Apparatus and method for mixing materials

Abstract

The present disclosure relates to a technique for mixing fired products that are materials for a positive electrode active material. Provided are an apparatus and a method for mixing materials, the method comprising: generating an Nth (N being an integer greater than or equal to 1) combination including a plurality of matching combinations in which a plurality of first materials and a plurality of second materials are classified; calculating matching suitability for each of the matching combinations on the basis of characteristic information of each material; calculating total suitability of the Nth combination on the basis of the matching suitability; selecting, as a candidate combination, a combination having the highest total suitability from among the total suitability of a first combination to the total suitability of the Nth combination that have been calculated; determining whether a preset criterion is satisfied; determining the candidate combination as an optimal combination if N satisfies the preset criterion; and generating an (N+1)th combination if N does not satisfy the preset criterion.

Description

Material mixing device and method

[0001] The present disclosure relates to a technique for mixing a sintered product that is a material for an anode.

[0002] Recently, due to the emergence of electric vehicles and the impact of climate change, research on secondary battery technology, which is more stable than primary batteries and can be used repeatedly through charging, is actively underway.

[0003] The cathode material, which is a key component of secondary batteries and determines the capacity and voltage of the battery, includes representative cathode materials such as NCM (nickel, cobalt, manganese, lithium), LFP (iron, phosphorus, lithium), and NCA (nickel, cobalt, aluminum, lithium). The cathode material can be produced through a process of mixing a precursor material with a raw material containing lithium or sodium, a process of calcining based on a preset temperature and then drying, and a process of mixing the materials produced afterward.

[0004] However, in the mixing process, the selection of materials to be mixed has traditionally been done manually, relying on the operator's experience and intuition. Consequently, there are issues such as inconsistent quality of the produced cathode materials and excessive production time. Furthermore, manual operation carries the risk of operator error, which in turn results in defective products being included in the manufactured cathode materials.

[0005] The present disclosure aims to provide a technique for mixing a sintered product, which is a material for an anode.

[0006] In one aspect, the present embodiments provide a material mixing apparatus comprising: a combination generating unit that generates a N (N is an integer greater than or equal to 1) combination including a plurality of matching combinations classified from a plurality of first materials and a plurality of second materials; a suitability calculation unit that calculates a matching suitability for each matching combination generated by the combination generating unit based on characteristic information of each material and calculates a total suitability of the N combination based on the matching suitability; a candidate combination selection unit that selects a combination having the highest total suitability among the total suitability of the N combinations from the total suitability of the first combination calculated earlier as a candidate combination; and an optimal combination determination unit that determines whether N satisfies a preset criterion, and if N satisfies the preset criterion, determines the candidate combination as the optimal combination, and if N does not satisfy the preset criterion, requests the combination generating unit to generate an N+1 combination.

[0007] In another aspect, the present embodiments provide a material mixing method comprising: a combination generation step for generating a combination N (where N is an integer greater than or equal to 1) including a plurality of matching combinations classified from a plurality of first materials and a plurality of second materials; a suitability calculation step for calculating a matching suitability for each matching combination generated in a combination generation unit based on characteristic information of each material and calculating a total suitability of the N combination based on the matching suitability; a candidate combination selection step for selecting a combination having the highest total suitability among the total suitability of the N combinations from the total suitability of the first combination calculated earlier as a candidate combination; and an optimal combination determination step for determining whether N satisfies a preset criterion, and if N satisfies the preset criterion, determining the candidate combination as the optimal combination, and if N satisfies the preset criterion, requesting the combination generation step to generate an N+1 combination.

[0008] The present disclosure may provide a technique for mixing a sintered product that is a material for an anode.

[0009] FIG. 1 is a drawing for explaining the configuration of a device for mixing a calcined product, which is a material of an anode material, according to one embodiment.

[0010] FIG. 2 is a flowchart for schematically explaining the process of manufacturing a cathode material according to one embodiment.

[0011] FIG. 3 is a diagram for schematically explaining the process of mixing a sintered product, which is a material of an anode material according to one embodiment.

[0012] FIG. 4 is a diagram illustrating the learning process of an artificial intelligence model that predicts the quality of a mixture according to one embodiment.

[0013] FIG. 5 is a flowchart for specifically explaining the process of selecting an optimal combination for mixing a sintered product according to one embodiment.

[0014] FIG. 6 is a flowchart for explaining the process of generating a new combination when an optimal combination according to one embodiment is not selected.

[0015] FIG. 7 is a flowchart for illustrating, by example, a method for generating a new combination according to one embodiment.

[0016] FIG. 8 is a drawing for explaining a method of mixing a sintered product, which is a material of an anode material according to one embodiment.

[0017] Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the exemplary drawings. In assigning reference numerals to the components of each drawing, the same components may have the same reference numeral as much as possible, even if they are shown in different drawings. Furthermore, in describing the embodiments, if it is determined that a detailed description of related known components or functions may obscure the essence of the technical concept, such detailed description may be omitted. Where terms such as "comprising," "having," or "consisting of" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it may include a plural unless otherwise specified.

[0018] Additionally, terms such as first, second, A, B, (a), (b), etc., may be used to describe the components of the present disclosure. These terms are used merely to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by such terms.

[0019] In describing the positional relationship of components, where it is stated that two or more components are "connected," "combined," or "joined," it should be understood that while the two or more components may be directly "connected," "combined," or "joined," they may also be "connected," "combined," or "joined" with other components "intervened." Here, the other components may be included in one or more of the two or more components that are "connected," "combined," or "joined" with one another.

[0020] In describing the temporal flow relationship regarding components, methods of operation, or methods of production, for example, when the temporal or sequential relationship is described using "after," "following," "next," or "before," it may include cases where the relationship is not continuous unless "immediately" or "directly" is used.

[0021] Meanwhile, where numerical values ​​or corresponding information regarding a component (e.g., levels, etc.) are mentioned, even without separate explicit notation, the numerical values ​​or corresponding information may be interpreted as including a range of error that may occur due to various factors (e.g., process factors, internal or external shocks, noise, etc.).

[0022] The embodiments are described in detail below with reference to the drawings.

[0023]

[0024] FIG. 1 is a drawing for explaining the configuration of a device for mixing a calcined product, which is a material of an anode material, according to one embodiment.

[0025] Referring to FIG. 1, the material mixing device (100) of the present disclosure includes a combination generating unit (110) that generates a N (N is an integer greater than or equal to 1) combination, which includes a plurality of matching combinations in which a plurality of first materials and a plurality of second materials are classified.

[0026] The cathode material used in secondary batteries is manufactured through a process in which a precursor and a raw material are mixed, a process in which the material is heated to a high temperature in a kiln and fired, a process in which impurities contained in the fired product are removed, a process in which the fired product from which impurities have been removed is dried, a process in which the dried fired product is mixed according to type, and a process in which the mixed product is coated. In this disclosure, a material that is heated to a high temperature in a kiln and fired may be referred to as a fired product.

[0027] In particular, regarding the process of mixing sintered products according to type during the manufacturing of cathode materials, as mentioned above, conventionally the types of sintered products to be mixed were determined based on the judgment of the operator; consequently, there is a problem where the quality of the produced cathode materials is inconsistent and defective products are produced.

[0028] The present disclosure proposes a method that enables optimal mixing matching between fired products by considering the characteristics of each fired product and the characteristics of the mixed fired products.

[0029] For example, the aforementioned first material may be either a small particle size or a large particle size among the sintered products produced after the cathode material sintering process, and if the first material is a small particle size, the second material may be a large particle size, and if the first material is a large particle size, the second material may be a small particle size.

[0030] The particle size of the sintered product may vary depending on the type and mixing ratio of the precursor and raw material. In this disclosure, sintered products with large particles are referred to as large particle size, and those with small particles are referred to as small particle size. For example, sintered products with a size of 10 micrometers or more may be classified as large particle size, and those with a size of less than 10 micrometers may be classified as small particle size. However, this is merely one example of classifying large particle size and small particle size, and the aforementioned preset temperature and the size criteria for distinguishing large particle size and small particle size may be set in various ways as needed. Additionally, in this disclosure, a sintered product that serves as a material for an anode material may be referred to as a material.

[0031] As another example, the combination generating unit (110) of the present disclosure may be configured such that in the case of the first combination, which is the first combination, the first material and the second material are included one each, and each matching combination may be configured such that the first material and the second material are randomly matched.

[0032] The present disclosure proposes a method in which, for a first combination which is the first combination generated through a combination generation unit (110), a matching combination is generated by randomly matching one of a plurality of first materials with one of a plurality of second materials, and for a second combination which is not the first combination generated, some of the matching combinations included in the combination selected as a candidate combination in the previous step are rematched based on the matching suitability of each matching combination and a preset algorithm. The aforementioned preset algorithm may include a genetic algorithm.

[0033] The material mixing device (100) of the present disclosure includes a suitability calculation unit (120) that calculates the degree of matching suitability for each matching combination generated in a combination generating unit (110) based on characteristic information of each material, and calculates the total suitability of the Nth combination based on the calculated degree of matching suitability.

[0034] For example, the suitability calculation unit (120) of the present disclosure may obtain first characteristic information of a plurality of first materials and second characteristic information of a plurality of second materials, and calculate a quality value for each of a plurality of matching combinations based on the first characteristic information and the second characteristic information.

[0035] For example, the material mixing device (100) of the present disclosure may match one of a plurality of first materials with one of a plurality of second materials, and each matching pair may be combined to form a single combination. The aforementioned matching pair may be referred to as a matching combination in the present disclosure. Additionally, characteristic information for each material may be obtained, and based on the characteristic information, a quality value of a matching combination in which one first material and one second material are mixed may be calculated.

[0036] The aforementioned characteristic information may include at least one of the particle size, magnetic impurities, and residual lithium concentration of each sintered product. However, this is merely one example representing the characteristics of each material, and various information may be included as needed.

[0037] As another example, the aforementioned quality figures can be calculated through an artificial intelligence model trained to minimize loss based on training data reflecting the characteristic information of each material and a Mean Square Error (MSE) loss function.

[0038] Specifically, the material mixing device (100) of the present disclosure can input characteristic information of each material into an artificial intelligence model to calculate a quality value, and perform training of the artificial intelligence model so that the difference between the output quality information and the actual quality information of the mixture is minimized using an MSE loss function, and through the trained artificial intelligence model, input quality information for two materials into the artificial intelligence model in an actual mixing process to calculate a quality value of the mixture of two materials. The aforementioned quality value may mean a real number between 0 and 100, or may mean a grade such as high, medium, or low.

[0039] The material mixing device (100) of the present disclosure can be configured to include multiple matching combinations in a single combination. The number of possible cases as matching combinations can be calculated in various ways depending on the number of each material. For example, if there are 4 first materials and 4 second materials each, and a matching combination is generated by matching any one of the first materials with any one of the second materials, the number of cases for multiple matching combinations can be 24. As another example, if there are 5 first materials and 5 second materials each, and a matching combination is generated by matching any one of the first materials with any one of the second materials, the number of cases for multiple matching combinations can be 120. As such, the number of cases for multiple matching combinations increases as the number of each material increases. The material mixing device (100) of the present disclosure can be configured to include only matching combinations in the combination where the calculated quality value among the various cases for matching combinations is greater than or equal to a preset value.

[0040] As another example, the aforementioned matching fit is calculated based on quality figures, and the aforementioned total fit can be calculated based on the sum of each matching fit.

[0041] The apparatus (100) for mixing the sintered product of the present disclosure includes a candidate combination selection unit (130) that selects a combination having the highest total suitability among the total suitability of the first combination calculated and the total suitability of the Nth combination as a candidate combination.

[0042] The material mixing device (100) of the present disclosure can classify a first material and a second material to generate a first combination including a plurality of matching combinations and calculate the total suitability for the first combination. Additionally, it can generate a second combination and calculate the total suitability for the second combination. In the same manner, it can generate an Nth combination and calculate the total suitability for the Nth combination. Once the total suitability for each combination is calculated, the material mixing device (100) of the present disclosure can select the combination with the highest total suitability among the first combination to the Nth combination as a candidate combination.

[0043] However, the material mixing device (100) of the present disclosure may determine the first combination as a candidate combination immediately if the generated combination is the first combination generated initially.

[0044] When a candidate combination is selected, the material mixing device (100) of the present disclosure can proceed with the optimal combination selection process by determining whether the number of combination generations or the number of candidate combination selections satisfies a preset standard.

[0045] When a candidate combination is selected, the apparatus (100) for mixing the sintered product of the present disclosure determines whether the number of combination generations or the number of candidate combination selections satisfies a preset standard, and if the preset standard is satisfied, determines the candidate combination as the optimal combination, and if the preset standard is not satisfied, requests the combination generation unit (110) to generate the N+1th combination. The optimal combination determination unit (140) includes an optimal combination determination unit (140).

[0046] The aforementioned preset criteria may include cases where the number of combination generation counts N is greater than or equal to a first threshold. Alternatively, the preset criteria may include cases where N is less than a preset first threshold and candidate combinations are selected consecutively for a preset second threshold.

[0047] The aforementioned first threshold and second threshold are integers greater than or equal to 1 and can be set in various ways as needed.

[0048] The fact that N is greater than or equal to the first threshold and that a specific combination has been selected consecutively as a candidate combination for a second threshold that is pre-set may mean that the last selected candidate combination can be recognized as being composed of optimal matching combinations.

[0049] For example, in cases where the number of combination generations or the number of candidate combination selections does not satisfy a preset standard, such that N is less than a preset first threshold, or N is greater than or equal to a preset first threshold and candidate combinations are not selected consecutively for a preset second threshold, the material mixing device (100) of the present disclosure may request the combination generation unit (110) to generate a new combination, the N+1 combination, as described above.

[0050] When the optimal combination determination unit (140) of the present disclosure requests the combination generation unit (110) to generate the N+1th combination, the combination generation unit (110) can rematch a plurality of matching combinations included in the aforementioned candidate combinations based on the matching suitability of the candidate combinations and a preset algorithm, and generate the N+1th combination based on the rematching result.

[0051] Specifically, the combination generation unit (110) can select a predetermined number of matching combinations with low matching suitability among a plurality of matching combinations included in the candidate combinations, and rematch the predetermined number of selected matching combinations through a predetermined algorithm.

[0052] The aforementioned preset number is an integer of 2 or more, and can be set in various ways as needed.

[0053] When the N+1 combination is generated, the material mixing device (100) of the present disclosure calculates the total suitability of the N+1 combination in the manner described above, performs a process of selecting candidate combinations, and determines whether they satisfy a preset criterion.

[0054] As another example, even though the N+1 combination has been generated, if a preset criterion is not satisfied, the material mixing device of the present disclosure may generate the N+2 combination and repeat the same process.

[0055] As another example, the optimal combination determination unit (140) of the present disclosure can determine a candidate combination as the optimal combination when N is greater than or equal to a preset first threshold, and can also determine a candidate combination as the optimal combination when N is less than a preset first threshold and the candidate combination is selected consecutively for a preset second threshold.

[0056] For example, if a first threshold is set to 10 and a second threshold is selected to 3, and a third combination is selected as a candidate combination when a third combination is generated, and a third combination is selected as a candidate combination when a fourth combination is generated, and a third combination is selected as a candidate combination when a fifth combination is generated, then the material mixing device (100) of the present disclosure can determine that the third combination is the optimal combination even though it is less than the first threshold because N is 5, and can generate an instruction to mix the materials according to the matching combination included in the combination.

[0057] Once the materials are mixed according to each matching combination, a coating process can be performed on each mixed object. After coating, a packaging process can be performed, and the cathode material production process can be completed.

[0058] The aforementioned preset algorithm may include a genetic algorithm.

[0059]

[0060] The present disclosure has the advantage of improving production efficiency by shortening production time and reducing costs by automatically selecting the mixing target between two types of materials, maintaining consistency in quality by determining the optimal combination according to certain standards, increasing work efficiency by reducing the burden on workers by preventing worker errors in advance, and saving power consumed in the workplace by reducing unnecessary calculations by creating conditions that allow the work to be terminated early through the setting of a first threshold and a second threshold.

[0061]

[0062] Below, the overall process of mixing the sintered product is explained in more detail with reference to the drawing.

[0063]

[0064] FIG. 2 is a flowchart for schematically explaining the process of manufacturing a cathode material according to one embodiment.

[0065] Referring to FIG. 2, the process of manufacturing a cathode material, which is mainly included as a component of a secondary battery, is shown. The cathode material can be manufactured by mixing raw materials, calcining at a high temperature, removing impurities, and then drying and coating.

[0066] Specifically, a co-precipitation process is performed to generate a precursor that serves as a basic raw material for the production of cathode materials (S200).

[0067] A precursor refers to a material in the preceding stage of a target material produced through a chemical reaction. In the battery production process, precursors can be considered raw materials for the production of cathode materials. For example, a cathode active material precursor may be composed of a chemical formula containing at least one element selected from nickel (Ni), cobalt (Co), manganese (Mn), aluminum (Al), zinc (Zn), boron (B), tungsten (W), chromium (Cr), and magnesium (Mg). For example, [Ni a Co b Mn c Mg d It may include compounds composed of the chemical formula ](OH)2(a, b, c, d are natural numbers greater than or equal to 1).

[0068] In addition, the raw material containing lithium or sodium may be, for example, a material containing lithium carbonate (Li2CO3) or lithium hydroxide (LiOH).

[0069] A precursor can be produced by dissolving at least one of the aforementioned nickel (Ni), cobalt (Co), manganese (Mn), aluminum (Al), zinc (Zn), boron (B), tungsten (W), chromium (Cr), and magnesium (Mg) to create a metal solution, and by adjusting the pH of the metal solution to wash and dry the material precipitated in the metal solution. Once the precursor is produced, a transition metal oxide can be produced by mixing the precursor with a raw material containing the aforementioned lithium carbonate (Li2CO3) or lithium hydroxide (LiOH).

[0070] When a transition metal oxide is produced through a co-precipitation process, a calcination process is performed to calcine the transition metal oxide at a high temperature (S210).

[0071] Calcination is a step of heating the mixture at a high temperature, which can be performed at a temperature of 700°C to 1000°C. If the calcination temperature is excessively low, the structural stability of the cathode material may be reduced even if it is completed. In addition, if the calcination temperature is excessively high, non-uniform growth of particles may occur, and a decrease in capacity may occur.

[0072] However, a firing temperature of 700°C to 900°C is merely an example, and can be performed at various temperatures depending on the contained material, firing environment, etc.

[0073] In addition, the firing time can be performed for as little as 5 hours or as long as 40 hours.

[0074] As described above, the material produced through the firing process may be referred to as a fired product in this disclosure.

[0075] When a sintered product is produced through a sintering process, a process of crushing / classifying / de-ironing the produced sintered product is performed (S220).

[0076] Crushing or classification refers to grinding the produced sintered product into powder, and deironing refers to the process of removing trace amounts of iron or non-ferrous metals contained in the sintered product.

[0077] Iron or non-ferrous metals contained in the fired product are removed, and when it becomes powder, a washing and drying process is performed to remove impurities contained in the fired product (S230).

[0078] The sintered product, which has been pulverized through a sintering process and a crushing / classification / de-ironing process, may be mixed with water to remove impurities. As the sintered product is mixed with water, impurities containing residual lithium in the cathode material are washed away by the water. This disclosure refers to this as an aqueous cathode material solution. Once the washing process is complete, a portion of the washing solution is separated from the aqueous cathode material solution through a filter, and a drying process may be performed to remove any remaining moisture from the sintered product.

[0079] The fired products from which impurities have been removed through washing and drying processes can undergo a mixing process in which the fired products are mixed (S240).

[0080] The fired products from which impurities have been removed can be classified by size. As described above, the present disclosure may refer to fired products with large particles as large-particle fired products or large-particle products, and fired products with small particles as small-particle fired products or small-particle products. Each large-particle product and each small-particle product may have physical or chemical properties, and in a mixing process, the large-particle and small-particle products may be mixed according to a certain ratio based on such properties.

[0081] Once the mixing process is completed, a coating and heat treatment process is performed on the mixed large and small particles (S250).

[0082] The coating is intended to suppress the reaction between the cathode material and the electrolyte, and the material used for the coating may include at least one of Ba, Ce, F, P, S, Zr, B, W, Mo, Cr, Nb, Mg, and Hf. The aforementioned material is a metal element, and instead of using the metal element itself in the coating, it may be used in the form of sulfides, hydroxides, oxides, acetates, sulfates, etc.

[0083] Heat treatment is performed to bond the cathode active material and the coating material and can be carried out at a temperature of 200°C to 400°C. However, the heat treatment temperature of 200°C to 400°C is merely an example and can be carried out at various temperatures depending on the given environment.

[0084] When coating and heat treatment are performed, a packaging process is performed, and when packaging is completed, an anode material is produced (S260).

[0085] The generated cathode material can be used as a component of a secondary battery.

[0086] The manufacturing process of the aforementioned cathode material is not limited to the aforementioned process and may further include a process for analyzing the quality of the produced cathode material.

[0087] The present disclosure proposes a method for predicting quality values ​​for each combination and calculating suitability in order to find an optimal combination in which large and small particle sizes from which impurities have been removed are respectively matched so that all matched combinations do not degrade the quality of the cathode material.

[0088]

[0089] FIG. 3 is a diagram for schematically explaining the process of mixing a sintered product according to one embodiment.

[0090] Referring to FIG. 3, the material mixing device of the present disclosure can perform a mixing process by matching the sintered products produced through the sintering process, taking into account the quality of the cathode material finally produced.

[0091] For example, the inventory status (300) for each sintered product produced through the sintering process can be stored in a pre-set database. The aforementioned inventory refers to sintered products that have finished the sintering process but have not entered the mixing process.

[0092] Additionally, the products produced through the firing process can be classified. For example, the products can be classified into large-sized products and small-sized products according to their size. Thus, the inventory status (310) for large-sized products can be stored in a pre-set database, and the inventory status (320) for small-sized products can be stored in another database.

[0093] The aforementioned inventory status may include at least one of the code, name, quantity, and characteristic information of the sintered product of large or small particle size generated through the sintering process.

[0094] The material mixing device of the present disclosure can match at least one large particle size and at least one small particle size stored in the aforementioned databases, obtain characteristic information of the large particle size and characteristic information of the small particle size, and predict the quality value of the mixture in which the large particle size and small particle size are mixed through a preset artificial intelligence model (330). The training data used in the aforementioned artificial intelligence model (330) may be information regarding the large particle size and small particle size generated during the actual cathode material production process, and if the training data is insufficient, the characteristic information of the large particle size and small particle size may be input into another artificial intelligence model to generate a sufficient amount of training data, and then input into the artificial intelligence model (330) that predicts the quality value to be used for training. The present disclosure may refer to the artificial intelligence model (330) that predicts the quality value (340) of the mixture as a mixture quality prediction model.

[0095] As described above, the material mixing device of the present disclosure can generate a combination including a plurality of matching combinations in which at least one large particle size and at least one small particle size are matched based on the large particle size inventory status (310) and the small particle size inventory status (320), and can predict a quality value (340) for the mixture by inputting characteristic information for each matching combination into a pre-trained artificial intelligence model (330).

[0096] The material mixing device of the present disclosure can predict a quality value (340) of the mixture for each of a plurality of matching combinations, and the present disclosure may refer to the aforementioned quality value as a heat-treated product quality prediction value (340) between the recalcined products. Accordingly, one combination includes a plurality of matching combinations, and each matching combination may correspond to a quality value predicted through an artificial intelligence model. The present disclosure may store this in a separate database.

[0097] Once the aforementioned combination is generated, at least one candidate combination is selected, and if the selected candidate combination is not determined to be the optimal combination, a new combination is generated, thereby repeating the selection of candidate combinations and the determination of whether the optimal combination is determined.

[0098] When a quality value is calculated for each matching combination, the matching suitability for each matching combination is calculated based on a preset algorithm (350), and the total suitability for the Nth combination can be calculated. In this disclosure, the algorithm used to calculate the matching suitability may be referred to as the matching algorithm (350).

[0099] When the total suitability is calculated, the combination with the highest total suitability among the first to N combinations can be selected as a candidate combination based on a pre-set matching algorithm (350), and by determining whether the candidate combination can be considered as the optimal combination, the optimal combination can be determined or a new combination can be created, and the selection of candidate combinations and the determination of the optimal combination can be performed repeatedly. When the optimal combination is determined, a work order (360) can be generated so that a mixing process is performed based on the matching combination included in the combination.

[0100]

[0101] FIG. 4 is a diagram illustrating the learning process of an artificial intelligence model that predicts the quality of a mixture according to one embodiment.

[0102] Referring to FIG. 4, the material mixing device of the present disclosure can predict the quality of a mixture of large and small particle sizes in advance through an artificial intelligence model based on characteristic information of large and small particle sizes.

[0103] For example, data regarding the precursor and raw materials that serve as raw materials for the cathode material before the calcination process is performed, and data regarding the semi-finished product generated after the calcination and mixing processes are performed, may be stored in a pre-set database. This allows the raw materials input throughout the entire process to be tracked and the relationship between the input raw materials and the generated semi-finished product to be verified. The present disclosure may refer to this as Lot tracking data (400). For example, Lot tracking data (400) may include at least one of the following information: a heat-treated product Lot number, a total input weight, a large-sized calcined product Lot number, a large-sized input weight, a small-sized calcined product Lot number, and a small-sized input weight.

[0104] When the firing process is performed after the crushing / classification / de-ironing process, multiple fired products are generated. Since the purpose of this disclosure is to find the optimal combination between large and small particle sizes, the generated fired products can be classified into large and small particle sizes according to their size, and characteristic information for each can be stored in a separate database. Each generated large particle size can be corresponded to a large particle fired product Lot number, and each small particle size can be corresponded to a small particle fired product Lot number. This disclosure may refer to these as large particle fired product inspection judgment history data (420) and small particle fired product inspection judgment history data (430), respectively. The large particle fired product inspection judgment history data (420) and small particle fired product inspection judgment history data (430) may include at least one of the following: the date the inspection for identifying characteristic information was performed, the sample Lot number, the packaging material, the sample order, the test order, the process line, the inspection item, and the judgment result.

[0105] In addition, the history regarding the quality of the mixture of large and small particle sizes that is actually mixed can be stored in a separate database. Each mixture can be associated with a heat-treated product Lot number. The present disclosure may refer to this as heat-treated product inspection judgment history data (410). The heat-treated product inspection judgment history data (410) may include at least one of the following data: incoming Lot number, incoming quantity, customer name, customer specifications, and judgment specifications.

[0106] The aforementioned lot numbers for heat-treated products, large-particle calcined products, and small-particle calcined products can be used as keys in each database, and through the lot numbers, information on the raw materials of the cathode material, characteristic information on the calcined product generated through the calcination process, quality information on the mixed mixture, information on the semi-finished product generated during the cathode material production process, and information on the produced cathode material can be tracked, managed, and observed.

[0107] The units of the aforementioned characteristic information of the fired product and the quality information of the mixture may vary depending on the type and characteristics of the fired product and the quality of the mixture. For example, the unit of mass information may be g, and the unit of volume information may be ml or cm³. -- 3 It may be possible. The present disclosure may compare or calculate each characteristic information and quality information for quality prediction. Accordingly, a process (440) of normalizing each characteristic information and quality information may be performed.

[0108] For example, the present disclosure may normalize each characteristic information to a value between -1 and 1 based on the mean or standard deviation. For example, as shown in FIG. 4, if there is characteristic information regarding density, and the average density for the large particle size is 2.1 g / ml, and the highest density component included in the large particle size is 2.4 g / ml and the lowest density component is 1.9 g / ml, the average can be set to 0, the highest component to 1, and the lowest component to -1, thereby normalizing the characteristic information for each component to a value between -1 and 1.

[0109] The present disclosure allows for the learning of a heat-treated product quality prediction model, which is a preset artificial intelligence model, using the aforementioned heat-treated product inspection judgment history data, large-sized fired product inspection judgment history data (420), small-sized fired product inspection judgment history data (430), and heat-treated product inspection as learning data (410).

[0110] The training data input to the artificial intelligence model may include information regarding the input ratio of large and small particle sizes. For example, the quality of the mixture may be predicted by setting the mixing ratio of large and small particle sizes to 0.7 to 0.3 (7 to 3) based on mass. However, the mixing ratio is just one example, and the training data learned through the artificial intelligence model may include various information related to the quality of the fired product.

[0111] Normalized large-sized product inspection judgment history data and normalized small-sized product inspection judgment history data are input into a preset artificial neural network, and when the result is output, the error can be calculated through the heat-treated product inspection judgment history data and the MSE (Mean Square Error) loss function (450). The present disclosure proposes a method of training an artificial intelligence model so that the error calculated through the MSE loss function (450) is minimized, in that a model that accurately predicts the quality of a mixture when two characteristic information is input must be used.

[0112] The aforementioned MSE loss function (450) is where La is the loss value, X true is heat-treated product inspection and judgment history data, X large is the quality measurement of large-diameter fired products, X small is a quality measurement of small-grained fired products, f nn is the quality prediction output from the quality prediction artificial neural network, m1 is the sum of the masses of the large-particle-size fired product and the small-particle-size fired product, m sum is from m1 to m n The sum up to, a and b, can be defined by mathematical formula 1, with the mixing ratio of large and small particle sizes as factors.

[0113]

[0114] Therefore, based on the aforementioned artificial neural network and MSE loss function, an artificial intelligence model for predicting the quality of heat-treated products is trained, and the quality of a mixture can be predicted by inputting normalized characteristic information of multiple sintered products into the trained artificial intelligence model.

[0115] For example, if normalized large particle diameter characteristic information and normalized small particle diameter characteristic information are input, a quality value for one matching combination can be calculated, and the material mixing device of the present disclosure may set a condition that the quality value of all matching combinations output through a learned artificial intelligence model in the generation of the Nth combination is greater than or equal to a preset threshold. Alternatively, a condition may be set that the average of the quality values ​​of all matching combinations output through the learned artificial intelligence model is between a preset range of values. For example, if the quality value of all matching combinations output through the learned artificial intelligence model has a value between -1 and 1, the material mixing device of the present disclosure may set a condition that the average of the quality values ​​of all matching combinations is between -0.3 and 0.3. However, the above-mentioned condition is merely an example, and the above-mentioned values ​​-1, 1, -0.3, and 0.3 can be set in various ways as needed.

[0116]

[0117] FIG. 5 is a flowchart for specifically explaining the process of selecting an optimal combination for mixing a sintered product according to one embodiment.

[0118] Referring to FIG. 5, the material mixing device of the present disclosure can generate the Nth combination, select one candidate combination from the first combination to the Nth combination, and determine whether one candidate combination is the optimal combination.

[0119] Specifically, the material mixing device of the present disclosure generates an Nth combination including a plurality of matching combinations to find an optimal combination for mixing between sintered products (S500).

[0120] The Nth combination includes multiple matching combinations, and each matching combination may include one large particle size and one small particle size. However, the number of large and small particle sizes included is merely an example for the convenience of explanation, and the types and quantities of the included large and small particle sizes can be set in various ways as needed.

[0121] The material mixing device of the present disclosure can generate a matching combination by randomly matching one of a plurality of first materials with one of a plurality of second materials in the case of a first combination where the generated combination is the first combination, and from the second combination which is not the first combination generated, it can generate a combination by rematching some of the matching combinations included in the combination selected as a candidate combination in the previous step based on the matching suitability of each matching combination and a preset algorithm.

[0122] The material mixing device of the present disclosure can obtain characteristic information for each of the large particle size and small particle size by querying a preset database, and can calculate quality values ​​for the mixture of sintered products included in each matching combination when the large particle size and small particle size are mixed through a preset artificial intelligence model.

[0123] For example, the material mixing device of the present disclosure can determine and generate the Nth combination, which includes only matching combinations in which the calculated quality value is greater than or equal to a preset value.

[0124] As another example, the quality score of each matching combination can be calculated through an artificial intelligence model trained to minimize loss based on training data reflecting the characteristic information of each material and an MSE loss function.

[0125] When an Nth combination including multiple matching combinations is generated, the material mixing device of the present disclosure calculates the degree of matching suitability for each matching combination and calculates the total degree of suitability of the Nth combination based on each degree of matching suitability (S510).

[0126] The material mixing device of the present disclosure can calculate the degree of matching suitability of each matching combination based on the quality value of each matching combination.

[0127] When the matching suitability of each combination is calculated, the material mixing device of the present disclosure can calculate the total suitability of the Nth combination based on the sum of the matching suitabilitys. In other words, the matching suitability is used to calculate the total suitability of the Nth combination, and the total suitability is used to select candidate combinations that are likely to become the optimal combination among the first combination to the Nth combination.

[0128] When the total suitability of the Nth combination is calculated, the material mixing device of the present disclosure selects one combination from the first combination to the Nth combination as a candidate combination that is likely to be the optimal combination (S520).

[0129] The material mixing device of the present disclosure selects a combination having the highest total fitness among the total fitness of the first combination calculated and the total fitness of the Nth combination as a candidate combination. Accordingly, the candidate combination may be the Nth combination, but one combination from the first combination to the N-1st combination that is not the Nth combination may be selected.

[0130] The material mixing device of the present disclosure can determine the candidate combination as the first combination immediately when the generated combination is the first combination.

[0131] When a candidate combination is selected, the material mixing device of the present disclosure checks whether the number of combination generations is greater than or equal to a first threshold value (S530).

[0132] The material mixing device of the present disclosure can determine whether the number of combinations generated from the first combination to the Nth combination is sufficient by comparing with a first threshold value, and if the number of combinations generated is sufficient, it can trust the matching combination included in the last selected candidate combination and determine the candidate combination as the optimal combination.

[0133] As described above, once the optimal combination is determined, the material mixing device of the present disclosure may stop the matching operation between the sintered products and generate a work instruction to allow the sintered products to be mixed according to the matching combination.

[0134] If N, which is the number of combination generations, is less than a first threshold, the material mixing device of the present disclosure calculates the number of times the most recently selected candidate combination is continuously selected as a candidate combination and checks whether the number is greater than or equal to a second threshold (S540).

[0135] The material mixing device of the present disclosure can determine that a candidate combination is reliable and select it as the optimal combination if it is selected above a second threshold value. This has the advantage of improving work speed by minimizing the computational process through early termination of the optimal combination selection process.

[0136] If the number of combinations generated is less than a first threshold and the number of times a selected candidate combination is selected consecutively is less than a second threshold, the material mixing device of the present disclosure generates the N+1th combination (S550).

[0137] The material mixing device of the present disclosure performs a process of repeatedly generating combinations from the first combination to the Nth combination to ensure that each matching combination included in the combination becomes an optimal combination in order to produce a cathode material having optimal quality and minimize the probability of producing a cathode material of poor quality, and to select candidate combinations by calculating quality values, matching suitability, and total suitability, and to determine whether the candidate combination is an optimal combination. However, if a combination determined to be an optimal combination is not derived even after generating up to the Nth combination, the material mixing device of the present disclosure attempts to derive an optimal combination by repeatedly generating one more combination until a combination determined to be an optimal combination is derived.

[0138] The material mixing device of the present disclosure can generate an N+1 combination based on the last selected candidate combination. A preset algorithm may be used for this purpose, and the preset algorithm may include a genetic algorithm. This is explained in more detail with reference to FIG. 6.

[0139]

[0140] FIG. 6 is a flowchart for explaining the process of generating a new combination when the optimal combination according to one embodiment is not determined.

[0141] Referring to FIG. 6, the material mixing device of the present disclosure can generate an N+1 combination when the optimal combination is not determined by generating combinations from the first to the Nth combination. The N+1 combination can be generated based on the last selected candidate combination.

[0142] Specifically, the material mixing device of the present disclosure calculates the matching suitability of each matching combination included in the candidate combination based on the candidate combination selected through the generation of the Nth combination (S600).

[0143] As described above, the matching fitness of each matching combination can be calculated based on quality values ​​derived through a pre-trained artificial intelligence model, and these quality values ​​can be calculated based on specific information regarding each large and small particle size. Alternatively, if the matching fitness of each matching combination included in the candidate combination has been previously calculated and stored in a pre-configured database, it may be obtained by querying that database.

[0144] When the matching suitability for each matching combination included in the most recently selected candidate combination is calculated or obtained, the material mixing device of the present disclosure selects a preset number of matching combinations with low matching suitability (S610).

[0145] The present disclosure proposes a method in which the most recently selected candidate combination can be regarded as the combination most likely to become the optimal combination, and among the matching combinations included in the candidate combination, only matching combinations with low matching fitness are selected in a predetermined number to perform rematching among them.

[0146] This allows the second combination to be generated as a better combination in terms of matching fitness than the first combination, and similarly, the N+1 combination to be generated as a better combination in terms of matching fitness than the N combination, thereby making the newly generated combination more optimized than the combination generated in the past, and since re-matching is not performed for all matching combinations, it can also be effective in terms of computational efficiency.

[0147] In selecting matching combinations with low matching fitness, the number of selected matching combinations can be up to a preset number.

[0148] When a preset number of matching combinations with low matching suitability are selected, the material mixing device of the present disclosure performs re-matching between the selected matching combinations (S620).

[0149] The material mixing device of the present disclosure can rematch a plurality of matching combinations included in the candidate combination with the highest total suitability among the first to N combinations. However, it may not rematch all matching combinations, but may perform rematching only between matching combinations with low matching suitability.

[0150] The method of performing rematching is explained in detail in Fig. 7.

[0151] When rematching between selected matching combinations is performed, the material mixing device of the present disclosure checks whether a mutation may appear during the rematching process (S630).

[0152] The aforementioned mutations may include cases where, when generating the N+1th combination, the quality score or matching fitness among the rematched combinations is significantly low. In such cases, since rematching may actually be detrimental to finding the optimal combination, if a rematched combination corresponding to such mutations is derived, it is determined that there is an error in the rematching process, and the rematching process is performed again.

[0153] The material mixing device of the present disclosure can perform the aforementioned rematching operation and the operation of determining whether a mutation has appeared using a preset algorithm. The aforementioned preset algorithm may include a genetic algorithm.

[0154] When the rematching process is performed, the material mixing device of the present disclosure generates an N+1 combination including the rematched matching combination and the matching combination that was not selected due to high matching suitability (S650).

[0155] When the N+1 combination of the material mixing device of the present disclosure is generated, it calculates the quality value of the matching combination included in the N+1 combination, calculates the combination suitability based on the calculated quality value, and calculates the total suitability based on the total sum of the combination suitability. When the total suitability is calculated, the material mixing device of the present disclosure selects the combination with the highest total suitability among the first combination to the N+1 combination as a candidate combination, determines the candidate combination as the optimal combination if N+1 is greater than or equal to a preset first threshold value, determines the candidate combination as the optimal combination if it is less than the first threshold value and the number of times the candidate combination is continuously selected as a candidate combination is greater than or equal to a second threshold value, and if it is less than the second threshold value, it additionally generates the N+2 combination through the aforementioned rematching process and repeats the aforementioned process again.

[0156]

[0157] FIG. 7 is a flowchart for illustrating, by example, a method for generating a new combination according to one embodiment.

[0158] Referring to FIG. 7, the material mixing device of the present disclosure may generate the N+1th combination based on the most recently selected candidate combination when the optimal combination is not selected through the generation of the Nth combination. However, in this case, instead of rematching all matching combinations included in the candidate combination, a predetermined number of matching combinations with low matching suitability are set, and rematching is performed between the selected matching combinations.

[0159] For example, the material mixing device of the present disclosure may select one combination based on total suitability (S700). The selected combination is the combination with the highest total suitability, which is the same as the most recently selected candidate combination.

[0160] According to Fig. 7, Xn 1 In this equation, 1 represents the first combination, and Xn represents the matching fitness of the matching combinations included in the first combination. Therefore, ∑Xn 1 represents the total fitness of the first combination. Likewise, ∑Xn 2 represents the total fitness of the second combination, and ∑Xn 3 represents the total fitness of the third combination, and ∑Xn 4 represents the total fitness of the 4th combination. Through a comparison of the total fitness between combinations, the 3rd combination was selected as the combination with the highest total fitness. However, the number of combinations and the details regarding the selected 3rd combination mentioned above are merely examples for the convenience of explanation and are not limited to this, and can be set in various ways as needed.

[0161] When one combination is selected based on total suitability, the material mixing device of the present disclosure selects a preset number of matching combinations based on the matching suitability of each matching combination included in the selected combination (S710).

[0162] When one combination is selected based on the overall fitness in the material mixing device of the present disclosure, the matching fitness of each matching combination included in the selected combination can be calculated or obtained, and only a preset number of matching combinations can be selected in ascending order of the matching fitness.

[0163] For example, referring to FIG. 7, there are a total of 6 matching combinations included in the third combination, and each combination may include 1 large inlet diameter and 1 small inlet diameter. 大1 shown in FIG. 7 means the large inlet diameter included in the first matching combination, and 小1 means the small inlet diameter included in the first matching combination. Similarly, 大2 and 小2 respectively mean the large inlet diameter and the small inlet diameter included in the second matching combination, 大3 and 小3 respectively mean the large inlet diameter and the small inlet diameter included in the third matching combination, 大4 and 小4 respectively mean the large inlet diameter and the small inlet diameter included in the fourth matching combination, 大5 and 小5 respectively mean the large inlet diameter and the small inlet diameter included in the fifth matching combination, and 大6 and 小6 respectively mean the large inlet diameter and the small inlet diameter included in the sixth matching combination. In addition, X1 to X6 respectively mean the matching fitness of the first matching combination to the sixth matching combination.

[0164] The above-mentioned matching fitness may be calculated in the rematching process by the material mixing device, or may be obtained through query when stored in a preset database in the previously calculated information.

[0165] Comparing the magnitudes of the matching fitness from the first matching combination to the sixth matching combination, as shown in FIG. 7, it can be X3 > X4 > X1 > X6 > X2 > X5, and among them, X2, X6, and X5 with low matching fitness can be selected as the rematching targets.

[0166] When the matching combinations to be rematched are selected based on the matching fitness, the material mixing device of the present disclosure can perform rematching between the selected matching combinations (S720).

[0167] The material mixing device of the present disclosure can generate a new matching combination by performing rematching on a selected matching combination according to a preset standard.

[0168] For example, according to FIG. 7, a new matching combination can be created by matching the small particle size of the 5th matching combination with the lowest matching suitability with the large particle size of the 2nd matching combination with the second lowest matching suitability, matching the small particle size of the 2nd matching combination with the second lowest matching suitability with the large particle size of the 6th matching combination with the third lowest matching suitability, and matching the small particle size of the 6th matching combination with the third lowest matching suitability with the large particle size of the 5th matching combination with the lowest matching suitability.

[0169] However, in this case, if the quality value for the matching of the small particle size of the 5th matching combination and the large particle size of the 2nd matching combination is less than a preset 3rd threshold, or if the matching fitness is less than a preset 4th threshold, the matching of the small particle size of the 5th matching combination and the large particle size of the 2nd matching combination is determined to be a mutation, the previously performed rematching is canceled, and rematching is performed again according to another preset criterion.

[0170] The aforementioned method of rematching large and small particle sizes is merely one example, and the rematching method can be set in various ways as needed.

[0171]

[0172] FIG. 8 is a drawing for explaining a method of mixing a sintered product, which is a material of an anode material according to one embodiment.

[0173] Referring to FIG. 8, the material mixing method of the present disclosure includes a combination generating step of generating a N (N is an integer greater than or equal to 1) combination comprising a plurality of matching combinations classified from a plurality of first materials and a plurality of second materials (S800).

[0174] The cathode material used in secondary batteries is manufactured through a process in which a precursor and a raw material are mixed, a process in which the material is heated to a high temperature in a kiln and fired, a process in which impurities contained in the fired product are removed, a process in which the fired product from which impurities have been removed is dried, a process in which the dried fired product is mixed according to type, and a process in which the mixed product is coated.

[0175] In particular, regarding the process of mixing dried sintered products according to type during the manufacturing process of cathode materials, as mentioned above, conventionally the types of sintered products to be mixed were determined based on the judgment of the operator; consequently, there is a problem where the quality of the produced cathode materials is inconsistent and defective products are produced.

[0176] The present disclosure proposes a method that enables optimal mixing matching between fired products by considering the characteristics of each fired product and the characteristics of the mixed fired products.

[0177] For example, the aforementioned first material may be either a small particle size or a large particle size among the sintered products produced after the cathode material sintering process, and if the first material is a small particle size, the second material may be a large particle size, and if the first material is a large particle size, the second material may be a small particle size.

[0178] The particle size of the fired product may vary depending on the type and mixing ratio of the precursor and raw materials. For example, a fired product with a size of 10 micrometers or more may be classified as large particle size, and one with a size of less than 10 micrometers may be classified as small particle size. However, this is merely one example of classifying large and small particle sizes, and the aforementioned preset temperature and the size criteria for distinguishing between large and small particle sizes may be set in various ways as needed.

[0179] As another example, the combination generating device of the present disclosure may be configured such that in the case of a first combination, which is the first combination, the first material and the second material are included one each, and each matching combination may be configured such that the first material and the second material are randomly matched.

[0180] The present disclosure proposes a method for generating a matching combination by randomly matching one of a plurality of first materials with one of a plurality of second materials in the case of a first combination, which is the first combination generated through a combination generation step, and generating a matching combination from the second combination onwards, which is not the first combination, by re-matching some of the matching combinations included in the combination selected as a candidate combination in the previous step based on the matching fitness of each matching combination and a preset algorithm. The aforementioned preset algorithm may include a genetic algorithm.

[0181] The material mixing method of the present disclosure includes a suitability calculation step (S810) which calculates the matching suitability for each matching combination generated in the combination generation step based on the characteristic information of each material, and calculates the total suitability of the Nth combination based on the calculated matching suitability.

[0182] For example, the suitability calculation step of the present disclosure may obtain first characteristic information of a plurality of first materials and second characteristic information of a plurality of second materials, and calculate a quality value for each of a plurality of matching combinations based on the first characteristic information and the second characteristic information.

[0183] For example, the material mixing device of the present disclosure can match one of a plurality of first materials with one of a plurality of second materials and combine each matching pair to create a single combination. In addition, the material mixing device of the present disclosure can obtain characteristic information for each material and, based on the characteristic information, calculate a quality value of a matching combination in which one first material and one second material are mixed.

[0184] The aforementioned characteristic information may include at least one of the particle size, magnetic impurities, and residual lithium concentration of each sintered product. However, this is merely one example representing the characteristics of each material, and various information may be included as needed.

[0185] As another example, the aforementioned quality figures can be calculated through an artificial intelligence model trained to minimize loss based on training data reflecting the characteristic information of each material and an MSE loss function.

[0186] Specifically, the material mixing device of the present disclosure can calculate quality values ​​by inputting characteristic information of each material into an artificial intelligence model and training the artificial intelligence model so that the difference between the output quality information and the actual quality information of the mixture is minimized using an MSE loss function, and through the trained artificial intelligence model, input quality information for two materials into the artificial intelligence model in an actual mixing process to calculate quality values ​​for the mixture of two materials. The aforementioned quality values ​​may refer to real numbers between 0 and 100, or may refer to grades such as high, medium, and low.

[0187] The material mixing device of the present disclosure can be configured to include multiple matching combinations in a single combination. The number of possible cases as matching combinations can be calculated differently depending on the number of each material. For example, if there are 4 first materials and 4 second materials each, and a matching combination is generated by matching any one of the first materials with any one of the second materials, the number of cases for multiple matching combinations can be 24. As another example, if there are 5 first materials and 5 second materials each, and a matching combination is generated by matching any one of the first materials with any one of the second materials, the number of cases for multiple matching combinations can be 120. As such, the number of cases for multiple matching combinations increases as the number of each material increases. The material mixing device of the present disclosure can be configured to include only matching combinations in the combination where the calculated quality value among the various cases for matching combinations is greater than or equal to a preset value.

[0188] As another example, the aforementioned matching fit is calculated based on quality figures, and the aforementioned total fit can be calculated based on the sum of each matching fit.

[0189] The method for mixing the sintered product of the present disclosure includes a candidate combination selection step of selecting a combination having the highest total suitability among the total suitability of the first combination calculated and the total suitability of the Nth combination as a candidate combination (S820).

[0190] The material mixing device of the present disclosure can classify a first material and a second material to generate a first combination including a plurality of matching combinations and calculate the total fitness for the first combination. Additionally, it can generate a second combination and calculate the total fitness for the second combination. In the same manner, it can generate an Nth combination and calculate the total fitness for the Nth combination. Once the total fitness for each combination is calculated, the material mixing device of the present disclosure can select the combination with the highest total fitness among the first combination to the Nth combination as a candidate combination.

[0191] However, the material mixing device of the present disclosure may determine the first combination as a candidate combination immediately if the generated combination is the first combination generated initially.

[0192] When a candidate combination is selected, the material mixing device of the present disclosure can proceed with the optimal combination selection process by determining whether the number of combination generations or the number of candidate combination selections satisfies a preset criterion.

[0193] When a candidate combination is selected, the method for mixing the sintered product of the present disclosure includes a step of determining the optimal combination, in which the number of combination generations or the number of candidate combination selections is determined to satisfy a preset standard, and if the preset standard is satisfied, the candidate combination is determined as the optimal combination, and if the preset standard is not satisfied, the generation of the N+1th combination is requested in the combination generation step (S830).

[0194] The aforementioned preset criteria may include cases where the number of combination generation counts N is greater than or equal to a first threshold. Alternatively, the preset criteria may include cases where N is less than a preset first threshold and candidate combinations are selected consecutively for a preset second threshold.

[0195] The aforementioned first threshold and second threshold are integers greater than or equal to 1 and can be set in various ways as needed.

[0196] The fact that N is greater than or equal to the first threshold and that a specific combination has been selected consecutively as a candidate combination for a second threshold that is pre-set may mean that the last selected candidate combination can be recognized as being composed of optimal matching combinations.

[0197] For example, in cases where the number of combination generations or the number of candidate combination selections does not satisfy a preset criterion, such that N is less than a preset first threshold, or N is greater than or equal to a preset first threshold and candidate combinations are not selected consecutively for a preset second threshold, the optimal combination determination step of the present disclosure may request the combination generation step to generate a new combination, the N+1th combination.

[0198] When the generation of the N+1th combination is requested to the combination generation step through the optimal combination determination step of the present disclosure, the combination generation step may rematch a plurality of matching combinations included in the aforementioned candidate combinations based on the matching fitness of the candidate combinations and a preset algorithm, and generate the N+1th combination based on the rematching result.

[0199] Specifically, the combination generation step selects a predetermined number of matching combinations with low matching fitness among multiple matching combinations included in the candidate combinations, and can rematch the predetermined number of selected matching combinations through a predetermined algorithm.

[0200] The aforementioned preset number is an integer of 2 or more, and can be set in various ways as needed.

[0201] When the N+1th combination is generated, the material mixing device of the present disclosure calculates the total suitability of the N+1th combination in the manner described above, performs a process of selecting candidate combinations, and determines whether they satisfy a preset criterion.

[0202] Even though the N+1th combination has been generated, if the number of combination generations or the number of candidate combination selections does not satisfy a preset criterion, the material mixing device of the present disclosure may generate the N+2th combination and repeat the same process.

[0203] As another example, the material mixing device of the present disclosure can determine a candidate combination as the optimal combination when N is greater than or equal to a preset first threshold, and can also determine a candidate combination as the optimal combination when N is less than a preset first threshold and the candidate combination is selected consecutively by a preset second threshold.

[0204] For example, if a first threshold is set to 10 and a second threshold is selected to 3, and a third combination is selected as a candidate combination when a third combination is generated, and a third combination is selected as a candidate combination when a fourth combination is generated, and a third combination is selected as a candidate combination when a fifth combination is generated, then the material mixing device of the present disclosure can determine that the third combination is the optimal combination, even though it is less than the first threshold because N is 5, and can generate an instruction to mix materials according to the matching combination included in the combination.

[0205] Once the materials are mixed according to each matching combination, a coating process can be performed on each mixed object. After coating, a packaging process can be performed, and the cathode material production process can be completed.

[0206] The aforementioned preset algorithm may include a genetic algorithm.

[0207]

[0208] Through the operation of the aforementioned configurations, the quality of the produced cathode material can be standardized, and errors caused by manual work and the defect rate of the produced cathode material can be reduced.

[0209] The foregoing description is merely an illustrative explanation of the technical concept of the present disclosure, and those skilled in the art to which the present disclosure pertains may make various modifications and variations within the scope of the essential characteristics of the technical concept. Furthermore, since these embodiments are intended to explain, not limit, the scope of the technical concept is not limited by these embodiments. The scope of protection of the present disclosure shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of the present disclosure.

[0210]

[0211] CROSS-REFERENCE TO RELATED APPLICATION

[0212] This patent application claims priority pursuant to Section 119(a) of the U.S. Patent Act (35 USC § 119(a)) to Korean Patent Application No. 10-2024-0142661 filed on October 18, 2024, all of which are incorporated by reference into this patent application. Additionally, this patent application claims priority in countries other than the United States for the same reasons as above, all of which are incorporated by reference into this patent application.

Claims

1. A combination generating unit that generates a N-th combination (N is an integer greater than or equal to 1) including a plurality of matching combinations classified from a plurality of first materials and a plurality of second materials; A suitability calculation unit that calculates the degree of matching suitability for each of the matching combinations generated by the combination generation unit based on the characteristic information of each material, and calculates the total degree of suitability of the Nth combination based on the degree of matching suitability; A candidate combination selection unit that selects as a candidate combination the combination having the highest total fitness among the total fitness of the Nth combinations, starting from the total fitness of the first combination calculated earlier; and A material mixing device comprising an optimal combination determining unit that determines whether a preset standard is satisfied, determines the candidate combination as the optimal combination when N satisfies the preset standard, and requests the combination generating unit to generate the N+1th combination when N does not satisfy the preset standard.

2. In Paragraph 1, The above-mentioned first material is, It is either a small particle size or a large particle size generated after the cathode material firing process, and A material mixing device characterized in that when the first material has a small particle size, the second material has a large particle size, and when the first material has a large particle size, the second material has a small particle size.

3. In Paragraph 1, The above suitability calculation unit is, A material mixing device characterized by acquiring first characteristic information of a plurality of first materials and second characteristic information of a plurality of second materials, and calculating a quality value for each of the plurality of matching combinations based on the first characteristic information and the second characteristic information.

4. In Paragraph 3, The above quality figures are, A material mixing device characterized by being calculated through an artificial intelligence model trained to minimize loss based on training data reflecting characteristic information of each of the above-mentioned materials and a Mean Square Error (MSE) loss function.

5. In Paragraph 3, The above matching fitness is, Calculated based on the above quality figures, The above total goodness of fit is, A material mixing device characterized by being calculated based on the sum of each of the above matching degrees.

6. In Paragraph 1, The above-mentioned preset criteria are, A material mixing device characterized by including at least one of the following cases: the number of combination generation counts N is greater than or equal to a preset first threshold, and N is less than a preset first threshold, and the candidate combinations are selected consecutively by a preset second threshold.

7. In Paragraph 1, The combination generation unit that received a request to generate the above N+1 combination, A material mixing device characterized by rematching the plurality of matching combinations included in the above candidate combinations based on the matching suitability and the above preset algorithm, and generating the N+1th combination based on the rematching result.

8. In Paragraph 7, The combination generation unit that received a request to generate the above N+1 combination, A material mixing device characterized by selecting a predetermined number of matching combinations with low matching suitability among the plurality of matching combinations included in the above candidate combinations, and re-matching the predetermined number of selected matching combinations through the predetermined algorithm.

9. In Paragraph 7, The above-mentioned preset algorithm is, A material mixing device characterized by including a genetic algorithm.

10. A combination generation step for generating an N-th combination (N is an integer greater than or equal to 1) comprising a plurality of matching combinations classified from a plurality of first materials and a plurality of second materials; A suitability calculation step for calculating the degree of matching suitability for each of the matching combinations generated in the combination generation step based on the characteristic information of each material, and calculating the total degree of suitability of the Nth combination based on the degree of matching suitability; A candidate combination selection step of selecting as a candidate combination the combination having the highest total fitness among the total fitness of the Nth combinations, starting from the total fitness of the first combination calculated earlier; and A material mixing method comprising an optimal combination determination step, wherein N determines whether a preset criterion is satisfied, and if N satisfies the preset criterion, determines the candidate combination as the optimal combination, and if N does not satisfy the preset criterion, requests the combination generation step to generate the N+1th combination.

11. In Paragraph 10, The above-mentioned first material is, It is either a small particle size or a large particle size generated after the cathode material firing process, and A method for mixing materials characterized in that when the first material has a small particle size, the second material has a large particle size, and when the first material has a large particle size, the second material has a small particle size.

12. In Paragraph 10, The above suitability calculation step is, A material mixing method characterized by obtaining first characteristic information of a plurality of first materials and second characteristic information of a plurality of second materials, and calculating a quality value for each of the plurality of matching combinations based on the first characteristic information and the second characteristic information.

13. In Paragraph 12, The above quality figures are, A method for mixing materials characterized by being calculated through an artificial intelligence model trained to minimize loss based on training data reflecting characteristic information of each of the above materials and a Mean Square Error (MSE) loss function.

14. In Paragraph 12, The above matching fitness is, Calculated based on the above quality figures, The above total goodness of fit is, A material mixing method characterized by being calculated based on the sum of each of the above matching suitability.

15. In Paragraph 10, The above-mentioned preset criteria are, A material mixing method characterized by including at least one of the following cases: the number of combination generation counts N is greater than or equal to a preset first threshold, and N is less than a preset first threshold, and the candidate combinations are selected consecutively by a preset second threshold.

16. In Paragraph 10, The combination generation step requested to generate the above N+1 combination is, A material mixing method characterized by rematching the plurality of matching combinations included in the above candidate combinations based on the matching suitability and the above preset algorithm, and generating the N+1th combination based on the rematching result.

17. In Paragraph 16, The combination generation step requested to generate the above N+1 combination is, A method for mixing materials characterized by selecting a predetermined number of matching combinations with low matching suitability among the plurality of matching combinations included in the above candidate combinations, and re-matching the predetermined number of selected matching combinations through the predetermined algorithm.

18. In Paragraph 16, The above-mentioned preset algorithm is, A method for mixing materials characterized by including a genetic algorithm.

Images