Low-voltage battery cell sorting device and its operating method

Abstract

A low-voltage battery cell sorting device according to one embodiment disclosed in the present application may include a data acquisition unit that acquires process data sets of battery cells included in each of a plurality of battery trays, a data pre-processing unit that uses the process data sets to generate induction variable data sets including at least two types of induction variables for each of the battery cells and standardizes the induction variable data sets, and a sorting unit that uses an artificial intelligence model to sort out, from the standardized induction variable data sets, battery cells having standardized induction variable data sets that fall outside a specified range as low-voltage battery cells.

Description

[Technical Field]

[0001] The present invention claims the benefit of priority based on Korean Patent Application No. 10-2023-0031451, filed on March 9, 2023, the entire contents of which are incorporated herein by reference. SUMMARY OF THE INVENTION The embodiments disclosed herein relate to a low-voltage battery cell sorting apparatus and a method of operation thereof. [Background technology]

[0002] In recent years, research and development into secondary batteries has been actively pursued. Here, secondary batteries are batteries that can be charged and discharged, and include both conventional Ni / Cd batteries, Ni / MH (metal hydride) batteries, and more recent lithium-ion batteries. Among secondary batteries, lithium-ion batteries have the advantage of having a much higher energy density than conventional Ni / Cd batteries, Ni / MH batteries, and other batteries. Furthermore, because lithium-ion batteries can be manufactured to be compact and lightweight, they are used as power sources for mobile devices. In recent years, their range of use has expanded to include power sources for electric vehicles, and they are attracting attention as a next-generation energy storage medium.

[0003] Defective battery cells may be produced during the battery cell manufacturing process, and the defects may include defects due to low voltage and internal disconnection.

[0004] If a defective battery cell is installed in an electronic product, the battery may explode and cause a fire, which may result in injury and property damage. Therefore, it is necessary to sort out the defective battery cells from the produced battery cells. Summary of the Invention [Problem to be solved by the invention]

[0005] In the conventional battery manufacturing process, a specified number of battery cells are placed in one tray and low-voltage battery cells are detected for each tray. For example, the OCV of the battery cells is measured for each tray, and low-voltage battery cells are sorted out using the average and standard deviation of the dOCV2 (OCV1 - OCV2) value for the battery cells.

[0006] However, since only deviations within a tray are considered and deviations between trays are not considered, there is no standard that can be applied to all trays, and therefore inspection must be performed on each tray, which takes a long time.

[0007] In addition, since the proportion of low-voltage battery cells among all produced battery cells is approximately 1% or less, there is a problem that sorting low-voltage battery cells by tray is inefficient in terms of the inspection time required for sorting.

[0008] The technical problems of the embodiments disclosed in the present application are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the following description. [Means for solving the problem]

[0009] A low-voltage battery cell sorting device according to one embodiment disclosed in the present application may include a data acquisition unit that acquires process data sets of battery cells included in each of a plurality of battery trays, a data pre-processing unit that uses the process data sets to generate induction variable data sets including at least two types of induction variables for each of the battery cells and standardizes the induction variable data sets, and a sorting unit that uses an artificial intelligence model to sort out, from the standardized induction variable data sets, battery cells having standardized induction variable data sets that fall outside a specified range as low-voltage battery cells.

[0010] In one embodiment, the data preprocessor can standardize the derived variable dataset using a mean and standard deviation based on quartiles of the derived variable dataset.

[0011] In one embodiment, the artificial intelligence model is trained with a training data set for normal battery cells, and can learn a distribution pattern for the training data set using a distribution map based on at least two of a plurality of induction variables.

[0012] In one embodiment, the specified range may be a boundary determined based on the distribution pattern, and the sorting unit may sort out a battery cell having a standardized induction variable data set that falls outside the determined boundary as the low-voltage battery cell.

[0013] In one embodiment, the process data set can include an open circuit voltage (OCV) measured for each of the battery cells in chronological order during the activation process.

[0014] An operating method of a low-voltage battery cell sorting device according to one embodiment disclosed herein may include the following operations: acquiring a process data set of battery cells included in each of a plurality of battery trays; generating an induction variable data set using the process data set, the induction variable data set including at least two types of induction variables for each of the battery cells; standardizing the induction variable data set; and using an artificial intelligence model to sort, from the standardized induction variable data set, battery cells having a standardized induction variable data set outside a specified range as low-voltage battery cells.

[0015] In one embodiment, the standardizing operation may standardize the derived variable dataset using a mean and standard deviation based on quartiles of the derived variable dataset.

[0016] In one embodiment, the artificial intelligence model is trained with a training data set for normal battery cells, and can learn a distribution pattern for the training data set using a distribution map based on at least two of a plurality of induction variables.

[0017] In one embodiment, the specified range may be a boundary determined based on the distribution pattern, and the sorting operation may classify battery cells having a standardized derivative variable data set that falls outside the determined boundary as the low-voltage battery cells.

[0018] In one embodiment, the process data set can include an open circuit voltage (OCV) measured for each of the battery cells in chronological order during the activation process. [Effects of the Invention]

[0019] The low-voltage battery cell sorting apparatus and its operating method according to various embodiments disclosed herein can consider variations within a tray by converting a process data set based on multiple induction variables. Furthermore, variations between trays can be considered by standardizing the induction variable data set. In other words, the present invention considers both variations within a tray and variations between trays and can manage the battery cells included in all trays in an integrated manner, thereby reducing the inspection time required to sort low-voltage battery cells.

[0020] This reduces the inspection time required to screen out low-voltage battery cells during the battery cell manufacturing process, allowing for a rapid response to any sudden increase in battery demand. The reduced inspection time also reduces costs.

[0021] The effects of the low-voltage battery cell sorting device and its operating method disclosed herein are not limited to the effects mentioned above, and other effects not mentioned herein will be apparent to those skilled in the art from the disclosure of the present application. [Brief explanation of the drawings]

[0022] [Figure 1] 1 illustrates the problems of the prior art method for screening low voltage battery cells. [Figure 2] FIG. 1 is a block diagram of a low-voltage battery cell sorting device according to an embodiment of the present disclosure. [Figure 3] 1 illustrates a first distribution diagram for data standardized based on a first derived variable and a second derived variable according to an embodiment of the present disclosure. [Figure 4] 10 illustrates a second distribution diagram for data standardized based on a third derived variable and a second derived variable according to an embodiment of the present disclosure. [Figure 5] 1 is a flowchart illustrating a method of operating a low-voltage battery cell sorting apparatus according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0023] Embodiments of the present invention will now be described with reference to the accompanying drawings, although it should be understood that this is not intended to limit the invention to the particular embodiments, but rather includes various modifications, equivalents, and / or alternatives to the embodiments of the present invention.

[0024] The embodiments and terms used in the present application are not intended to limit the technical features described in the present application to a specific embodiment, but should be understood to include various modifications, equivalents, or alternatives of the embodiment. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of the item, unless the relevant context clearly dictates otherwise.

[0025] As used herein, each phrase such as "A or B," "at least one of A and B," "at least one of A or B," "A, B, or C," "at least one of A, B, and C," and "at least one of A, B, or C" may include any one or all possible combinations of the items listed with that phrase. Terms such as "first," "second," "first," "second," "A," "B," "(a)," or "(b)" may be used simply to distinguish one element from another, and do not limit the element in other respects (e.g., importance or order) unless specifically stated to the contrary.

[0026] In this application, when a (e.g., first) component is referred to as being "coupled," "coupled," or "connected" to another (e.g., second) component, with or without the terms "functionally" or "communicatively," this means that the component may be coupled to the other component directly (e.g., by wire or wirelessly) or indirectly (e.g., via a third component).

[0027] The method according to various embodiments disclosed herein may be provided in a computer program product. The computer program product may be traded between a seller and a buyer as a commodity. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., a compact disc read-only memory, CD-ROM) or may be distributed online (e.g., downloaded or uploaded) via an application store or directly between two user devices. In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored or temporarily generated in a machine-readable storage medium such as the memory of a manufacturer's server, an application store server, or an intermediary server.

[0028] According to embodiments disclosed herein, each of the aforementioned components (e.g., modules or programs) may include one or more entities, and some of the entities may be located separately in other components. According to embodiments disclosed herein, one or more of the aforementioned components or operations may be omitted, or one or more other components or operations may be added. Alternatively or additionally, multiple components (e.g., modules or programs) may be integrated into a single component. In this case, the integrated component may perform one or more functions of each of the multiple components in the same or similar manner as those performed by the respective components of the multiple components before the integration. According to embodiments disclosed herein, operations performed by modules, programs, or other components may be performed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be performed in a different order, omitted, or one or more other operations may be added.

[0029] FIG. 1 illustrates the problem of the prior art method for screening out defective battery cells. Referring to FIG. 1 , a first graph 10 shows the dOCV2 of each battery cell included in the first tray. Here, dOCV2 (OCV1-OCV2) may be the difference between OCV1 and OCV2 for each battery cell. Here, OCV1 may be the open circuit voltage (OCV) measured earlier than OCV2. The open circuit voltage may be measured during an activation process. The activation process may be a process for checking the safety of a battery after the battery assembly process. During the activation process, the battery is repeatedly charged or discharged, and the safety of the battery can be tested based on the data measured for the battery.

[0030] A cutoff 100 for the first tray may be set based on the average and standard deviation of dOCV2 for the battery cells included in the first tray. Battery cells below the cutoff 100 for the first tray may be classified as normal battery cells, and battery cells above the cutoff 100 for the first tray may be classified as defective battery cells. Here, a tray may be a box containing a plurality of battery cells, with the produced battery cells being classified into a certain number of batches. For example, if 3,000 battery cells are produced, each of 100 trays may contain 30 battery cells. The cutoff may be one of the reference values ​​for sorting low-voltage battery cells. The cutoff may be a value set based on the average and standard deviation (σ) of dOCV2 for the battery cells for each tray, excluding battery cells suspected of being low-voltage. For example, the cutoff may be a value based on the sum of the average dOCV2 and 4σ.

[0031] Hereinafter, the defective battery cells may be referred to as low-voltage battery cells. However, this is merely a premise for explaining the present invention, and the defective battery cells are not limited to low-voltage battery cells, and may be referred to as battery cells corresponding to various types of defects.

[0032] A second graph 11 shows the dOCV2 of each battery cell included in the second tray. A cutoff 110 for the second tray can be set based on the average and standard deviation of the dOCV2 for the battery cells included in the second tray. Battery cells below the cutoff 110 for the second tray can be classified as normal battery cells, and battery cells above the cutoff 110 for the second tray can be classified as low-voltage battery cells.

[0033] Referring to third graph 13, which is a combination of first graph 10 and second graph 11, data 102 for a battery cell classified as a normal battery cell on first graph 10 is equal to or greater than cutoff 110 for the second tray, so this battery cell can be classified as a low-voltage battery cell. Similarly, data 112 for a battery cell classified as a low-voltage battery cell on second graph 11 is less than cutoff 100 for the first tray, so this battery cell can be classified as a normal battery cell.

[0034] That is, since the average and standard deviation of dOCV2 for the battery cells differ for each tray and the cutoff for determining low voltage may differ, it is necessary to derive an integrated reference value that can be applied without tray classification.

[0035] FIG. 2 is a block diagram of a low-voltage battery cell sorting apparatus 20 according to one embodiment of the present disclosure. Referring to FIG. 2, the low-voltage battery cell sorting device 20 can be connected to a user terminal 24 via wire and / or wireless connection.

[0036] In one embodiment, the connection 23 between the low-voltage battery cell sorting apparatus 20 and the user terminal 24 may be a communication connection via a wired and / or wireless network. In one embodiment, the wired network may be based on a local area network (LAN) communication or a power line communication. In one embodiment, the wireless network may be based on a local area network (e.g., Bluetooth, WiFi (wireless fidelity), or IrDA (infrared data association)) or a wide area network (e.g., a cellular network, a 4G network, or a 5G network).

[0037] In another embodiment, the connection 23 between the low-voltage battery cell sorting apparatus 20 and the user terminal 24 may be a connection via an inter-device communication method (e.g., a bus, a general purpose input and output (GPIO), a serial peripheral interface (SPI), or a mobile industry processor interface (MIPI)).

[0038] In one embodiment, each of the plurality of trays 221, 223, 225 can contain a plurality of fabricated battery cells. In one embodiment, the user terminal 24 may be a mobile device (eg, a mobile phone, a laptop computer, a smart phone, a smart pad) or a PC (personal computer) that controls the low-voltage battery cell sorting apparatus 20.

[0039] In one embodiment, the low-voltage battery cell sorting apparatus 20 may include a communication circuit 200, a sensor 220, a memory 240, and a processor 260. According to an embodiment, the low-voltage battery cell sorting apparatus 20 shown in FIG. 2 may further include at least one component (e.g., a display, an input device, or an output device) other than the components shown in FIG.

[0040] In one embodiment, the communication circuit 200 can establish a wired and / or wireless communication channel between the low-voltage battery cell sorting device 20 and the user terminal 24, and transmit and receive data to and from the user terminal 24 via the established communication channel.

[0041] In one embodiment, the sensor 220 can acquire process data sets for battery cells included in the multiple trays 221, 223, and 225. In one embodiment, the process data sets can include data on the open circuit voltages (OCVs) of the battery cells. Here, the process data sets can include the open circuit voltages (OCVs) measured for each battery cell in the order of the activation process. For example, the process data sets can include a first measured OCV (OCV1), a second measured OCV (OCV2), and a third measured OCV (OCV3) for the battery cell, where each order can be distinguished in the order of the activation process.

[0042] In one embodiment, memory 240 may include volatile memory and / or non-volatile memory. In one embodiment, the memory 240 can store data used by at least one component (e.g., the processor 260) of the low-voltage battery cell sorting apparatus 20. For example, the data can include software (or associated instructions), input data, or output data. In one embodiment, the instructions, when executed by the processor 260, can cause the low-voltage battery cell sorting apparatus 20 to perform the operation defined by the instructions.

[0043] In one embodiment, the memory 240 may include one or more pieces of software (eg, a data acquisition unit 242, a data preparation unit 244, a selection unit 246, and a model training unit 248).

[0044] In one embodiment, processor 260 may include a central processing unit, an application processor, a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor.

[0045] In one embodiment, the processor 260 can execute software (e.g., the data acquisition unit 242, the data preprocessing unit 244, the sorting unit 246, and the model learning unit 248) and control at least one other component (e.g., a hardware or software component) of the low-voltage battery cell sorting apparatus 20 coupled to the processor 260, and can perform various data processing or calculations.

[0046] Hereinafter, with reference to Figures 2 to 5, a method will be described in which the low-voltage battery cell sorting device 20 sorts low-voltage battery cells from produced battery cells via the data acquisition unit 242, the data pre-processing unit 244, the sorting unit 246, and the model learning unit 248.

[0047] The data acquisition unit 242 can acquire process data sets for the plurality of battery trays. The data acquisition unit 242 can acquire process data sets for the battery cells included in the plurality of battery trays via the sensor 220.

[0048] The data preprocessor 244 can preprocess raw data. The data preprocessor 244 can preprocess process data sets.

[0049] The data preprocessing unit 244 can preprocess the process data set by converting it into an induction variable data set. The data preprocessing unit 244 can convert the process data set into induction variable data for each of a plurality of induction variables using the OCV values ​​of the process data set. Here, the induction variable data set can include induction variable data for each of the plurality of induction variables. For example, the data preprocessing unit 244 can generate an induction variable data set including induction variable data obtained by converting the process data set into the form of each of three induction variables using the OCV values ​​of a battery cell. The data preprocessing unit 244 can further preprocess the process data set by standardizing the induction variable data set based on the process data set. The data preprocessing unit 244 can standardize each of the induction variable data included in the induction variable data set to generate a standardized induction variable data set.

[0050] In one embodiment, the data pre-processing unit 244 can convert the OCV value included in the process data set of the battery cell into a form corresponding to each of a plurality of induction variables. The data pre-processing unit 244 can generate induction variable data by converting the OCV value included in the process data set of the battery cell into a form corresponding to each of a plurality of induction variables. Here, the induction variables may be variables based on the OCV value of the battery cell. For example, the number of induction variables may be three, and the three induction variables may be expressed based on Equations 1 to 3.

[0051] [Formula 1] First induction variable = (dOCV2-μ) / σ

[0052] Here, dOCV2 may be the difference between OCV1 and OCV2 measured for the battery cell, μ may be the average dOCV2 of the battery cells for each tray, and σ may be the standard deviation of dOCV2 of the battery cells for each tray.

[0053] [Formula 2] Second induction variable = dOCV2-cutoff

[0054] Here, the cutoff may be one of the criteria used in the past for sorting low-voltage battery cells. The cutoff may be a value set based on the average and standard deviation of dOCV2 of battery cells for each tray, excluding battery cells suspected of being low-voltage. For example, the cutoff may correspond to the sum of the average dOCV2 and 4σ.

[0055] [Formula 3] Third induction variable = dOCV2 / OCV1

[0056] In one embodiment, the data preprocessing unit 244 can standardize the induced variable dataset using the mean and standard deviation based on the quartiles of the induced variable dataset. The data preprocessing unit 244 can calculate the mean and standard deviation of the induced variable dataset within the first quartile Q1 to the third quartile Q3 based on the IQR for the induced variable dataset, and standardize the induced variable dataset using the calculated mean and standard deviation. Here, the IQR serves as an interquartile range and can indicate the degree of distribution of data excluding the top 25% and bottom 25% of data for each of the multiple induced variables when the induced variable datasets for all battery cells are sorted in ascending order. The induced variable dataset can be standardized based on Equation 4 (Mathematical Expression 1).

[0057] [Formula 4]

number

[0058] Here, X may be at least one induced variable data included in the induced variable dataset, Q1 may be induced variable data corresponding to the first quartile, and Q3 may be induced variable data corresponding to the third quartile.

number

[0059] The reason why the data pre-processing unit 244 performs standardization is that a data set that is distributed at both ends of all data sets may be a data set corresponding to a defective battery cell due to a cause other than low voltage. By the data pre-processing unit 244 eliminating the data sets that are distributed at both ends of all data sets and performing standardization, the present invention can ultimately improve the accuracy of sorting out low-voltage battery cells.

[0060] The data preprocessing unit 244 can reflect deviations within a tray by converting the values ​​of the process data set into an induction variable data set based on induction variables, and can also reflect deviations between trays by standardizing the induction variable data set. As a result, the present invention can integrate and manage data with different distribution patterns for each tray into standardized data.

[0061] Below, we will explain a method for training an artificial intelligence model for sorting low-voltage battery cells based on preprocessed data, and a method for sorting low-voltage battery cells using a pre-trained artificial intelligence model.

[0062] How artificial intelligence models are trained The model learning unit 248 may use a training data set to train the artificial intelligence model. Here, the training data set may be a process data set for normal battery cells, an induced variable data set for normal batteries, or a standardized induced variable data set for normal batteries.

[0063] The model learning unit 248 can learn a distribution pattern for normal battery cells using the training data set. For example, the model learning unit 248 can learn a distribution pattern for normal battery cells based on the distribution of the training data set by the first induction variable and the second induction variable.

[0064] How to sort low-voltage battery cells The sorting unit 246 can sort out low-voltage battery cells using a pre-trained artificial intelligence model. The sorting unit 246 can sort out low-voltage battery cells using an artificial intelligence model trained by the model learning unit 248.

[0065] In one embodiment, the screening unit 246 can screen battery cells having a standardized process data set that falls outside a specified range from the learned distribution pattern as low-voltage battery cells.

[0066] In one embodiment, the sorting unit 246 may determine a boundary between normal battery cells and low-voltage battery cells. The sorting unit 246 may determine the boundary between normal battery cells and low-voltage battery cells based on the learned distribution pattern. The sorting unit 246 may classify battery cells having a standardized process data set that is not included in the determined boundary as low-voltage battery cells. Here, the determined boundary may be a boundary value of a region based on the distribution pattern for normal battery cells. Using the determined boundary as a reference, a region inside the determined boundary may be determined as a normal battery cell, and a region outside the determined boundary may be determined as a low-voltage battery cell.

[0067] FIG. 3 illustrates a first distribution diagram 30 for data standardized based on a first derived variable and a second derived variable according to one embodiment of the present disclosure. Referring to FIG. 3, a first distribution chart 30 for normalized data based on the first and second induction variables can show the distribution of normalized induction variable data sets for a plurality of battery cells.

[0068] The model training unit 248 can learn a distribution pattern for normal battery cells based on the distribution of the standardized induced variable data. In one embodiment, the model training unit 248 can cause the artificial intelligence model to learn a distribution pattern for normal battery cells based on the data density on the first distribution diagram 30. In another embodiment, the model training unit 248 can cause the artificial intelligence model to learn a distribution pattern for normal battery cells based on the mean and standard deviation of the standardized induced variable data set.

[0069] The selection unit 246 can select battery cells corresponding to data that falls outside a range designated in the distribution pattern derived by the model learning unit 248 as defective battery cells due to low voltage.

[0070] A determined boundary 300 may be further shown on the first distribution diagram 30. Here, the determined boundary 300 may refer to a reference boundary value that can determine whether a battery cell is normal or not based on the learned distribution pattern.

[0071] The sorting unit 246 may sort out battery cells corresponding to a standardized induction variable data set that is not included in the determined boundary 300 as defective battery cells due to low voltage. For example, the sorting unit 246 may use the determined boundary 300 as a reference and sort out battery cells having a standardized induction variable data set that is distributed within the determined boundary 300 as normal battery cells.

[0072] FIG. 4 illustrates a second distribution plot 40 for data standardized based on the third derived variable and the second derived variable according to one embodiment of the present disclosure. Referring to FIG. 4, a second distribution chart 40 for normalized data based on the third and second induction variables can show the distribution of normalized induction variable data sets for each of a plurality of battery cells.

[0073] Similar to the above-described FIG. 3, the model learning unit 248 can learn a distribution pattern for normal battery cells based on the distribution of the standardized induction variable data set.

[0074] The sorting unit 246 can sort out battery cells having standardized induction variable data sets that fall outside a specified range from the distribution pattern derived by the model learning unit 248 as defective battery cells due to low voltage.

[0075] A determined boundary 400 may further be shown on the second distribution diagram 40. The sorting unit 246 may sort out battery cells having a standardized induction variable data set that is not included in the determined boundary 400 as defective battery cells due to low voltage. For example, the sorting unit 246 may classify battery cells having a standardized induction variable data set that is distributed within the determined boundary 400 as normal battery cells using the determined boundary 400 as a reference.

[0076] Although the first distribution map 30 and the second distribution map 40 described above in Figures 3 and 4 relate to planar distribution maps based on two induction variables, this is for illustrative purposes only, and the present invention is not limited thereto. The present invention may also use a spatial distribution map based on three induction variables to sort low-voltage battery cells.

[0077] FIG. 5 is a flowchart illustrating a method of operation of the low-voltage battery cell sorting apparatus 20 according to one embodiment of the present disclosure. 5, in operation 500, the data acquisition unit 242 can acquire process data sets for the plurality of battery trays. The data acquisition unit 242 can acquire process data sets for the battery cells included in the plurality of battery trays via the sensors 220.

[0078] In operation 502, the data pre-processor 244 may pre-process the raw data. The data pre-processor 244 may pre-process the process data set.

[0079] The data preprocessing unit 244 can preprocess the process data set by converting the process data set into an induction variable data set. The data preprocessing unit 244 can convert the process data set into induction variable data for each of a plurality of induction variables using the OCV values ​​of the process data set.

[0080] In one embodiment, the data pre-processing unit 244 may convert the OCV value included in the process data set of the battery cell into a form corresponding to each of a plurality of derived variables. The data pre-processing unit 244 may generate derived variable data by converting the OCV value included in the process data set of the battery cell into a form corresponding to each of a plurality of derived variables.

[0081] In one embodiment, the data preprocessing unit 244 can standardize the derived variable dataset using a mean and standard deviation based on the quartiles of the derived variable dataset. The data preprocessing unit 244 can calculate the mean and standard deviation of the derived variable dataset that falls within the first quartile Q1 to the third quartile Q3 based on the IQR for the derived variable dataset, and can standardize the derived variable dataset using the calculated mean and standard deviation.

[0082] In operation 504, the sorting unit 246 can sort out low-voltage battery cells using a pre-trained artificial intelligence model. The sorting unit 246 can sort out low-voltage battery cells using an artificial intelligence model trained by the model learning unit 248.

[0083] In one embodiment, the screening unit 246 can screen battery cells having a standardized process data set that falls outside a specified range from the learned distribution pattern as low-voltage battery cells.

[0084] In one embodiment, the sorting unit 246 can determine a boundary between a normal battery cell and a low-voltage battery cell. The sorting unit 246 can determine a boundary between a normal battery cell and a low-voltage battery cell based on the learned distribution pattern. The sorting unit 246 can classify battery cells having a standardized process data set that is not included in the determined boundary as low-voltage battery cells.

[0085] As used above, terms such as "comprise," "constitute," or "have" mean that the relevant element can be contained within the term, unless otherwise specified, and should be interpreted as meaning that other elements may be included, rather than excluding other elements. All terms, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments disclosed herein belong, unless otherwise defined. Commonly used terms, such as dictionary-defined terms, should be interpreted in a manner consistent with the context of the relevant art, and should not be interpreted in an idealized or overly formal sense unless expressly defined in this application.

[0086] The above description is merely an illustrative example of the technical concepts disclosed in the present application, and various modifications and variations are possible within the scope of the essential characteristics of the embodiments disclosed in the present application, provided that such modifications and variations are not departing from the scope of the essential characteristics of the embodiments disclosed in the present application. Therefore, the embodiments disclosed in the present application are intended to illustrate, rather than limit, the technical concepts of the embodiments disclosed in the present application, and such embodiments do not limit the scope of the technical concepts disclosed in the present application. The scope of protection of the technical concepts disclosed in the present application is to be interpreted by the scope of the claims below, and all technical concepts within the scope equivalent thereto are to be interpreted as being within the scope of the present application.

Claims

1. a data acquisition unit that acquires a process data set of a plurality of battery cells included in each of a plurality of battery trays; a data preprocessing unit that uses the process data set to generate an induction variable data set including at least two types of induction variables for each of the battery cells and standardizes the induction variable data set; a sorting unit that uses an artificial intelligence model to sort out, as a low-voltage battery cell, a battery cell having a standardized induction variable data set that is outside a specified range from among the standardized induction variable data sets.

2. The low-voltage battery cell sorting apparatus of claim 1 , wherein the data preprocessing unit standardizes the derived variable data set using a mean and standard deviation based on quartiles of the derived variable data set.

3. The artificial intelligence model is trained on a training dataset of normal battery cells; The low-voltage battery cell sorting device according to claim 1 , wherein a distribution pattern for the training data set is learned using a distribution map based on at least two types of induction variables among a plurality of induction variables.

4. the specified range is a boundary determined based on the distribution pattern, The low-voltage battery cell sorting device according to claim 3 , wherein the sorting unit sorts out, as the low-voltage battery cell, a battery cell having a standardized induction variable data set that falls outside the determined boundary.

5. 2. The low-voltage battery cell sorting device of claim 1, wherein the process data set includes an open circuit voltage (OCV) measured for each of the battery cells in chronological order of an activation process.

6. A method of operating a low voltage battery cell sorting device, comprising: acquiring a process data set for battery cells contained in each of a plurality of battery trays; generating an induction variable data set including at least two types of induction variables for each of the battery cells using the process data set; standardizing the derived variable data set; and using an artificial intelligence model to classify battery cells having standardized induction variable data sets outside a specified range from among the standardized induction variable data sets as low-voltage battery cells.

7. The method of claim 6 , wherein the act of standardizing standardizes the derived variable data set using a mean and standard deviation based on quartiles of the derived variable data set.

8. The artificial intelligence model is trained on a training dataset of normal battery cells; The method of claim 6 , further comprising learning a distribution pattern for the training data set using a distribution map based on at least two of the plurality of induction variables.

9. the specified range is a boundary determined based on the distribution pattern, The method of claim 8 , wherein the act of sorting comprises sorting out as the low-voltage battery cell a battery cell having a standardized induction variable data set that falls outside the determined boundary.

10. The method of claim 6 , wherein the process data set includes an open circuit voltage (OCV) measured for each of the battery cells in chronological order during an activation process.

Images