Machine learning model-based method for manufacturing carbon nanomaterial and system for performing same

Abstract

The present invention relates to a machine learning model-based method for manufacturing a carbon nanomaterial. The method may comprise the steps of: acquiring first control information about a process of synthesizing a carbon nanomaterial; acquiring analysis information about the synthesized carbon nanomaterial in real time on the basis of the first control information; managing the first control information and the analysis information in a database; training a machine learning model by using the information managed in the database; and applying second control information, obtained by adjusting the first control information about the process, on the basis of the trained machine learning model, thereby synthesizing the carbon nanomaterial.

Description

Method for manufacturing carbon nanomaterials based on machine learning model and system for performing the same

[0001] An embodiment of the present specification relates to a technology capable of continuously mass-producing carbon nanomaterials having uniform properties based on a machine learning model.

[0002] Carbon nanotubes (CNTs), an example of carbon nanomaterials, are graphite sheets rolled into nanometer-sized diameters. They possess excellent mechanical, thermal, and electrical properties. For example, the tensile strength of CNTs is approximately 100 times that of steel, their thermal conductivity is comparable to that of diamond, and their electrical conductivity is approximately 1,000 times that of copper. Accordingly, active research is being conducted across various industrial fields to utilize the outstanding properties of carbon nanomaterials such as CNTs.

[0003] However, carbon nanomaterials such as carbon nanotubes, even when manufactured under identical manufacturing process conditions, may not maintain consistent physical properties due to various external factors that impact the manufacturing process. In other words, the resulting product can vary depending on various external factors and manufacturing process conditions, making it difficult to continuously mass-produce carbon nanomaterials with uniform physical properties.

[0004] While research has been conducted to apply machine learning models to general manufacturing processes to improve quality prediction and yield, no research has been conducted on their application to specialized manufacturing processes, such as carbon nanomaterials. Therefore, technology is needed to apply machine learning models to carbon nanomaterial manufacturing processes to enable continuous, mass-produced carbon nanomaterials with uniform properties.

[0005] The embodiments of this specification are proposed to address the aforementioned issues and relate to a technology for continuously mass-producing carbon nanomaterials with uniform properties based on a machine learning model. The technical challenges addressed by the embodiments are not limited to the aforementioned technical challenges, and other technical challenges can be inferred from the following embodiments.

[0006] In order to achieve the above-described task, a method for manufacturing a carbon nanomaterial based on a machine learning model according to an embodiment of the present specification may include: a step of obtaining first control information for a process of synthesizing a carbon nanomaterial; a step of obtaining analysis information on the synthesized carbon nanomaterial in real time based on the first control information; a step of managing the first control information and the analysis information in a database; a step of training a machine learning model using information managed in the database; and a step of applying second control information in which the first control information for the process is adjusted based on the trained machine learning model to synthesize the carbon nanomaterial.

[0007] According to an embodiment, the analysis information may be characterized by including information analyzed in real time for the synthesized carbon nanomaterial using a Raman spectroscopy and a camera.

[0008] According to an embodiment, the analysis information may be characterized by including information on the Raman peak intensity ratio of the synthesized carbon nanomaterial analyzed in real time using the Raman spectrometer and information on the width and thickness of the synthesized carbon nanomaterial analyzed in real time using the camera.

[0009] According to an embodiment, the analysis information may further include information on real-time analysis of the electrical conductivity of the synthesized carbon nanomaterial.

[0010] According to an embodiment, the carbon nanomaterial may be characterized by including a carbon nanotube (CNT).

[0011] According to an embodiment, the carbon nanotube may be characterized in that it is synthesized by at least one method among chemical vapor deposition (CVD), arc discharge, and laser ablation.

[0012] According to an embodiment, the first control information may be characterized by including information on the temperature of a heat source corresponding to a preset position in an electric furnace required for the process of synthesizing the carbon nanomaterial.

[0013] According to an embodiment, the first control information may be characterized by including information on the input amount and input speed of raw materials input into the process for synthesizing the carbon nanomaterial.

[0014] According to an embodiment, the first control information may be characterized by including information on the amount of transport gas input into the process of synthesizing the carbon nanomaterial.

[0015] According to an embodiment, the first control information may be characterized by including information on the speed of a roller transporting the carbon nanomaterial synthesized according to the process.

[0016] According to an embodiment, the machine learning model may be characterized in that it is updated based on the acquired first control information, the acquired analysis information, and the history information accumulated in the database.

[0017] According to an embodiment, the method may include a non-transitory computer-readable recording medium having recorded thereon a program for executing a method for manufacturing a carbon nanomaterial based on a machine learning model on a computer.

[0018] In order to achieve the above-described task, a system for performing a method for manufacturing a carbon nanomaterial based on a machine learning model according to an embodiment of the present specification may include: a communication unit; a database for managing at least one piece of information; and a control unit for confirming first control information for a process for synthesizing a carbon nanomaterial, confirming analysis information for the synthesized carbon nanomaterial in real time based on the first control information, instructing management of the first control information and the analysis information in the database, training a machine learning model using information managed in the database, and applying second control information in which the first control information for the process is adjusted based on the machine learning model to instruct synthesis of the carbon nanomaterial.

[0019] According to an embodiment, the analysis information may be characterized by including information on the Raman peak intensity ratio of the synthesized carbon nanomaterial analyzed in real time using the Raman spectrometer and information on the width and thickness of the synthesized carbon nanomaterial analyzed in real time using the camera.

[0020] According to an embodiment, the analysis information may further include information on real-time analysis of the electrical conductivity of the synthesized carbon nanomaterial.

[0021] According to an embodiment, the carbon nanomaterial may be characterized by including a carbon nanotube synthesized by at least one method among a chemical vapor deposition method, an arc discharge method, and a laser deposition method.

[0022] According to an embodiment, the first control information may be characterized by including information on the temperature of a heat source corresponding to a preset position in an electric furnace required for the process of synthesizing the carbon nanomaterial.

[0023] According to an embodiment, the first control information may be characterized by including information on the input amount and input speed of raw materials input into the process for synthesizing the carbon nanomaterial.

[0024] According to an embodiment, the first control information may be characterized by including information on the amount of transport gas input into the process of synthesizing the carbon nanomaterial.

[0025] According to an embodiment, the first control information may be characterized by including information on the speed of a roller transporting the carbon nanomaterial synthesized according to the process.

[0026] Specific details of other embodiments are included in the detailed description and drawings.

[0027] According to the embodiments of the present specification, since the properties of carbon nanomaterials are changed by various conditions other than controllable conditions during the process, the relationship between these conditions can be analyzed using a machine learning model.

[0028] In addition, according to an embodiment of the present specification, control information can be adjusted and applied based on a machine learning model so as to continuously mass-produce carbon nanomaterials having uniform properties that satisfy reference values.

[0029] In addition, according to the embodiment of the present specification, since control information can be adjusted by considering analysis information obtained by analyzing carbon nanomaterials in real time, efficiency can be improved in the process of modifying a process for synthesizing carbon nanomaterials.

[0030] In addition, according to an embodiment of the present specification, carbon nanomaterials having uniform properties can be continuously mass-produced in various external environments and manufacturing processes by a machine learning model.

[0031] In addition, according to the embodiments of the present specification, the process can be monitored and controlled remotely, so that a quick response can be made when an abnormality occurs, thereby improving the efficiency of the process.

[0032] The effects of the embodiments are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the claims.

[0033] FIGS. 1A to 1C are drawings according to one embodiment for explaining a process for synthesizing carbon nanomaterials.

[0034] FIG. 2 is a diagram illustrating a process of adjusting control information based on a machine learning model in a process of synthesizing a carbon nanomaterial according to one embodiment.

[0035] FIG. 3 is a diagram for explaining environmental information, control information, and analysis information according to one embodiment.

[0036] FIG. 4 illustrates a carbon nanotube sheet (CNT sheet) according to one embodiment for explaining analysis information.

[0037] Figure 5 shows a picture of a database obtained by analyzing a CNT film through machine vision according to one embodiment.

[0038] FIG. 6 is a flowchart of a method for synthesizing carbon nanomaterials based on a machine learning model according to one embodiment.

[0039] FIG. 7 is a block diagram of a system performing the method of FIG. 6 according to one embodiment.

[0040] The terms used in the embodiments have been selected from widely used and common terms, taking into account the functions of the present disclosure. However, these terms may vary depending on the intentions of those skilled in the art, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, terms may be arbitrarily selected by the applicant, in which case their meanings will be described in detail in the relevant description. Therefore, the terms used in this disclosure should not be defined simply as names of terms, but rather based on the meanings of the terms and the overall content of the present disclosure.

[0041] When a part of the specification is said to "include" a component, this does not exclude other components, but rather implies the inclusion of other components, unless otherwise specifically stated. Furthermore, terms such as "part" and "module" used in the specification mean a unit that processes at least one function or operation, which may be implemented in hardware, software, or a combination of hardware and software.

[0042] The expression “at least one of a, b, and c” described throughout the specification may encompass ‘a alone’, ‘b alone’, ‘c alone’, ‘a and b’, ‘a and c’, ‘b and c’, or ‘all of a, b, and c’.

[0043] Below, embodiments of the present disclosure are described in detail with reference to the attached drawings so that those skilled in the art can easily implement the present disclosure. However, the present disclosure may be implemented in various different forms and is not limited to the embodiments described herein.

[0044] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings.

[0045] FIGS. 1A to 1C are drawings according to one embodiment for explaining a process for synthesizing carbon nanomaterials.

[0046] Specifically, FIGS. 1A and 1B are perspective views and cross-sectional views of a device for synthesizing carbon nanomaterials according to one embodiment, and FIG. 1C is a cross-sectional view of the X and Y portions shown in FIG. 1B according to one embodiment.

[0047] Here, the carbon nanomaterial may include graphene, graphite, and carbon nanotubes as examples. For example, the carbon nanomaterial may include not only carbon nanotubes, which are short, single fibers, but also carbon nanotube yarns, which are long fibers made by connecting multiple carbon nanotubes, and carbon nanotube sheets (CNT sheets) processed therefrom. The carbon nanotube sheets may include yarns in which carbon nanotubes are aligned, and specifically, a carbon nanotube sheet having an array structure in which carbon nanotubes are aggregated and aligned in one direction may be manufactured. The term "yarn" refers to carbon nanotubes grown and formed in the form of fibers, or a plurality of carbon nanotubes aggregated and / or fused into fibers.

[0048] Figures 1a to 1c illustrate a device (100) for manufacturing carbon nanotube yarn using a direct spinning method according to one embodiment. The direct spinning method involves injecting carbon raw materials and a catalyst together with a transport gas into a high-temperature electric furnace, synthesizing carbon nanotubes at the top of the electric furnace, and winding the continuous carbon nanotube aggregates that descend to the bottom to obtain fibers. The direct spinning method has the advantage of high production efficiency because continuous fiber manufacturing is possible by supplying the materials necessary for carbon nanotube synthesis to the electric furnace. A device (100) for manufacturing carbon nanotube yarn includes a supply unit (110) for supplying materials necessary for synthesizing carbon nanotube yarn, a main body (120) having a space for synthesizing the materials supplied through the supply unit (110) into carbon nanotube yarn, a heating unit (130) for controlling the temperature inside the main body (120), a space unit (140) formed at the lower portion of the main body (120) and having a space inside, a winding unit (150) connected to the space unit (140) for winding the carbon nanotube yarn synthesized in the main body (120), and a suction unit (170) for sucking gas inside the space unit (140). For example, the supply unit (110) can supply materials necessary for synthesizing carbon nanotube yarn into the main body (120) described below. The necessary materials refer to carbon raw materials, a catalyst, and a transport gas. The supply unit (110) may have an inlet through which a material is injected and a flow path through which the injected material can flow inside. More specifically, the supply unit (110) may include a carbon supply unit (111), a catalyst supply unit (112), a gas supply unit (113), and a nozzle unit (114). The carbon supply unit (111) supplies a carbon raw material, and the catalyst supply unit (112) supplies a catalyst, respectively. The carbon supply unit (111) and the catalyst supply unit (112) are formed on the upper portion of the device (100) for manufacturing carbon nanotube yarn. The shapes of the carbon supply unit (111) and the catalyst supply unit (112) may be formed approximately in the shape of a "Y".The upper part of the carbon supply unit (111) is inclined to the right, and may have a shape that extends downward from the middle part. The carbon supply unit (111) may include a flow path through which a fluid can flow inside. The catalyst supply unit (112) has an upper part that is inclined to the left, and like the carbon supply unit (111), the middle part extends downward. The catalyst supply unit (112) is connected to the carbon supply unit (111) by being inserted into the upper part of the carbon supply unit (111).

[0049] According to one embodiment, the carbon raw material supplied through the carbon supply unit (111) may be mixed with the catalyst supplied through the catalyst supply unit (112) at the bottom of the carbon supply unit (111). Here, the supply of the catalyst needs to be appropriately controlled in consideration of the supply amount of the supplied carbon raw material. This is because an excessive amount of catalyst may act as an impurity included in the synthesized carbon nanotube yarn, thereby lowering the purity of the synthesized carbon nanotube yarn. For this purpose, the diameter of the carbon supply unit (111) may be formed to be larger than the diameter of the catalyst supply unit (112). Preferably, the diameter of the carbon supply unit (111) may be formed to be 2 to 50 times the diameter of the catalyst supply unit (112).

[0050] According to one embodiment, the carbon raw material supplied through the carbon supply unit (111) may be a carbon compound in a gaseous or liquid state. For example, the carbon raw material may be a single or combination of known substances such as ethane, methane, ethanol, methanol, propanol, etc.

[0051] According to one embodiment, the catalyst supplied through the catalyst supply unit (112) may perform a function of initiating the synthesis of carbon nanotube yarns. For example, the catalyst may be a compound such as ferrocene, cobaltocene, or osmocene. In addition to the catalyst, a catalyst activator that promotes the action of the catalyst may also be supplied to the catalyst supply unit (112). The carbon supply unit and the catalyst supply unit described above do not necessarily have to be formed as separate supply pipes. The carbon supply unit and the catalyst supply unit may also be formed as a single supply pipe.

[0052] According to one embodiment, the gas supply unit (113) can supply a transport gas. In the embodiment, the gas supply unit (113) is mounted at the middle portion of the carbon supply unit (111). Specifically, the gas supply unit (113) can include a gas inlet (113a) and a circulation unit (113b). One end of the gas inlet (113a) is connected to a transport gas storage unit (not shown) and the other end is connected to the circulation unit (113b). The circulation unit (113b) can have an approximately cylindrical shape. The circulation unit (113b) is coupled to the outer peripheral surface of the middle portion of the carbon supply unit (111). A plurality of injection holes (111a) can be formed at the portion of the outer peripheral surface of the carbon supply unit (111) that is coupled to the circulation unit (113b) (see drawing (a) of FIG. 1C). The injection holes (111a) of the carbon supply unit (111) may be formed at regular intervals on the outer surface of the carbon supply unit (111). The transport gas introduced into the gas inlet (113a) moves to the circulation unit (113b) and circulates inside the circulation unit (113b) and is then injected into the carbon supply unit (111) through a plurality of injection holes (111a) formed on the outer surface of the carbon supply unit (111). When the transport gas is mixed with the carbon raw material and the catalyst and supplied into the main body (120), it can function to allow the carbon nanotube yarn to easily move downward in the main body (120) without sticking to the inner wall of the main body (120). Therefore, the transport gas needs to be uniformly mixed with the carbon raw material and the catalyst. In the embodiment, as described above, an injection hole (111a) is formed on the outer surface of the carbon supply unit (111), so that the transport gas can be uniformly injected from various directions, thereby allowing the transport gas to be uniformly mixed with the carbon raw material and the catalyst.

[0053] In the embodiment, the transport gas may be argon (Ar) gas only, hydrogen (H2) gas only, or a mixture of argon gas and hydrogen gas. Argon gas is an inert gas that transports the synthesized carbon nanotube yarn without reacting with the carbon raw material inside the high-temperature main body (120). In addition, argon gas does not affect the activity of the catalyst. Hydrogen gas is a reducing gas that can contribute to the synthesis of high-density carbon nanotube yarn by removing amorphous carbon. In the present embodiment, the injection ratio of hydrogen gas and argon gas may be set to 0.5:100 to 50:100, but the mixing ratio of the transport gas is not necessarily limited thereto. In addition, the inert gas and the reducing gas are not limited to argon gas and hydrogen gas, and the inert gas may be changed to nitrogen gas, and the reducing gas may be changed to a known gas such as ammonia.

[0054] According to one embodiment, a mixture of a carbon raw material, a catalyst, and a transport gas (hereinafter referred to as “mixture”) may be injected into the main body (120) described below through the nozzle part (114) of the supply part (110). The nozzle part (114) may be formed at the lower portion of the supply part (110) and may be located at the upper portion inside the main body (120). The nozzle part (114) may be formed in an approximately conical shape whose diameter decreases as it goes downward. A plurality of injection holes (114a) may be formed on the outer circumferential surface of the nozzle part (114) (see (b) of FIG. 1C). The mixture injected from the upper portion of the supply part (110) may reach the nozzle part (114) and be uniformly injected into the main body (120) through the plurality of injection holes (114a) formed in the nozzle part (114). Uniform spraying of the mixture can contribute to the synthesized carbon nanotube yarn having uniform density and properties.

[0055] In addition to the configuration of the supply unit (110) described above, the supply unit (110) may be equipped with a mass flow controller (MFC) and a pump, thereby controlling the supply pressure, supply amount, etc. of the fluid flowing within the supply unit (110) in real time. In addition, the structure of the supply unit described above is an exemplary structure for explaining an embodiment, and the present invention is not limited thereto. The shape of the supply unit or the position of the supply unit connected to the main body may be variously changed.

[0056] According to one embodiment, carbon nanotube yarn is synthesized inside the main body (120). The main body (120) has a cylindrical shape that extends vertically and includes a hollow space inside. The upper part of the main body (120) is connected to the supply part (110) described above. When a mixture is supplied into the main body (120) through the supply part (110), the mixture moves downward inside the main body (120) by gravity or the supply pressure of the mixture, and is synthesized into carbon nanotube yarn in this process. At this time, the internal temperature of the main body (120) can be heated to a temperature for synthesizing carbon nanotube yarn by the heating part (130) described below. The synthesized carbon nanotube yarn is discharged to the lower part of the main body (120). The main body (120) can be formed of a heat-resistant material or formed by having the interior coated with a heat-resistant material. The shape of the main body (120) described above does not limit the scope of the present invention. The shape of the main body (120), particularly the structure of the internal space where the carbon nanotube yarn is synthesized, can be varied. For example, it is also possible to form a separate injection port in the middle of the main body (120) to inject additives that improve the properties of the carbon nanotube yarn.

[0057] According to one embodiment, the heating unit (130) can control the internal temperature of the main body (120) to a preset temperature. Here, the preset temperature does not mean a temperature of a fixed specific value. The synthesis speed of the carbon nanotube yarn can change depending on the temperature, and the synthesis speed can affect the physical properties of the carbon nanotube yarn, so the device operator can appropriately design the preset temperature by taking this into consideration. In the present embodiment, the electric furnace is formed as a heating unit (130), and the heating unit (130) can be formed as a variety of devices - for example, a gas heating type, an electric heating type, etc. - regardless of the heating method, as long as it is a device that can uniformly heat the internal temperature of the main body (120) as a whole.

[0058] According to one embodiment, carbon nanotube yarns discharged from the lower portion of the main body (120) move to a space portion (140). The space portion (140) is connected to the lower portion of the main body (120) and may be formed in a box shape having a space inside. Carbon nanotube yarns and a gas including a transport gas may be present in the space portion (140).

[0059] According to one embodiment, the carbon nanotube yarn moved to the space portion (140) can be wound by a winding unit (150) connected to one side of the space portion (140). The winding unit (150) can include a motor and a winch connected to the motor, and the carbon nanotube yarn can be wound around the winch as the winch rotates in one direction by the motor. The rotation speed of the winch winding the carbon nanotube yarn can be appropriately designed in consideration of the synthesis speed of the carbon nanotube yarn, etc.

[0060] A device (100) for manufacturing carbon nanotube yarn according to one embodiment may further include a receiving section (160) for improving the density of the carbon nanotube yarn synthesized between the main body section (120) and the winding section (150). The receiving section (160) may be formed inside the space section (140). The receiving section (160) may contribute to producing a high-density carbon nanotube yarn by shrinking the synthesized carbon nanotube aggregate. The receiving section (160) includes a container (161) containing a solvent. The solvent may be water, acetone, dimethylformamide (DMF), or the like. The receiving section (160) may further include one or more transport rollers (162). In this embodiment, one transport roller (162) is formed on the lower side of the solvent-containing container (161) and one on the upper side of the container (161), and the transport roller (162) formed on the upper side is connected to the winding unit (150).

[0061] Information obtained during the process, such as control information and analysis information described in this specification, may be information obtained not only in the process as shown in FIGS. 1A to 1C, but also in a different process.

[0062] FIG. 2 is a diagram illustrating a process of adjusting control information based on a machine learning model in a process of synthesizing a carbon nanomaterial according to one embodiment.

[0063] Referring to Fig. 2, the electric furnace (201) is a heating device that can control the temperature of an area heated by the electric furnace (201). Specifically, the area heated by the electric furnace (201) can be divided into preset locations, and the temperature of the heat source can be controlled for each preset location, thereby changing the properties of the carbon nanomaterial being synthesized. In addition to the electric furnace (201), many pieces of equipment (not shown) can be additionally used in the process of synthesizing carbon nanomaterials, and information related to each piece of equipment can be transmitted and received to a system that performs a method for manufacturing a carbon nanomaterial via a communication device (203). Accordingly, the system can remotely monitor and control each piece of equipment via the communication device (203).

[0064] In order to mass-produce carbon nanomaterials with consistently uniform properties, it is necessary to analyze the properties of the synthesized carbon nanomaterials in real time and adjust control information accordingly. Various methods known in the related art can be applied to analyze the properties of carbon nanomaterials, but among them, analysis information such as image information, Raman peak intensity ratio, and electrical conductivity can be analyzed in real time during the manufacturing process. Specifically, a camera (205) can acquire image information of the synthesized carbon nanomaterials in real time, and a Raman spectrometer (207) can acquire Raman peak intensity ratio of the synthesized carbon nanomaterials in real time. In addition, electrical conductivity of the synthesized carbon nanomaterials can be analyzed and acquired in real time. However, a method for analyzing properties, such as a scanning electron microscope (SEM), cannot acquire analysis information in real time and thus may be excluded from the method for analyzing the properties of carbon nanomaterials. Information analyzed in real time by each device can be transmitted and received to the system via a communication device (209).

[0065] The system can check first control information and history control information (211) and analysis information and history analysis information (213). Here, the history control information and the history analysis information are information accumulated in the past and managed in a database, and the first control information may be information received through a communication device (203) and the analysis information may be information received through a communication device (209). That is, the system can check the history control information and the history analysis information managed in the database, and check the first control information and the analysis information received from the communication devices (203, 209). The first dashboard (215) can display the control information and the analysis information. Alternatively, the first dashboard (215) can additionally display the history control information and the history analysis information. Therefore, an administrator managing the system can not only monitor the process based on the information displayed on the first dashboard (215), but also check the analysis results for the synthesized carbon nanomaterial.

[0066] The system can train a machine learning model (217) based on historical control information and historical analysis information, and update the machine learning model (217) based on first control information and real-time analysis information. The machine learning model (217) can display second control information (221) through a second dashboard (219), and the second control information (221) can be transmitted to related equipment through a communication device (203, 209). Accordingly, an administrator managing the system can remotely monitor and control the process.

[0067] Here, the second control information is control information in which the first control information is adjusted based on the machine learning model (217). Specifically, if the analysis information confirmed in real time through the first dashboard (215) does not reach the reference value, the machine learning model (217) can output the second control information so as to correspond to the reference value. At this time, the reference value may be set differently depending on the field to which the manufactured carbon nanomaterial is applied. For example, the reference values ​​corresponding to carbon nanomaterials applied to semiconductors and carbon nanomaterials applied to air conditioning devices may be set differently, and the machine learning model can output the second control information by considering the different reference values.

[0068] The machine learning model can output second control information based on first control information, historical control information, analysis information, and historical analysis information, and can be reinforced by granting a reward when the properties (i.e., analysis information) of the carbon nanomaterial synthesized based on the second control information satisfy a reference value. Alternatively, the machine learning model can be created using reinforcement learning, but it can also be created using other algorithms. For example, various machine learning algorithms such as unsupervised learning and rule-based machine learning algorithms can be applied.

[0069] FIG. 3 is a diagram for explaining environmental information, control information, and analysis information according to one embodiment.

[0070] Referring to FIG. 3, environmental information (310), control information (320), and analysis information (330) according to one embodiment can be confirmed.

[0071] Here, the environmental information (310) according to the embodiment is information related to the environment of the space where carbon nanomaterials are synthesized, and may include, for example, information on the temperature in the atmosphere, the relative humidity in the atmosphere, the atmospheric pressure, and the temperature and relative humidity of the acquisition space inside the facility where carbon nanotubes, which are an example of carbon nanomaterials, are manufactured. Specifically, the environmental information (310) is information measured within a certain range by a sensor in relation to the temperature in the atmosphere, the relative humidity in the atmosphere, the atmospheric pressure, and the temperature and relative humidity of the acquisition space. At this time, unlike the controllable control information (320), the environmental information (310) is uncontrollable information, and the physical properties of the carbon nanomaterial may be changed by the environmental information (310) even under the same conditions as the control information (320). Therefore, the machine learning model can learn the relationship between the control information and the analysis information by taking the environmental information into consideration.

[0072] The control information (320) according to the embodiment is information that can be adjusted in the process of synthesizing carbon nanomaterials, and nine of them are illustrated in FIG. 3 as an example. In addition to the nine pieces of information illustrated in FIG. 3, additional information may be included in the control information (320). Specifically, the electric furnace, which is a heating device, is a long piece of equipment, and the temperature of the area heated by the electric furnace can be changed at each preset location, thereby adjusting the properties of the carbon nanomaterial. At this time, the number of areas heated by the electric furnace is not limited to three, and may be determined by considering various factors such as the performance of the electric furnace.

[0073] Furthermore, Acetone is an example of a raw material, Ferrocene is an example of a catalyst, and Thiophene is an example of a catalyst activator. The amount and rate of introduction of the catalyst and catalyst activator can be controlled within a relative range compared to the raw material, and the properties of the carbon nanomaterial can be controlled based on this. Here, the catalyst activator acts as a promoter during carbon nanotube conversion, increasing the carbon diffusion rate, thereby enabling carbon nanotube synthesis in a short period of time.

[0074] In addition, raw material transport gas 01 is a gas that transports raw materials, catalysts, and catalyst activators and is regulated in a predetermined unit within a preset range, and transport gas 02 is a gas that is introduced to uniformly cause a carbon nanomaterial synthesis reaction having uniform properties and can be regulated in a predetermined unit within a preset range. Raw material transport gas 01 and transport gas 02 may be the same or different gases.

[0075] In addition, the roller can transport a carbon nanotube sheet, which is a carbon nanomaterial, or a film obtained by obtaining a carbon nanotube sheet, and the properties of the carbon nanotube sheet can be changed by the roller speed that is adjusted in a constant speed unit within a preset speed range.

[0076] Analysis information (330) according to an embodiment is information analyzed in real time for a carbon nanomaterial synthesized by a camera, a Raman spectrometer, and other equipment. In addition, even if information is acquired by equipment capable of analyzing the properties of carbon nanomaterials, analysis information that is not acquired in real time may be excluded from the analysis information (330). Information related to the width and thickness of a CNT sheet can be analyzed in real time from image information acquired through a camera. Fig. 4 illustrates a carbon nanotube sheet (CNT sheet) according to an embodiment for explaining analysis information (330). For example, the CNT sheet width (410) may correspond to information confirmed in units of a predetermined length in a preset length range based on an average value in the image information. In addition, the average thickness of the entire CNT sheet may be information confirmed by applying a preset method to the image information for the synthesized CNT sheet. In addition, the uniformity of the thickness of the CNT sheet area, which indicates whether the thickness is uniform, may be information corresponding to the standard deviation of information confirmed by applying a preset method to the image information. In addition, unlike the area thickness uniformity of the CNT sheet, the vertical thickness uniformity of the CNT sheet may be information confirmed in a predetermined unit within a preset range as the thickness uniformity for the area (430) in the direction in which the CNT sheet is manufactured. In addition, the Raman peak intensity ratio (Id / Ig) for the carbon nanotube sheet acquired through a Raman spectrometer may be information confirmed in a predetermined unit within a preset range. In addition, the electrical conductivity of the carbon nanotube sheet may be information confirmed in real time within a preset range.

[0077] Figure 5 shows a picture of a database obtained by analyzing a CNT film through machine vision according to one embodiment.

[0078] Referring to FIG. 5, as a result of analyzing and database-izing a CNT film through a machine vision camera, an area (510) may correspond to the CNT film. Specifically, a CNT film may be photographed through a machine vision camera, and the X-axis and Y-axis of the photographed image may be divided into preset pixels. For example, an image of a CNT film photographed through a camera may be divided into 7200 pixels in the X-axis and 3600 pixels in the Y-axis. Here, 7200 pixels and 3600 pixels are merely examples, and the present technology may also be applied to a case where a CNT film is photographed through a camera having different performance. When the contrast is quantified for each pixel, it may be displayed as in FIG. 5.

[0079] A machine learning model can be trained based on multiple databases of information collected by analyzing CNT films using a machine vision camera. The machine learning model can analyze the properties of carbon nanomaterials captured in real time by the camera and adjust control information to produce carbon nanomaterials with properties that meet standard values.

[0080] FIG. 6 is a flowchart of a method for synthesizing carbon nanomaterials based on a machine learning model according to one embodiment.

[0081] Referring to FIG. 6, first control information for a process of synthesizing carbon nanomaterials in S610 can be obtained.

[0082] The first control information is information that can be adjusted during the process, and thus the properties of the carbon nanomaterial can vary. For example, the first control information may include information about the temperature of a heat source corresponding to a preset location in an electric furnace, which is a heating device, in the process of synthesizing carbon nanomaterials. For example, the first control information may include information about the input amount and input speed of raw materials input into the process of synthesizing carbon nanomaterials. For example, the first control information may include information about the input amount of transport gas input into the process of synthesizing carbon nanomaterials. For example, the first control information may include information about the speed of a roller that transports carbon nanomaterials synthesized according to the process.

[0083] In S620, analysis information on carbon nanomaterials synthesized based on the first control information can be acquired in real time. Only information analyzed in real time (e.g., image information, Raman peak intensity ratio, electrical conductivity, etc.) on carbon nanomaterials can be included, and information not analyzed in real time (e.g., scanning electron microscope) can be excluded. Information analyzed in real time by each device can be transmitted to the system via a communication device.

[0084] Specifically, the analysis information may include real-time analysis of the Raman peak intensity ratio of the synthesized carbon nanomaterial using a Raman spectrometer and real-time analysis of the width and thickness of the synthesized carbon nanomaterial using a camera. Furthermore, the analysis information may further include real-time analysis of the electrical conductivity of the synthesized carbon nanomaterial.

[0085] Here, the carbon nanomaterial includes carbon nanotubes, and the carbon nanotubes can be synthesized by at least one method selected from the group consisting of chemical vapor deposition (CVD), arc discharge, and laser ablation. Anything known in the art regarding chemical vapor deposition, arc discharge, and laser ablation can be applied. For example, the chemical vapor deposition method is a method in which a catalyst metal film is formed on a substrate and then etched with an etching gas, thereby allowing a plurality of catalyst fine particles to react with a raw material in the formed state to synthesize carbon nanotubes. For example, the arc discharge method is a method in which, for example, a pure carbon electrode is used as a cathode and a metal-doped carbon electrode is used as an anode, an arc discharge is generated between the two carbon electrodes, and carbon vaporized from the anode moves to the cathode surface, thereby synthesizing carbon nanotubes synthesized on the cathode surface. For example, laser deposition is a method of synthesizing carbon nanotubes by vaporizing graphite using a laser in a high-temperature atmosphere filled with an inert gas. Additionally, known synthesis methods in the field of carbon nanomaterial manufacturing technology can be applied.

[0086] At S630, first control information and real-time acquired analysis information can be managed in a database. At S640, a machine learning model can be trained using the information managed in the database. The machine learning model can be updated based on the first control information, real-time analyzed analysis information, and accumulated historical information in the database. At S650, carbon nanomaterials can be synthesized by applying second control information, in which the first control information for the process is adjusted based on the machine learning model.

[0087] FIG. 7 is a block diagram of a system that performs the method of FIG. 6 according to one embodiment. FIG. 7 illustrates components related to the present embodiment, but is not limited thereto. In addition to the components illustrated in FIG. 7, other general-purpose components may be included. Any content that overlaps with the descriptions provided above may be omitted.

[0088] The system (700) of FIG. 7 may be a device that performs the aforementioned method. Referring to FIG. 7, the system (700) may include a communication unit (710), a database (720), a control unit (730), and a bus (740). The database (720) and the control unit (730) may communicate with each other via the bus (740). Each of the database (720) and the control unit (730) may mean a unit that processes at least one function or operation, and may be implemented by hardware, software, or a combination of hardware and software.

[0089] In an embodiment, the system (700) may include a database (720) that stores various information related to the system (700). For example, the database (720) may store at least one instruction for operating the system (700). In this case, the database (720) and the control unit (730) may perform various operations based on these instructions. The control unit (730) may perform the aforementioned operations as the instructions stored in the database (720) are executed by the control unit (730). The database (720) may be a volatile memory or a non-volatile memory. For example, the database (720) may store data processed by the control unit (730) and data to be processed. The database (720) may include random access memory (RAM) such as dynamic random access memory (DRAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), CD-ROM, Blu-ray or other optical disk storage, hard disk drive (HDD), solid state drive (SSD), or flash memory.

[0090] In an embodiment, a command is executed in a control unit (730), and the control unit (730) can check control information for a process of synthesizing a carbon nanomaterial, check analysis information for the synthesized carbon nanomaterial in real time based on the control information, instruct to manage the control information and the analysis information in the database, train a machine learning model using information managed in the database, and instruct to adjust the control information related to the synthesis of the carbon nanomaterial based on the machine learning model.

[0091] The systems and various devices according to the above-described embodiments may include a control unit, a memory for storing and executing program data, a permanent storage unit such as a disk drive, a communication port for communicating with an external device, and a user interface device such as a touch panel, a key, a button, etc. Methods implemented as software modules or algorithms may be stored on a computer-readable recording medium as computer-readable codes or program commands executable on the control unit. Here, the computer-readable recording medium includes a magnetic storage medium (e.g., read-only memory (ROM), random-access memory (RAM), floppy disk, hard disk, etc.) and an optical reading medium (e.g., CD-ROM, DVD: Digital Versatile Disc)). The computer-readable recording medium may be distributed to computer systems connected to a network, so that the computer-readable code may be stored and executed in a distributed manner. The medium may be readable by a computer, stored in a memory, and executed by the control unit.

[0092] The present embodiment may be represented by functional block configurations and various processing steps. These functional blocks may be implemented by various numbers of hardware and / or software configurations that perform specific functions. For example, the embodiment may employ direct circuit configurations such as memory, processing, logic, look-up tables, etc., which may perform various functions under the control of one or more microcontrollers or other control devices. Similarly, the present embodiment may be implemented in a programming or scripting language such as C, C++, C#, Python, Java, assembler, etc., including various algorithms implemented by a combination of data structures, processes, routines, or other programming configurations.

[0093] The above-described embodiments are merely examples, and other embodiments may be implemented within the scope of the claims set forth below.

Claims

1. A step of obtaining first control information for a process of synthesizing carbon nanomaterials; A step of obtaining analysis information on the synthesized carbon nanomaterial in real time based on the first control information; A step of managing the first control information and the analysis information in a database; A step of training a machine learning model using information managed in the above database; and A step of synthesizing the carbon nanomaterial by applying second control information in which the first control information for the process is adjusted based on the learned machine learning model, A method for manufacturing carbon nanomaterials based on a machine learning model.

2. In paragraph 1, The above analysis information is characterized in that it includes information analyzed in real time for the synthesized carbon nanomaterial using a Raman spectroscopy and a camera. A method for manufacturing carbon nanomaterials based on a machine learning model.

3. In paragraph 2, The above analysis information is characterized in that it includes information analyzed in real time using the Raman spectrometer regarding the Raman peak intensity ratio of the synthesized carbon nanomaterial and information analyzed in real time using the camera regarding the width and thickness of the synthesized carbon nanomaterial. A method for manufacturing carbon nanomaterials based on a machine learning model.

4. In paragraph 3, The above analysis information is characterized in that it further includes information on the electrical conductivity of the synthesized carbon nanomaterial analyzed in real time. A method for manufacturing carbon nanomaterials based on a machine learning model.

5. In paragraph 1, The above carbon nanomaterial is characterized in that it includes a carbon nanotube (CNT). A method for manufacturing carbon nanomaterials based on a machine learning model.

6. In paragraph 5, The above carbon nanotube is characterized in that it is synthesized by at least one method among chemical vapor deposition (CVD), arc discharge, and laser ablation. A method for manufacturing carbon nanomaterials based on a machine learning model.

7. In paragraph 5, The first control information is characterized in that it includes information on the temperature of a heat source corresponding to a preset position in an electric furnace required for the process of synthesizing the carbon nanomaterial. A method for manufacturing carbon nanomaterials based on a machine learning model.

8. In paragraph 5, The first control information is characterized in that it includes information on the input amount and input speed of raw materials input into the process of synthesizing the carbon nanomaterial. A method for manufacturing carbon nanomaterials based on a machine learning model.

9. In paragraph 5, The first control information is characterized in that it includes information on the amount of transport gas input into the process of synthesizing the carbon nanomaterial. A method for manufacturing carbon nanomaterials based on a machine learning model.

10. In paragraph 5, The first control information is characterized in that it includes information on the speed of the roller transporting the carbon nanomaterial synthesized according to the process. A method for manufacturing carbon nanomaterials based on a machine learning model.

11. In paragraph 1, The machine learning model is characterized in that it is updated based on the acquired first control information, the acquired analysis information, and the history information accumulated in the database. A method for manufacturing carbon nanomaterials based on a machine learning model.

12. A non-transitory computer-readable recording medium having recorded thereon a program for executing the method of paragraph 1 on a computer.

13. Ministry of Communications; a database that manages at least one piece of information; and A control unit that checks first control information for a process of synthesizing a carbon nanomaterial, checks analysis information on the synthesized carbon nanomaterial in real time based on the first control information, instructs to manage the first control information and the analysis information in the database, trains a machine learning model using information managed in the database, and applies second control information in which the first control information for the process is adjusted based on the machine learning model to instruct the synthesis of the carbon nanomaterial. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

14. In paragraph 13, The above analysis information is characterized in that it includes information analyzed in real time using a Raman spectrometer regarding the Raman peak intensity ratio of the synthesized carbon nanomaterial and information analyzed in real time using a camera regarding the width and thickness of the synthesized carbon nanomaterial. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

15. In paragraph 14, The above analysis information is characterized in that it further includes information on the electrical conductivity of the synthesized carbon nanomaterial analyzed in real time. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

16. In paragraph 13, The above carbon nanomaterial is characterized in that it includes a carbon nanotube synthesized by at least one method among chemical vapor deposition, arc discharge, and laser deposition. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

17. In paragraph 16, The first control information is characterized in that it includes information on the temperature of a heat source corresponding to a preset position in an electric furnace required for the process of synthesizing the carbon nanomaterial. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

18. In paragraph 16, The first control information is characterized in that it includes information on the input amount and input speed of raw materials input into the process of synthesizing the carbon nanomaterial. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

19. In paragraph 16, The first control information is characterized in that it includes information on the amount of transport gas input into the process of synthesizing the carbon nanomaterial. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

20. In paragraph 16, The first control information is characterized in that it includes information on the speed of the roller transporting the carbon nanomaterial synthesized according to the process. A system for performing a method for manufacturing carbon nanomaterials based on a machine learning model.

Images