diagram showing an outline of the hardware configuration of a core quality estimating system according to a first embodiment. [Figure 4] FIG. 1 is a diagram showing a functional configuration of a core quality estimating system according to a first embodiment. [Diagram 5]FIG. 4 is a flow diagram of a core quality estimation process according to the first embodiment. [Figure 6] FIG. 11 is a diagram showing the functional configuration of a core quality estimating system according to a second embodiment. [Figure 7] FIG. 11 is a diagram illustrating a functional configuration of a machine learning machine according to a second embodiment. [Figure 8] FIG. 13 is a diagram showing a neural network used for machine learning according to the second embodiment. [Figure 9] FIG. 11 is a flowchart showing the flow of a learning process according to the second embodiment. [Figure 10] FIG. 11 is a flow diagram of a core quality estimation process according to a second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0029] Hereinafter, specific embodiments to which the present invention is applied will be described in detail with reference to the drawings. However, the present invention is not limited to the following embodiments. In addition, the following description and drawings are appropriately simplified for clarity of explanation.

[0030] First, before describing a core quality estimating system according to an embodiment to which the present invention is applied, a core manufacturing apparatus for manufacturing a core whose quality is estimated by the core quality estimating system will be briefly described. Figures 1 and 2 are overall cross-sectional views showing an outline of the core manufacturing apparatus.

[0031] As shown in FIGS. 1 and 2, the core making apparatus 100 includes a kneading tank 110, a raw material supplying means 120, a seat 130, a piston 140, a cylinder 150, and a die 160.

[0032] The kneading tank 110 is a container for producing mixed sand by kneading raw materials. The kneading tank 110 is, for example, a cylindrical bottomed container with an open top and a bottom. The kneading tank 110 has an inner diameter of about 150 to 350 mm and a height of about 150 to 350 mm, for example, an inner diameter of about 250 mm and a height of about 250 mm. The kneading tank 110 is supplied with liquid additives such as core sand, water glass, water, and surfactant, which are raw materials for the core, from the open top by raw material supply means 120 (core sand supply means 121, water glass supply means 122, water supply means 123, surfactant supply means 124). Here, the water glass plays the role of a binder. The binder is not limited to water glass, and inorganic substances such as clays and cement can be used.

[0033] At the bottom of the kneading tank 110, a through hole 111 is provided for ejecting the kneading material S produced by kneading the raw materials with a kneading blade (not shown) inside the kneading tank 110. In addition, the through hole 111 is provided with, for example, a rubber valve with a cut, which prevents the kneading material S from leaking out of the kneading tank 110 and opens when ejected.

[0034] The pedestal 130 is a member for supporting the kneading vessel 110. A convex portion that fits into a through hole 111 provided in the bottom of the kneading vessel 110 is formed on the upper surface of the pedestal 130, and holds down a valve provided in the through hole 111 by supporting it from below.

[0035] In this way, after the raw materials are mixed in the kneading tank 110 placed on the pedestal 130, the kneading tank 110 containing the mixing material S is transferred from the pedestal 130 to the mold 160. The transfer can be performed, for example, by a conveying device having a driving means such as a motor and a holding means for holding the kneading tank 110. Note that FIG. 1 shows a state in which the kneading tank 110 is placed on the pedestal 130, and the kneading tank 110 on the mold 160 is shown by a two-dot chain line. FIG. 2 shows a state in which the kneading tank 110 is placed on the mold 160 after the transfer, and the kneading tank 110 on the pedestal 130 is shown by a two-dot chain line.

[0036] 2, after the kneading tank 110 is placed on the metal mold 160, the kneading material S in the kneading tank 110 is injected and filled into the metal mold 160 by the piston 140. Here, the piston 140 can be moved vertically by the cylinder 150. When the cylinder 150 is lowered by a control unit (not shown), the piston 140 is lowered, and the kneading material S is injected and filled into the metal mold 160.

[0037] The kneading material S filled in the mold 160 is heated by the mold 160 whose temperature is adjusted by the mold heat controller 161 in the mold 160, and hardened along the inner wall of the mold 160, thereby forming a core having the shape of the mold. Here, the mold heat controller 161 heats, cools, and maintains the temperature of each part of the mold 160, and may be provided inside or outside the mold 160. Specifically, the mold temperature can be adjusted by a heater, a cooling water circulator, or the like. In addition, the mold 160 is provided with a temperature sensor 162, which continuously or non-continuously measures the mold temperature of the mold 160. The temperature sensor 162 may be, for example, a contact temperature sensor such as a thermocouple, or a non-contact temperature sensor. The installation location and number of the temperature sensors 162 are appropriately determined according to the shape of the mold 160, etc.

[0038] The shaped core is removed from the die 160. For example, the left die 163 is configured to be able to move relatively to the right die 164, and the core can be removed from the die 160 by moving the left die 163 away from the right die 164.

[0039] Here, we have described an example of a core making device that makes cores whose quality is estimated by the core quality estimation system of the present invention, but various modifications may be made to the configuration and making flow, and in the core quality estimation system of the present invention, the core making device that makes the cores that are the subject of estimation is not limited to the configuration of the core making device described above.

[0040] <First embodiment> Next, a description will be given of a core quality estimating system according to the first embodiment. Fig. 3 is a block diagram showing an outline of the hardware configuration of a core quality estimating system 200 according to this embodiment.

[0041] As shown in Fig. 3, core quality estimation system 200 has computer resources that are possessed by a general information processing device such as a personal computer. Specifically, it has a CPU (Central Processing Unit) 201, a ROM (Read Only Memory) 202, a RAM (Random Access Memory) 203, a HDD (Hard Disk Drive) 204, a communication interface (I / F) 205, and an input / output interface (I / F) 206. Furthermore, each of these parts is connected to each other via a data bus 207 so that they can communicate with each other.

[0042] Note that, although an example will be described here in which each unit of the core quality estimation system 200 is realized by an information processing device incorporating the control function of the core making apparatus 100, they may also be realized by an information processing device provided independently of a control device having the control function of the core making apparatus 100, and some or all of these functions may be realized in an external device such as an edge or server (cloud). Specifically, they may be realized in an in-factory server that functions as a platform for aggregating information on each piece of equipment in a factory and managing and analyzing the data, or in an inter-factory server that manages equipment across multiple factories.

[0043] CPU 201 is a microprocessor that provides overall control of core quality estimation system 200. Specifically, it reads out various control programs executed in this embodiment that are stored in ROM 202 and HDD 204, and executes these programs deployed on RAM 203. Here, an SSD (Solid State Drive) may be provided as a storage device instead of or in addition to HDD 204.

[0044] The communication interface 205 communicates between the core quality estimation system 200 and external devices. The communication can be realized by various communication technologies, whether wired or wireless. In this embodiment, for example, mold temperature information of the mold 160 is received from a temperature sensor 162 provided in the core formation apparatus 100.

[0045] The input / output interface 206 performs input / output between the core quality estimation system 200 and the outside. For example, it is equipped with an output device such as a display that displays information related to the estimated core quality, and an input device for use by an operator.

[0046] Next, a description will be given of the functional configuration of the core quality estimating system 200 according to this embodiment. Fig. 4 is a block diagram showing the functional configuration of the core quality estimating system 200 according to this embodiment.

[0047] As shown in Figure 4, core quality estimation system 200 comprises, as its functional components, a mold temperature information acquisition unit 301, an environmental information acquisition unit 302, an estimation unit 303, a control unit 304, and a display unit 305. These function when CPU 201 executes various control programs stored in ROM 202 or the like. Some or all of the functions of core quality estimation system 200 may be realized by hardware circuits.

[0048] The mold temperature information acquisition unit 301 acquires mold temperature information of the mold 160 into which the mixed sand S is filled. Specifically, the mold temperature information is acquired from a temperature sensor 162 that is provided in the mold 160 and measures the mold temperature. In addition, if the mold 160 is provided with, for example, a plurality of temperature sensors 162, position information of the temperature sensors 162 in the mold 160 can be associated with the mold temperature and acquired as mold temperature information. Here, only the mold temperature may be acquired from the temperature sensor 162, and the mold temperature information acquisition unit 301 may associate the mold temperature with the position information and acquire it as mold temperature information.

[0049] The environmental information acquisition unit 302 acquires environmental information related to the surrounding environment in which the core is to be formed. Specifically, based on the location information in which the core forming apparatus 100 is used, it can acquire environmental information such as temperature, humidity, and air pressure by utilizing weather information and weather forecast information corresponding to the location information provided from the outside. For example, when the environmental information acquisition unit 302 acquires the location information in which the core forming apparatus 100 is installed, it transmits the location information together with a weather information request to a weather information providing server connected via a communication network such as the Internet. The weather information providing server reads out weather information and the like for the location specified by the received location information from a database and transmits it to the environmental information acquisition unit 302. In this way, the environmental information acquisition unit 302 can acquire weather information and the like corresponding to the location information.

[0050] Regarding location information here, core making apparatus 100 and core quality estimating system 200 are equipped with a function for acquiring location information such as a GPS function, and location information may be acquired using the GPS function, or location information input by a user of core quality estimating system 200 may be used. By acquiring environmental information based on location information in this way, it is not necessary to install sensors or the like for acquiring environmental information, thereby reducing costs. Of course, the environmental information may also be acquired via wired or wireless communication means from sensors that measure environmental information such as temperature, humidity, and air pressure installed in the vicinity of core making apparatus 100.

[0051] The estimation unit 303 estimates the quality of the core to be molded based on the mold temperature information acquired by the mold temperature information acquisition unit 301 and the environmental information acquired by the environmental information acquisition unit 302. For example, the estimation unit 303 compares the quality of the core with a predetermined reference value and estimates whether the quality is higher than the reference value (i.e., a non-defective product) or whether the quality is below the reference value (i.e., a defective product). In this case, the reference value does not need to be one, and there may be multiple values. Furthermore, the degree of quality may be expressed by a numerical value or a rank, or information indicating the specific content of the quality may be generated.

[0052] Specifically, the quality of the core is estimated by analyzing the behavior of the mixed sand S in the mold 160 using CAE (Computer Aided Engineering). By using an analysis using CAE, it is possible to estimate quality that is difficult to estimate using a calculation formula, for example. The software used for this analysis may be appropriately selected from various commercially available analysis software, or commercially available analysis software modified or improved for this analysis.

[0053] First, the subject of the analysis is the behavior of the mixed sand S in the die 160 (including the gate portion which is the entrance of the die 160) that is filled into the die 160 by the piston 140 from the mixing tank 110, and an analysis model is constructed using drawing data (CAD data) of these target parts. The constructed analysis model is stored in the estimation unit 303, and by inputting analysis conditions to this analysis model, behavior analysis of the mixed sand S in the die 160 is performed, and the analysis results are output. Various output forms of the analysis results can be selected using various software.

[0054] Here, the behavior of the mixed material S in the mold 160 is influenced by the amount of moisture contained in the mixed sand S, the kinetic viscosity, and the like. The mixed sand S is a mixture of core sand, water glass, water, surfactant, and the like. Since moisture evaporates during mixing, the amount of moisture contained in the mixed sand S and the kinetic viscosity of the mixed sand S when actually filled into the mold are influenced by the surrounding environmental information. In this embodiment, at least mold temperature information and environmental information are input as analysis conditions to perform behavior analysis, so that the behavior analysis can be performed accurately by taking into account these influences. Here, as the environmental information, any one of temperature information, humidity information, and atmospheric pressure information may be used, or two or more pieces of information may be used in combination, such as temperature information and humidity information, humidity information and atmospheric pressure information, or temperature information and atmospheric pressure information. With this configuration, the influence on the amount of moisture contained in the mixed sand S and the kinetic viscosity is considered in detail, and the behavior analysis can be performed more accurately.

[0055] In this embodiment, behavior analysis is performed using mold temperature information and any analysis conditions other than environmental information, such as the raw material type of the mixed sand S, raw material ratios such as the amount of water added during mixing, the kinetic viscosity of the mixed sand S, and molding conditions such as the injection pressure by the piston 140, as conditions previously set or stored in the estimation unit 304, but this information can also be input during analysis to perform behavior analysis. In this case, the core quality estimation system 200 can be further equipped with a function to acquire molding conditions such as the raw material type of the mixed sand S, raw material ratios such as the amount of water, the kinetic viscosity of the mixed sand S, and the injection pressure, so that this information can be input during analysis to perform behavior analysis.

[0056] In this way, the quality of the core to be molded is estimated based on information obtained by the behavior analysis of the mixed sand S using CAE. Specifically, the quality of the core is estimated based on mold temperature information of the mold 160 when the mixed sand S is filled or heated. That is, the mold temperature of the part that contacts the mixed sand S when filling or heating greatly affects the quality of the core to be molded, so if the mold temperature deviates from the predetermined specified temperature range, defects occur. For example, if the mold temperature is higher than the specified range, defects such as breakage or cracking of the core occur. On the other hand, if the mold temperature is lower than the specified range, defects such as the core sticking to the mold 160 occur. In addition, in this embodiment, the quality of the core is estimated based on mold temperature information obtained by using the behavior analysis, so the mold temperature at the time of filling can be estimated before the mixed sand S is actually filled into the mold 160, and more detailed temperature distribution information can be estimated than by the temperature sensor 162, so the quality of the core can be accurately estimated.

[0057] Furthermore, the estimation unit 303 can estimate the quality of the core based on behavior information of the mixed sand S at the time of filling, obtained by using the behavior analysis of the mixed sand S. Specifically, the quality of the core is estimated based on flow rate information and energy information of the mixed sand S in the mold 160. That is, since the behavior information of the mixed sand S in the mold 160 greatly affects the quality of the core to be molded, defects occur when the flow rate and energy are out of the predetermined range. For example, if the kinetic viscosity of the mixed sand S is high and the flow rate is slow, defects such as sand clogging and wrinkles occur in the molded core. On the other hand, if the kinetic viscosity of the mixed sand S is low and the flow rate is fast, defects such as sticking to the mold 160 and deformation defects occur in the molded core. Therefore, the estimation unit 303 can estimate the quality of the core based on these pieces of information. The above-mentioned specified values ​​of the mold temperature, flow rate, and energy can be set by confirming the quality in advance by experiments, etc.

[0058] The control unit 304 controls the molding conditions for molding the core based on the estimation result by the estimation unit 303. Specifically, when the estimation unit 303 estimates that the core to be molded will have a defect, the control unit 304 adjusts the mold temperature of the mold 160 and the amount of water input during kneading in the kneading tank 110 to prevent the defect from occurring in the core to be molded. For example, when the estimation unit 303 estimates that the mold temperature of the mold 160 when the mixed sand S is filled is out of the specified range, the control unit 304 controls the mold heat controller 161 to control the mold temperature so that the mold temperature is within the specified range. When the flow rate is estimated to be slower than the specified range, the control unit 304 controls the water supply means 123 to adjust the amount of water input during kneading in the kneading tank 110 to control the flow rate to be within the specified range. It should be noted that the control of the molding conditions described here is merely one example, and similarly, the moisture content of the mixed sand S may be adjusted by controlling the core sand supplying means 121, the water glass supplying means 122, the surfactant supplying means 124, etc., or by adjusting the mixing time, or the injection pressure by the piston 140 may be adjusted by controlling the operation of the cylinder 150, or by controlling various other molding conditions, thereby preventing defects from occurring in the core. Here, the control unit 304 does not necessarily have to be provided in the core quality estimation system 200, and for example, defects may be prevented from occurring in the core by an operator operating a separately provided control device based on the estimation result by the estimation unit 303.

[0059] Display unit 305 displays the estimation results of core quality estimated by estimation unit 303, as well as the manufacturing conditions for preventing defects determined by control unit 304. By displaying information and warnings related to the occurrence of core quality defects using display unit 305, the information can be notified to the worker, allowing the worker to adjust the core manufacturing conditions or cancel manufacturing, for example. Also, while an example has been described here in which core quality estimation system 200 is equipped with display unit 305 as an information processing device, display unit 305 may also be provided in core manufacturing apparatus 100, or a configuration in which display unit 305 is provided on a tablet terminal held by the worker may also be used.

[0060] Next, the estimation flow in the core quality estimation system 200 according to this embodiment, i.e., the core quality estimation method, will be described. Fig. 5 is a flow diagram of the core quality estimation process according to this embodiment. This flow is not limited to this, but for example, in the core making apparatus 100, it starts at the timing when the mixed sand S is mixed in the mixing tank 110 each time a core making process is performed.

[0061] First, in step S11, the mold temperature information acquisition unit 301 acquires mold temperature information of the mold 160 into which the mixed sand S is to be filled. Next, in step S12, the environmental information acquisition unit 302 acquires environmental information related to the surrounding environment in which the core is molded. The environmental information acquisition unit 302 may acquire the environmental information each time an estimation process is performed, or may acquire the environmental information at regular intervals and use the same value for multiple estimation processes. The order of steps S11 and S12 may be changed as appropriate.

[0062] Next, in S13, the estimation unit 303 estimates the quality of the core to be formed based on the mold temperature information acquired by the mold temperature information acquisition unit 301 and the environmental information acquired by the environmental information acquisition unit 302. Here, the estimation by the estimation unit 303 is preferably performed before the filling of the mixed sand S into the mold 160 is completed in the molding process of the core to be estimated. By performing the estimation before the filling of the mixed sand S is completed, the mold temperature of the mold 160 can be controlled based on the estimation result, and defects occurring in the core can be prevented. Furthermore, the estimation by the estimation unit 303 is more preferably performed before the mixing of the mixed sand S in the kneading tank 110 is completed in the molding process of the core to be estimated. By performing the estimation before the mixing of the mixed sand S in the kneading tank 110 is completed, the amount of water input during mixing in the kneading tank 110 can be controlled, and defects occurring in the core can be prevented. Furthermore, if it is desired to estimate the quality of the core earlier than a specified timing in the core making process, this can be achieved by adjusting the timing for starting this flow to match the analysis time, or by degenerating the analysis software used for the behavior analysis described above.

[0063] Next, in step S14, the control unit 304 controls the molding conditions for molding the core based on the estimation result by the estimation unit 303, and this flow ends. Although omitted in this flow, a step may be added in which the display unit 305 displays the estimation result of the quality of the core estimated by the estimation unit 303 and the manufacturing conditions for preventing defects determined by the control unit 304. Also, here, an example has been described in which the estimation process is performed each time a core molding process is performed in the core molding apparatus 100, but the estimation process may be performed for each of multiple core molding processes, or may be performed in a virtual environment independent of the core molding process. In other words, the timing to start this flow and the timing of estimation by the estimation unit 303 can also be appropriately determined. For example, in the next molding process, the molding conditions may be controlled using the estimation result in the current molding process.

[0064] As explained above, the core quality estimation system in this embodiment comprises mold temperature information acquisition unit 301 which acquires mold temperature information of the mold, environmental information acquisition unit 302 which acquires environmental information related to the surrounding environment in which the core is formed, and estimation unit 303 which estimates the quality of the core based on the mold temperature information and the environmental information. Therefore, the quality of the core to be formed is estimated taking into account the environmental information related to the surrounding environment in which the core is formed, and the quality of the core can be accurately estimated taking into account changes in the moisture content of the mixed sand.

[0065] <Second embodiment> Next, a core quality estimating system 200 according to a second embodiment will be described. The hardware configuration of the core quality estimating system 200 according to the second embodiment is similar to that of the core quality estimating system 200 according to the first embodiment, so a description thereof will be omitted here. In addition, the functional configuration also includes common parts, so only the differences will be described here.

[0066] Fig. 6 is a block diagram showing the functional configuration of core quality estimating system 200 according to this embodiment. As shown in Fig. 6, core quality estimating system 200 comprises, as its functional configuration, a mold temperature information acquisition unit 301, an environmental information acquisition unit 302, an estimation unit 303, a control unit 304, a display unit 305, a correction acceptance unit 306, and a model update unit 307.

[0067] The estimation unit 303 of the core quality estimating system 200 according to the second embodiment estimates the quality of the core to be molded, based on the mold temperature information acquired by the mold temperature information acquisition unit 301 and the environmental information acquired by the environmental information acquisition unit 302, using a trained model that has been machine-learned using teacher data that takes as input mold temperature information and environmental information and outputs quality information related to the quality of the core. Here, the trained model is trained in advance in the core quality estimation system 200 or an external device such as an edge or server (cloud) using teacher data that takes as input mold temperature information and environmental information and outputs quality information related to the quality of the core, and is stored in the estimation unit 303, but the specific learning method will be described in detail later.

[0068] The correction receiving unit 306 receives corrections to the estimation result of the quality of the core by the estimation unit 303. Specifically, when the quality of the molded core is confirmed by a means other than the estimation by the estimation unit 303, such as an operator or an inspection device inspecting the actually molded core, if the estimation result by the estimation unit 303 differs from the confirmation result by the other means, the correction by the other means is accepted. Here, even if the results match, information that no correction is required may be accepted. In addition to the above, the confirmation by the other means may use mold temperature information obtained from the temperature sensor 162 when the mixed sand S is actually filled into the mold 160.

[0069] When the correction receiving unit 306 receives a correction, the model updating unit 307 updates the trained model stored in the estimation unit 303 based on the corrected content. Specifically, the trained model that has already been trained is re-trained by a learning method described below using a training data set including the corrected content. With this configuration, the trained model stored in the estimation unit 303 can be trained appropriately, and the quality of the core can be estimated with greater accuracy.

[0070] Next, the machine learning device 400 according to the second embodiment will be described. Here, detailed description of the hardware configuration of the machine learning device 400 will be omitted, but like the core quality estimation system 200, the machine learning device 400 has computer resources that a general information processing device has. In addition, here, an example will be described in which each unit of the machine learning device 400 is realized by an information processing device independent of the core quality estimation system 200, but it may be realized as a configuration incorporated inside the core quality estimation system 200, and some or all of these functions may be realized in an external device such as an edge or a server (cloud). Specifically, in a factory or the like in which a large number of core manufacturing devices 100 exist, each core manufacturing device 100 and each core quality estimation system 200 existing in each cell in the factory is connected to a fog server via a network, and the fog server provided for each cell is connected to a cloud server via a network. In such a configuration, the machine learning device 400 may be provided on the fog server or on the cloud server. In this way, by providing the machine learning machine 400 on the server, it is possible to collect information from multiple core molding devices and the like via the network and learn from it.

[0071] Next, a description will be given of the functional configuration of the machine learning device 400 according to this embodiment. Fig. 7 is a block diagram showing the functional configuration of the machine learning device 400 according to this embodiment.

[0072] The machine learning device 400 includes a mold temperature information acquisition unit 301, an environmental information acquisition unit 302, a quality information acquisition unit 401, and a learning unit 402. As described above, the mold temperature information acquisition unit 301 and the environmental information acquisition unit 302 respectively acquire mold temperature information of the mold 160 into which the mixed sand S is filled, and environmental information on the surrounding environment in which the core is molded.

[0073] The quality information acquisition unit 401 acquires quality information on the quality of the molded core. Specifically, the quality information on the quality of the core molded under these conditions is acquired in association with the mold temperature information acquired by the mold temperature information acquisition unit 301 and the environmental information acquired by the environmental information acquisition unit 302. Here, in this embodiment, an example will be described in which mold temperature information of the mold 160 measured when filling the mixed sand S or when heating is used as the quality information when the core is actually molded, but similarly, measured behavior information when filling the mixed sand S may be used as the quality information, and an operator may inspect the actually molded core, and the quality of the core input as inspection result information may be used as the quality information.

[0074] The learning unit 402 receives as input the mold temperature information acquired by the mold temperature information acquisition unit 301 and the environmental information acquired by the environmental information acquisition unit 302, creates and stores teacher data that outputs quality information associated with this information, and learns the quality of the core to be molded based on the teacher data, performing so-called supervised learning. Note that supervised learning refers to learning teacher data, that is, a learning data set of an input and an output for that input, to generate a learned model that estimates the output from the input, and various methods used in supervised learning can be used in the present invention.

[0075] Here, using a neural network as an example, we will explain the learning phase in which the machine learning device 400 learns the quality of the cores to be produced, and the operation phase in which the core quality estimation system 200 uses the learned model to estimate the quality of the cores to be produced.

[0076] First, in the learning phase, a learning dataset is created. For example, mold temperature information at multiple locations in the mold 160 acquired by the mold temperature information acquisition unit 301 and temperature information, humidity information, and atmospheric pressure information of the surrounding environment acquired by the environment information acquisition unit 302 are used as input values, and mold temperature information of the mold 160 when the mixed sand S is filled into the mold 160 when a core is actually molded under these conditions is used as an output value, and one learning dataset is created and stored. By preparing a large amount of data that indicates the relationship between these inputs and outputs, a learning dataset to be used in machine learning is created.

[0077] Next, the neural network used for learning will be described with reference to Fig. 8. Fig. 8 shows a "multiple input, single output" hierarchical neural network. Here, for simplicity, the hidden layer is described as having two layers, but in reality, it goes without saying that the network will have more layers, and the number of nodes in the input layer and hidden layer can be changed as desired depending on the learning data set.

[0078] By inputting input values ​​to each node of the input layer of such a neural network using the above-mentioned learning data set, the mold temperature information of the mold 160 when the mixed sand S is filled into the mold 160 is output as an estimated value from the output layer. Then, learning is performed so that the estimated value matches the mold temperature information of the mold 160 when the mixed sand S is filled into the mold 160 when the core is actually molded, which is prepared as a learning data set. Specifically, learning is performed using backpropagation or the like until the error of these values ​​converges to a predetermined set error or less, and the weights and biases (hereinafter collectively referred to as weights) of the neural network are learned. In this way, a neural network (trained model) is generated that outputs mold temperature information of the mold 160 when the mixed sand S is filled into the mold 160 in response to input of the mold temperature information acquired by the mold temperature information acquisition unit 301 and the environmental information acquired by the environmental information acquisition unit 302.

[0079] The neural network whose weights have been learned in this way is output to the estimation unit 303 of the core quality estimation system 200, and the operation phase is executed. In the operation phase, as shown in Fig. 8, the input values ​​of mold temperature information acquired by the mold temperature information acquisition unit 301 and environmental information acquired by the environmental information acquisition unit 302 are input to the input layer of the trained neural network, and mold temperature information of the mold 160 when the mixed sand S is filled into the mold 160 is output as an estimated value from the output layer. The estimation unit 303 compares the mold temperature information of the mold 160 output by the neural network with a preset specified range, and estimates the quality of the core. It is also possible to generate a neural network that outputs behavior information when the mixed sand S is filled, and a neural network that outputs the quality of the core, using a similar method. Furthermore, in this embodiment, an example has been described in which mold temperature information and environmental information are used as input values, but in addition to this, various molding conditions such as the raw material type of the mixed sand S, raw material ratios such as moisture content, kinetic viscosity of the mixed sand S, and injection pressure can be used as input values.

[0080] Next, the flow of the learning phase in the machine learning device 400 according to this embodiment, that is, the learning method, will be described. FIG. 9 is a flow diagram of the learning process according to this embodiment. This flow starts, for example, in a state where a predetermined amount or more of learning data sets have been accumulated in the machine learning device 400. Also, when a new learning data set of a predetermined amount or more has been accumulated since the previous learning, this learning process may be executed to perform re-learning.

[0081] First, in step S21, the learning unit 402 reads each data of the learning data set stored in the machine learning device 400. Next, in step S22, the learning unit 402 reads the number of nodes in each of the input layer, hidden layer, and output layer of the neural network, and creates a neural network.

[0082] In step S23, the learning unit 402 learns the weights of the neural network using the loaded learning data set. Specifically, the learning unit 402 inputs the input values ​​of the learning data set to the input layer, and executes learning of the weights of the neural network using backpropagation so that the error between the output values ​​of the neural network to be output and the output values ​​of the learning data set becomes small. After learning has been sequentially performed for all data, the process proceeds to step S24.

[0083] In step S24, the learning unit 402 determines whether the error between the output value in the neural network and the output value of the learning data set has converged to a predetermined set error or less, and if it is determined that the error has converged to the set error or less, the process proceeds to step S25, where the learned weights of the neural network are stored. On the other hand, if it is determined that the error has not converged to the set error or less, the learning unit 402 re-learns the weights in step S23, and continues learning until the error between the output value in the neural network and the output value of the learning data set has converged to the set error or less.

[0084] Next, the flow of the operation phase in core quality estimating system 200 according to this embodiment, i.e., the core quality estimating method, will be explained. Figure 10 is a flow diagram of core quality estimating processing according to this embodiment, but as most of it is common to the flow of core quality estimating processing according to the first embodiment, only the differences will be explained here. This flow starts when the learning process described above has been completed and the learned model has been stored in estimation unit 303.

[0085] First, in steps S31 and S32, similarly, mold temperature information and environmental information are acquired, and then in step S33, the estimation unit 303 estimates the quality of the core to be molded using the trained model for which the above-mentioned learning process has been completed, based on the mold temperature information and the environmental information. Here, as in the first embodiment, the estimation by the estimation unit 303 is preferably performed in the molding process of the core to be estimated by the time the filling of the mixed sand S into the mold 160 is completed, and more preferably by the time the mixing of the mixed sand S in the mixing tank 110 is completed. In this way, when it is desired to estimate the quality of the core earlier than a predetermined timing in the core molding process, this can be achieved by selecting a simple model structure for the trained model as described above, or by reducing the information used as input.

[0086] Next, in S34, correction receiving unit 306 receives corrections to the results of estimation of the quality of the core by estimation unit 303. Next, in S35, if the correction receiving unit 306 receives a correction, the model updating unit 307 updates the trained model stored in the estimation unit 303 based on the corrected content, and this flow ends. Although omitted in this flow, as explained in the flow of the core quality estimation process according to the first embodiment, a step of controlling the molding conditions by the control unit 304 and a step of displaying the estimation results of the core quality by the display unit 305 may be added.

[0087] In the core quality estimation system of this embodiment in particular, the estimation unit 303 estimates the quality of the core based on the mold temperature information and environmental information, using a trained model trained through machine learning using teacher data that inputs mold temperature information and environmental information and outputs quality information related to the quality of the core.Therefore, the quality of the core to be molded is estimated using a trained model that has previously learned the relationship between the mold temperature information, environmental information, and the quality information of the core, thereby achieving the effect of being able to accurately estimate the quality of the core taking into account changes in the moisture content of the mixed sand.

[0088] <Other embodiments> In the first and second embodiments, an example has been described in which a core quality estimation system 200 and a machine learning machine 400 are provided for each core making apparatus 100, but these may also be shared among a plurality of core making apparatuses 100 and configured to estimate and learn the quality of cores made by the plurality of core making apparatuses 100. Also, in the first and second embodiments, an example has been described in which the estimation unit 303 estimates the quality of a core using an analysis of the behavior of the mixed sand S in the mold 160 using CAE, or an example in which the estimation unit 303 estimates the quality of a core using a trained model trained by machine learning using teacher data, but the estimation method is not limited to these, and various methods such as estimation using a formula calculated in advance and multivariate analysis can be used.

[0089] The present invention is not limited to the above-described embodiment, and can be embodied in various forms by making appropriate modifications without departing from the spirit of the present invention. [Explanation of symbols]

[0090] 100 core molding device, 110 kneading tank, 111 through hole, 120 raw material supply means, 121 core sand supply means, 122 water glass supply means, 123 water supply means, 124 surfactant supply means, 130 pedestal, 140 piston, 150 cylinder, 160 mold, 161 mold heat controller, 162 temperature sensor, 163 left mold, 164 right mold, S kneaded sand, 200 core quality estimation system, 201 CPU, 202 ROM, 203 RAM, 204 HDD, 205 communication I / F, 206 input / output I / F, 207 data bus, 301 mold temperature information acquisition unit, 302 environmental information acquisition unit, 303 estimation unit, 304 control unit, 305 display unit, 306 correction reception unit, 307 model update unit, 400 Machine learning machine, 401 quality information acquisition unit, 402 learning unit

Claims

1. A core quality estimation system for estimating the quality of a core obtained through a process of moving a kneading tank containing kneaded sand up to a mold, filling the mold with the kneaded sand from the moved kneading tank, and shaping by heating, comprising: a mold temperature information acquisition unit that acquires mold temperature information from a temperature sensor provided in the mold; an environmental information acquisition unit that acquires environmental information regarding the surrounding environment in which the core is shaped; an estimation unit that estimates the quality of the core based on the mold temperature information of the mold during filling or heating of the kneaded sand and the behavior information during filling of the kneaded sand, which are obtained by performing an analysis of the behavior of the kneaded sand in the mold using CAE based on the mold temperature information and the environmental information; a control unit that controls the water input amount, core sand input amount, surfactant input amount, kneading time included in the shaping conditions during kneading of the kneaded sand, and the mold temperature of the mold included in the shaping conditions when filling the mold with the kneaded sand, based on the estimation result by the estimation unit; A core quality estimation system comprising the above.

2. A core quality estimation system for estimating the quality of a core obtained through a process of moving a kneading tank containing kneaded sand up to a mold, filling the mold with the kneaded sand from the moved kneading tank, and shaping by heating, comprising: a mold temperature information acquisition unit that acquires mold temperature information of the mold; an environmental information acquisition unit that acquires environmental information regarding the surrounding environment in which the core is shaped; an estimation unit that estimates the quality of the core based on the mold temperature information of the mold during filling or heating of the kneaded sand and the behavior information during filling of the kneaded sand, which are obtained based on the mold temperature information and the environmental information, using a learned model obtained by machine learning with teacher data that takes the mold temperature information and the environmental information as input and outputs quality information regarding the quality of the core; a control unit that controls the water input amount, core sand input amount, surfactant input amount, kneading time included in the shaping conditions during kneading of the kneaded sand, and the mold temperature of the mold included in the shaping conditions when filling the mold with the kneaded sand, based on the estimation result by the estimation unit; A core quality estimation system comprising the above.

3. a correction reception unit that receives corrections to the estimation result by the estimation unit; and a model update unit that updates the learned model based on the corrected content when the correction reception unit receives a correction; further comprising the core quality estimation system according to Claim 2.

4. ​ The environmental information includes at least one of humidity information and atmospheric pressure information of the surrounding environment. The core quality estimation system according to any one of claims 1 to 3.

5. The environmental information acquisition unit acquires the environmental information based on the position information where the core is shaped. The core quality estimation system according to any one of claims 1 to 4.

6. A core quality estimation method for estimating, by a computer, the quality of a core obtained through a process of moving a kneading tank containing kneaded sand to above a mold, filling the mold with the kneaded sand from the moved kneading tank, and shaping by heating, comprising: a mold temperature information acquisition step of acquiring mold temperature information of the mold from a temperature sensor provided in the mold; an environmental information acquisition step of acquiring environmental information regarding the surrounding environment where the core is shaped; an estimation step of estimating the quality of the core based on the mold temperature information of the mold during filling or heating of the kneaded sand and the behavior information during filling of the kneaded sand, which are obtained by performing an analysis of the behavior of the kneaded sand in the mold using CAE based on the mold temperature information and the environmental information; a control step of controlling the water input amount, core sand input amount, surfactant input amount, kneading time included in the shaping conditions during kneading of the kneaded sand, and the mold temperature of the mold included in the shaping conditions when filling the mold with the kneaded sand, based on the estimation result of the estimation step; A core quality estimation method comprising the above steps.

7. A program for executing a core quality estimation method for estimating the quality of a core obtained through a process of moving a kneading tank containing kneaded sand to above a mold, filling the mold with the kneaded sand from the moved kneading tank, and shaping by heating, comprising: a mold temperature information acquisition step of acquiring mold temperature information of the mold from a temperature sensor provided in the mold; an environmental information acquisition step of acquiring environmental information regarding the surrounding environment where the core is shaped; an estimation step of estimating the quality of the core based on the mold temperature information of the mold during filling or heating of the kneaded sand and the behavior information during filling of the kneaded sand, which are obtained by performing an analysis of the behavior of the kneaded sand in the mold using CAE based on the mold temperature information and the environmental information; Based on the estimation result of the estimation step, a control step of controlling the water input amount, the core sand input amount, the surfactant input amount, the kneading time included in the molding conditions during the kneading of the kneaded sand, and the mold temperature of the mold included in the molding conditions when filling the mold with the kneaded sand; A program that causes a computer to execute.

Images