Data Learning Based Sand Casting System
The data learning-based sand casting system addresses inconsistent quality in conventional sand casting by automatically adjusting the mixing ratio using a neural network, enhancing product consistency and environmental adaptability.
Patent Information
- Application Number
- KR1020250028584
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2025-03-06
- Publication Date
- 2026-11-11
- Estimated Expiration
- 2045-03-06
AI Technical Summary
Conventional self-hardening sand casting relies on manual adjustment of the mixing ratio of molding sand, resin, and hardener based on worker experience, leading to inconsistent casting quality due to environmental variations.
A data learning-based sand casting system that uses a mixer device, environment measuring means, and a controller to automatically adjust the mixing ratio based on data learning, employing a ConvLSTM neural network to optimize the supply amounts of molding sand, resin, and hardener.
Improves casting quality by reducing dependence on worker skill and enabling flexible responses to environmental changes, ensuring consistent product quality.
Smart Images

Figure 112025025184911-PAT00001_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to a data learning-based sand casting system, and more specifically, to a data learning-based sand casting system capable of automatically adjusting the mixing ratio of molding sand, resin, and hardener added for bonding the molding sand to an optimal level based on data learning. Background Technology
[0003] Generally, self-hardening sand casting is a casting method that uses a mold that is cured at room temperature by adding a binder (hereinafter referred to as 'resin'), which is a type of sand casting, with furfuryl alcohol as the main ingredient, and an acidic curing agent to the molding sand. The binder serves to bind the molding sand, and the curing agent promotes bonding to increase the strength of the mold.
[0004] For molding sand, heat resistance, dimensional stability, and recyclability are important; therefore, it must possess refractoriness to prevent sticking even when in contact with high-temperature molten metal, minimize expansion and contraction during cooling after being heated in contact with the molten metal, and allow for repeated use without the sand particles easily breaking down.
[0005] In the case of such molding sand, the mixing ratio of binder and hardener is closely related to the quality of the casting. Since environmental factors such as the temperature and humidity of the molding sand also act as determinants of quality, the mixing ratio of the binder and hardener depends significantly on the molding sand environment, specifically its temperature and humidity, in order to produce uniform and high-quality castings.
[0006] However, conventionally, the mixing ratio of molding sand, resin, and hardener was determined based on the experience and know-how of skilled workers. Since skilled workers independently assessed the environment of the molding sand to determine the mixing ratio of the resin and hardener and manually controlled the mixing process, the quality of the castings was uneven, and there was a problem where the quality varied depending on changes in the environment.
[0007] Casting technology is one of the six major root technologies and is utilized throughout the manufacturing industry. As it is an important technology that forms the foundation of the industry and is actively used in the manufacturing processes of upstream industries such as automobiles and shipbuilding, it is necessary to conduct research to solve the problems of conventional self-hardening sand casting as described above. The problem to be solved
[0009] The present invention has been devised to solve the aforementioned problems, and the objective of the present invention is to provide a data learning-based sand casting system that can automatically adjust the mixing ratio of the molding sand, the resin and hardener added for bonding the molding sand to the optimal level based on data learning, thereby improving the quality of sand products that depend on the skill level of workers and flexibly responding to changing environmental conditions. means of solving the problem
[0011] To achieve the above objective, the present invention comprises: a mixer device that receives molding sand, resin, and a hardener, mixes the resin and the hardener with the molding sand, and discharges the mixture into a mold; one or more molding sand environment measuring means that measure the molding sand environment, including the temperature and humidity, in advance before forming a casting using the molding sand; and a controller that controls the mixer device, generates supply amount information while adjusting the supply amounts of the molding sand, the resin, and the hardener to the mixer device, collects the molding sand environment information measured by the molding sand environment measuring means, and receives casting quality information of the casting formed using the molding sand. The present invention provides a data learning-based sand casting system comprising: a data server in which a database and a learning module are established, receiving information on the molding sand supply amount, the resin supply amount, the hardener supply amount, the molding sand environment information, and the casting quality information for the mixer device from the controller, storing it in the database, and learning conditions for forming optimal casting quality using the learning module based on the information stored in the database; wherein the controller adjusts one or more of the supply amounts of the molding sand, the resin, and the hardener for the mixer device according to the conditions learned from the data server.
[0012] In a preferred embodiment, the learning module is characterized by using a ConvLSTM (Convolutional LSTM) neural network learning model in which a Long Short-Term Memory (LSTM) neural network and a Multilayer Perceptron (MLP) neural network are configured between the Input Layer and the Output Layer.
[0013] In addition, the data learning-based sand casting system is provided with a sand receiving space for receiving sand supplied from a sand supply unit and a sand receiving space for receiving recycled sand supplied from a recycled sand supply unit, which is separated from the sand receiving space, on the upper side of the mixer device, and a sand supply hopper that receives recycled sand supplied from a recycled sand supply unit and forms a recycled sand receiving space, and the mixer device is characterized by supplying sand or recycled sand discharged alone from the sand supply hopper, or a mixed sand of sand and recycled sand discharged together, as the sand or recycled sand discharged together, to the mixer device as the sand or recycled sand discharged alone from the sand supply hopper, under the control of the controller.
[0014] Additionally, the recycled sand supply unit is connected to a recycled sand sample measuring device capable of measuring information to determine whether the recycled sand supplied to the molding sand supply hopper meets usage standards, and the recycled sand sample measuring device comprises: a recycled sand sample receiving means for receiving a recycled sand sample; a heating means for heating the recycled sand sample receiving means; and a weight measuring device that measures the weight of the recycled sand sample receiving means before heating by the heating means and the weight of the recycled sand sample receiving means after heating by the heating means and transmits this to the controller. The controller receives the information measured by the weight measuring device, compares the weight of the recycled sand sample receiving means before and after heating to determine the amount of weight loss of the recycled sand sample, and controls the supply of the recycled sand to the molding sand supply hopper only when the amount of weight loss of the recycled sand sample is below the standard value.
[0015] In addition, the molding sand environment measuring means comprises one or more molding sand environment measuring means for measuring the molding sand environment, including the temperature and humidity of the molding sand, in advance before forming the molding sand, and one or more recycled sand environment measuring means for measuring the recycled sand environment, including the temperature and humidity of the recycled sand, in advance before forming the molding sand; the controller collects molding sand environment information, including molding sand state information measured by the molding sand state measuring means and recycled sand state information measured by the recycled sand state measuring means, and further collects information on the mixing ratio of the molding sand and recycled sand of the mixed sand; the data server receives the molding sand environment information, the recycled sand environment information, and the mixing ratio of the molding sand and recycled sand of the mixed sand from the controller and stores them in the database; further utilizes the molding quality information, the molding sand supply amount information, the molding sand environment information, the recycled sand environment information, and the mixing ratio of the molding sand and recycled sand of the mixed sand to learn conditions for forming optimal molding quality; and the controller adjusts the mixing amount of the molding sand and the recycled sand for the molding sand according to the conditions learned by the data server. Effects of the invention
[0017] The data learning-based sand casting system of the present invention can automatically adjust the mixing ratio of the molding sand, resin, and hardener added for bonding the molding sand to an optimal level based on data learning, thereby improving the quality of sand products that depend on the skill level of workers and enabling flexible response to changing environmental conditions. Brief explanation of the drawing
[0019] Figure 1 is a structural diagram of a data learning-based sand casting system according to an embodiment of the present invention. Figure 2 is a diagram illustrating a learning model of a learning module of the data server presented in Figure 1. Figure 3 is a diagram showing the configuration of the controller and data server presented in Figure 1. Specific details for implementing the invention
[0020] The terms used in this invention have been selected to be as widely used as possible; however, in specific cases, terms have been arbitrarily selected by the applicant. In such cases, the meaning should be understood by considering the meaning described or used in the detailed description of the invention, rather than merely the name of the term.
[0021] Hereinafter, the technical configuration of the present invention will be described in detail with reference to preferred embodiments illustrated in the attached drawings.
[0022] However, the present invention is not limited to the embodiments described herein and may be embodied in other forms. Throughout the specification, the same reference numerals indicate the same components.
[0023] FIG. 1 is a structural diagram of a data learning-based sand casting system according to an embodiment of the present invention, FIG. 2 is a diagram illustrating a learning model of a learning module of a data server presented in FIG. 1, and FIG. 3 is a diagram showing the configuration of a controller and a data server presented in FIG. 1.
[0024] Referring to FIGS. 1 to 3, a data learning-based sand casting system according to an embodiment of the present invention comprises a mixer device (200), a sand environment measuring means (not shown in the drawings), a controller (300), and a data server (400).
[0025] In detail, the mixer device (200) is configured to receive molding sand, resin, and a hardener. Here, the molding sand is supplied from a molding sand supply hopper (100) located above the mixer device (200), the resin is supplied from a resin tank (150) in which the resin is stored, and the hardener is supplied from a hardener tank (160) in which the hardener is stored.
[0026] In the case of the aforementioned molding sand, only new molding sand (i.e., newly used molding sand) may be used; however, considering economic conditions, it is preferable to mix recycled sand (i.e., molding sand that has already been used) with the new molding sand.
[0027] Accordingly, in the present invention, the molding sand supply hopper (100) that supplies molding sand to the mixer device (200) as a preferred form is not exemplified as a single space that accommodates only sand, but is exemplified as a plurality of spaces capable of forming sand and molding sand, respectively.
[0028] However, as explained above, since only a gentleman can be used, it should be understood that, unlike the drawing, the casting sand supply hopper (100) can be formed as a single space.
[0029] A more detailed explanation regarding the case where two types of molding sand, new sand and recycled sand, are supplied to the molding sand supply hopper (100) will be provided later.
[0030] Meanwhile, the resin stored in the resin tank (150) can be transferred through a resin transfer pipe (152) connected from the resin tank (150) to a mixer device (200), and the hardener stored in the hardener tank (160) can be transferred through a hardener transfer pipe (162) connected from the hardener tank (160) to a mixer device (200).
[0031] Here, the molding sand supply hopper (100) is provided with an opening / closing door (130) that opens and closes the molding sand supply hopper (100), and the amount of molding sand supplied to the mixer device (200) is controlled according to the degree of opening of the opening / closing door (130), and the resin transfer pipe (152) and the hardener transfer pipe (162) are each provided with valves (154, 164), and the amount of resin and hardener supplied to the mixer device (200) can be controlled according to the degree of opening of the valves (154, 164).
[0032] The above-mentioned opening / closing door (130) and valves (154, 164) are all controlled by the controller (300) to be described later.
[0033] The above mixer device (200) is equipped with a mixing screw (210) inside to mix and transport the supplied molding sand, resin, and hardener, and to discharge them into the mold (10). Here, various types of molding sand such as silica sand, ceramic sand, and zircon sand can be used as the type of molding sand, the resin is a binder with furfuryl alcohol as the main ingredient, and the hardener is an acidic hardener.
[0034] By mixing the above-mentioned molding sand, resin, and hardener, and then fabricating and curing the mold, it is possible to cure at room temperature without the injection of gases such as CO2 or a separate drying procedure.
[0035] The molding sand environment measuring means is a means for measuring the molding sand environment, including the temperature and humidity of the molding sand, in advance before forming a casting using the molding sand. Here, the molding sand environment measuring means may be a temperature sensor capable of measuring temperature or a humidity sensor capable of measuring humidity, and a plurality of means (112, 114, 122, 124, 222, 224) may be provided at the beginning of the molding sand supply hopper (100) or the mixer device (200).
[0036] The controller (300) controls the mixer device (200) and can control the opening / closing door (130) of the molding sand supply hopper (100) described above, the valve (154, 164) of the resin transfer pipe (152) or the hardener transfer pipe (162) to control the supply amount of molding sand, resin, and hardener to the mixer device (200).
[0037] Here, the controller (300) generates supply amount information for each material while controlling the supply amounts of molding sand, resin, and hardener. That is, it generates molding sand supply amount information while controlling the molding sand supply amount, generates resin supply amount information while controlling the resin supply amount, and generates hardener supply amount information while controlling the hardener supply amount.
[0038] Additionally, the controller (300) is connected to a molding sand environment measuring means and collects molding sand environment information regarding the molding sand environment measured by the molding sand environment measuring means.
[0039] And, the controller (300) can receive information on the quality of the casting formed using molding sand from the operator.
[0040] The data server (400) is configured with a database (440) and a learning module (440).
[0041] The data server (400) receives information generated, collected, or input from the controller (300) and stores it in the database (440) to prepare big data for learning. That is, it receives information on the molding sand supply amount, resin supply amount, hardener supply amount, molding sand environment information, and molding quality information regarding the mixer device (200) from the controller (300) and stores it in the database (440).
[0042] In addition, the data server (400) performs learning using a learning module (440) built using big data stored in the database (440). That is, it performs learning using information such as molding sand supply amount, resin supply amount, hardener supply amount, molding sand environment information, and casting quality information stored in the database (440).
[0043] Here, the learning module (440) learns what conditions can form the optimal casting quality by comparing the casting quality information according to the temperature or humidity of the casting sand, identifying the appropriate amount of casting sand, resin, or hardener, etc.
[0044] The learned condition information is transmitted back to the controller (300), and the controller (300) performs control according to the learned conditions from the data server (400), and adjusts one or more of the supply amounts of molding sand, resin, and hardener so as to produce optimal casting quality relative to the environment, such as the temperature or humidity of the molding sand.
[0045] Meanwhile, as shown in FIG. 2, the learning module may use a ConvLSTM (Convolutional LSTM) neural network learning model in which a Long Short-Term Memory (LSTM) neural network and a Multilayer Perceptron (MLP) neural network are configured between the Input Layer and the Output Layer.
[0046] However, this is merely an exemplary learning model presented to aid understanding, and implementation is not necessarily limited to the learning model above.
[0047] Meanwhile, as described above, the present invention may also use a mixture of new sand and recycled sand for casting.
[0048] To this end, a sand receiving space (110) and a recycled sand receiving space (120) separated from the sand receiving space are formed in the casting sand supply hopper (100).
[0049] The above-mentioned shrine receiving space (110) receives a shrine supplied from a shrine supply unit (not shown in the drawing) and receives a shrine supplied from a recycled shrine supply unit (not shown in the drawing).
[0050] The above-mentioned sand supply unit is provided with a sand storage tank (170) for storing sand, a sand transfer pipe (172) for transferring sand stored in the sand storage tank (170) to the sand receiving space (110), and a valve (174) for controlling the supply of sand to the molding sand supply hopper (100) in the sand transfer pipe (172).
[0051] Likewise, the recycled sand supply unit is provided with a recycled sand storage tank (180) for storing recycled sand, a recycled sand transfer pipe (182) for transferring the recycled sand stored in the recycled sand storage tank (180) to the recycled sand receiving space (120), and a valve (184) for controlling the supply of recycled sand to the molding sand supply hopper (100) on the recycled sand transfer pipe (182).
[0052] Here, the valves (174, 184) provided in the above-mentioned new material transfer pipe (172) and the above-mentioned new material transfer pipe (182), respectively, are controlled by a controller (300).
[0053] The sand delivered from the sand supply unit to the sand receiving space (110) and the recycled sand delivered from the recycled sand supply unit to the recycled sand receiving space (120) are separated from each other within the molding sand supply hopper (100), and then, by opening the opening / closing door (130) of the controller (300), only the sand is discharged, only the recycled sand is discharged, or the recycled sand and sand are discharged together and mixed.
[0054] That is, the molding sand consists of the above-mentioned sand, the above-mentioned recycled sand, or a mixture of the above-mentioned sand and recycled sand.
[0055] Here, the opening and closing door (130) of the casting sand supply hopper (100) is provided for each receiving space (110, 120) so that the discharge of new sand and recycled sand can be controlled separately.
[0056] Meanwhile, a recycled sand sample measuring device (500) is connected to the recycled sand supply unit to measure information that can determine whether the recycled sand supplied to the molding sand supply hopper (100) meets the usage standard.
[0057] The above-described regenerated yarn sample measuring device (500) includes a regenerated yarn sample receiving means (510) for receiving a regenerated yarn sample, a heating means (530) for heating the regenerated yarn sample receiving means (510), and a weight measuring device (520) that measures the weight of the regenerated yarn sample receiving means (510) before heating by the heating means (530) and the weight of the regenerated yarn sample receiving means (510) after heating by the heating means, and transmits the results to the controller (300).
[0058] A controller (300) that receives information measured from the above-mentioned weight measuring device (520) can determine the amount of weight loss of the recycled yarn sample by comparing the weight before and after heating of the recycled yarn sample receiving means (510).
[0059] For example, if the weight of the recycled yarn sample receiving means (510) is 130 kg before heating and the weight of the recycled yarn sample receiving means (510) is 100 kg after heating, the weight of the recycled yarn sample receiving means (510) does not change, so it can be determined that the weight loss of the recycled yarn sample is 30 kg.
[0060] Here, the weight loss amount is an indicator that can determine the degree of contamination of the recycled yarn. If the weight loss amount is large, it means that the specific gravity of the resin or hardener attached to the recycled yarn is large, and if the weight loss amount is small, it means that the specific gravity of the resin or hardener attached to the recycled yarn is small.
[0061] For example, if we compare a weight loss of 30 kg with a weight loss of 10 kg, a weight loss of 30 kg means that all 30 kg of resin and hardener have evaporated and only recycled sand remains, and a weight loss of 10 kg means that all 10 kg of resin and hardener have evaporated and only recycled sand remains.
[0062] Based on this, the controller (300) determines that the use of recycled sand is suitable only when the weight loss of the recycled sand sample is below a reference value, and opens the valve (184) of the recycled sand transfer pipe (182), thereby allowing the recycled sand to be supplied to the molding sand supply hopper (100).
[0063] Meanwhile, in the present invention, when mixing new sand and recycled sand to use as molding sand, the mixing ratio of new sand and recycled sand that can form optimal casting quality is also learned, and the controller (300) can adjust the amount of recycled sand mixed according to the learned condition.
[0064] Specifically, the molding sand condition measuring means includes one or more molding sand condition measuring means for measuring the molding sand condition, including the temperature and humidity of the molding sand, in advance before forming the casting, and one or more recycled sand condition measuring means for measuring the recycled sand condition, including the temperature and humidity of the recycled sand, in advance before forming the casting.
[0065] Here, the means for measuring the sand condition may be, for example, a temperature sensor (112) and a humidity sensor (114) provided in the sand receiving space (110) of the molding sand supply hopper (100), and the means for measuring the recycled sand condition may be a temperature sensor (122) and a humidity sensor (124) provided in the recycled sand receiving space (120) of the molding sand supply hopper (100).
[0066] And, the controller (300) is configured to collect, as the molding sand environment information, the molding sand state information measured by the molding sand state measuring means and the regenerated sand state information measured by the regenerated sand state measuring means, as well as further collect the molding sand and regenerated sand mixing ratio information of the mixed sand.
[0067] And, the data server (400) receives the molding state information, the recycled sand state information, and the mixing ratio information of the molding sand and recycled sand from the controller (300), stores them in the database, and further learns the conditions for forming optimal casting quality by utilizing the casting quality information, the casting sand supply amount information, the molding state information, the recycled sand state information, and the mixing ratio information of the molding sand and recycled sand.
[0068] And, again, the controller (300) adjusts the amount of the new sand and the recycled sand mixed with the molding sand according to the conditions learned from the data server.
[0069] As illustrated in FIG. 3, the controller (300) is equipped with a molding sand supply control means (310) that generates molding sand supply amount information while controlling the molding sand supply, a resin supply control means (320) that generates resin supply amount information while controlling the resin supply, and a curing agent supply control means (330) that generates curing agent supply amount information while controlling the curing agent supply.
[0070] In addition, a temperature and humidity information collection means (340) for collecting measurement information regarding the molding sand environment, including temperature and humidity information, from temperature sensors (112, 122, 222) and humidity sensors (114, 124, 224) provided in a molding sand supply hopper (100) or a mixer device (200), and a molding quality information input means (350) for receiving molding quality information are provided.
[0071] In addition, when the molding sand is to be used together with the new sand and the recycled sand, a recycled sand supply control means (360) that generates information on the amount of recycled sand supplied while controlling the supply of recycled sand, and a sand supply control means (370) that generates information on the amount of sand supplied while controlling the supply of new sand are provided.
[0072] A data server (400) connected to the above controller (300) and exchanging various information and learned condition data, etc., is equipped with a database (440) that stores data received from the controller (300) and converts it into big data, and a learning module (440) that performs learning based on the information stored in the database (440).
[0073] In addition, a means for determining the amount of molding sand supplied according to the learning performed in the learning module (440) and a means for determining the amount of resin supplied according to the learning performed in the learning module (440) and a means for determining the amount of curing agent supplied according to the learning performed in the learning module (440) and a means for determining the amount of recycled sand supplied according to the learning performed in the learning module (440) and a means for determining the mixing ratio of recycled sand to molding sand according to the learning performed in the learning module (440) are provided.
[0074] The molding sand supply amount information determined by the molding sand supply amount determining means (410) is transmitted to the molding sand supply control means (310) of the controller (300) so that the controller (300) controls according to the information, and the resin supply amount information determined by the resin supply amount determining means (420) is transmitted to the resin supply control means (320) of the controller (300) so that the controller (300) controls according to the information.
[0075] In addition, the curing agent supply amount information determined by the curing agent supply amount determining means (430) is transmitted to the curing agent supply control means (330) of the controller (300) so that the controller (300) controls according to the information, and the regenerated sand supply amount information determined by the regenerated sand supply amount determining means (460) is transmitted to the regenerated sand supply control means (360) and the sand supply control means (370) of the controller (300) so that the controller (300) controls according to the information.
[0076] The data learning-based sand casting system according to the embodiment of the present invention described above can automatically adjust the mixing ratio of the molding sand, resin, and hardener added for bonding the molding sand to an optimal level based on data learning, thereby improving the quality of sand products that depend on the skill level of workers and flexibly responding to changing environmental conditions.
[0078] As described above, the present invention has been illustrated and explained with reference to preferred embodiments, but it is not limited to the aforementioned embodiments, and various changes and modifications may be made by those skilled in the art within the scope of the invention without departing from the spirit of the invention. Explanation of the symbols
[0080] 10 : Mold 100: Molding sand supply hopper 110: Gentleman's reception area 112: Temperature sensor 114: Humidity sensor 120 : Regeneration reception space 122: Temperature sensor 124: Humidity sensor 130 : Open / close door 150 : Resin tank 152 : Resin transfer pipe 154 : Valve 160: Hardener tank 162: Curing agent transfer pipe 164: Valve 170 : Gentleman Storage Tank 172 : Shinsa transfer pipe 174 : Valve 180 : Regenerator storage tank 182: Recycled sand transfer pipe 184: Valve 200 : Mixer device 210: Mixing screw 222: Temperature sensor 224: Humidity sensor 300 : Controller 310: Molding sand supply control means 320: Resin supply control means 330: Means for controlling curing agent supply 340: Means for collecting temperature and humidity information 350: Means for inputting casting quality information 360: Means for controlling the supply of recycled sand 370: Gentleman supply control means 400: Data Server 410: Means for determining molding sand supply amount 420: Means for determining resin supply amount 430: Means for determining curing agent supply amount 440: Database 450: Learning Module 460: Means for Determining Recycled Yarn Supply Amount 500 : Recycled yarn sample measuring device 510: Means for receiving recycled yarn samples 520: Weighing device 530: Heating means
Claims
Claim 1 A mixer device that receives molding sand, resin, and a hardener, mixes the resin and the hardener with the molding sand, and discharges the mixture into a mold; one or more molding sand environment measuring means that measure the molding sand environment, including the temperature and humidity, in advance before forming a casting using the molding sand; and a controller that controls the mixer device, generates supply amount information while adjusting the supply amounts of the molding sand, the resin, and the hardener to the mixer device, collects the molding sand environment information measured by the molding sand environment measuring means, and receives casting quality information of the casting formed using the molding sand. A data learning-based sand casting system comprising: a data server in which a database and a learning module are established, receiving information on the molding sand supply amount, the resin supply amount, the hardener supply amount, the molding sand environment information, and the molding quality information for the mixer device from the controller, storing it in the database, and learning conditions for forming optimal molding quality using the learning module based on the information stored in the database; wherein the controller adjusts one or more of the supply amounts of the molding sand, the resin, and the hardener for the mixer device according to the conditions learned from the data server; and a molding sand supply hopper is provided on the upper side of the mixer device, which includes a sand receiving space for receiving sand supplied from a sand supply unit and a regenerated sand receiving space for receiving regenerated sand supplied from a regenerated sand supply unit, which is separated from the sand receiving space; and wherein the mixer device is characterized by supplying sand or regenerated sand discharged alone from the molding sand supply hopper, or a mixture of sand and regenerated sand discharged together, as the molding sand under the control of the controller. Claim 2 A data learning-based casting system according to claim 1, characterized in that the learning module uses a ConvLSTM (Convolutional LSTM) neural network learning model configured between an input layer and an output layer, comprising a Long short-term memory (LSTM) neural network and a Multilayer perceptron (MLP) neural network. Claim 3 delete Claim 4 In claim 1, the recycled sand supply unit is connected to a recycled sand sample measuring device capable of measuring information to determine whether the recycled sand supplied to the molding sand supply hopper meets usage standards, and the recycled sand sample measuring device comprises: a recycled sand sample receiving means for receiving a recycled sand sample; a heating means for heating the recycled sand sample receiving means; and a weight measuring device that measures the weight of the recycled sand sample receiving means before heating by the heating means and the weight of the recycled sand sample receiving means after heating by the heating means and transmits this to the controller; and the controller receives the information measured by the weight measuring device, compares the weight of the recycled sand sample receiving means before and after heating to determine the amount of weight loss of the recycled sand sample, and controls the recycled sand to be supplied to the molding sand supply hopper only when the amount of weight loss of the recycled sand sample is below the standard value, characterized in that it is a data learning-based sand casting system. Claim 5 In claim 1, the molding sand environment measuring means comprises one or more sand environment measuring means for pre-measuring the sand environment, including the temperature and humidity of the sand, before forming the casting, and one or more recycled sand environment measuring means for pre-measuring the recycled sand environment, including the temperature and humidity of the recycled sand, before forming the casting; the controller collects sand state information measured by the sand environment measuring means and recycled sand state information measured by the recycled sand environment measuring means as molding sand environment information, and further collects information on the mixing ratio of the sand and recycled sand of the mixed sand; the data server receives the sand environment information, the recycled sand environment information, and the mixing ratio of the sand and recycled sand of the mixed sand from the controller and stores them in the database; further utilizes the casting quality information, the molding sand supply amount information, the sand environment information, the recycled sand environment information, and the mixing ratio of the sand and recycled sand of the mixed sand to learn conditions capable of forming optimal casting quality; and the controller adjusts the mixing amount of the sand and recycled sand for the molding sand according to the conditions learned by the data server, characterized by a data learning-based sand casting System.
Citation Information
Patent Citations
A system and method for evaluation of sand compactibility
US20230221687A1
