Method and electronic device for predicting open circuit voltage according to degradation

Abstract

An electronic device according to an embodiment of the present disclosure may be configured to: input, to an encoder, at least one piece of first CCV data corresponding to a first charge cycle of a battery cell; acquire, from the encoder, first latent vector data related to the at least one piece of first CCV data; input the first latent vector data and the at least one piece of first CCV data to a trained deep learning model; acquire, from the deep learning model, second latent vector data related to OCV data corresponding to the at least one piece of first CCV data; input the second latent vector data to a decoder; and acquire, from the decoder, OCV data according to an SoC.

Description

Method for predicting open circuit voltage due to degradation and electronic device

[0001] The present disclosure relates to a method and electronic device for predicting the Open Circuit Voltage (OCV)-SoC relationship according to degradation, and specifically, to a method and electronic device for predicting the Open Circuit Voltage (OCV)-SoC relationship according to degradation through Closed Circuit Voltage (CCV) data based on a deep learning model.

[0002] This application claims the benefit of priority based on Korean Patent Application No. 10-2024-0144925 dated October 22, 2024, and all contents disclosed in the document of said Korean Patent Application are incorporated herein as part of this specification.

[0003] The OCV-SoC relationship is an inherent characteristic of battery cells, and identical battery cells exhibit the same OCV-SoC relationship. Conventional NCM (nickel, cobalt, manganese) battery cells have the characteristic that the OCV-SoC relationship does not change with degradation, and most BMS (battery management system) control algorithms are established based on NCM battery cells. On the other hand, Mn-rich battery cells exhibit a characteristic where the OCV-SoC relationship changes with degradation, making it difficult to estimate the SoC based on existing BMS control algorithms.

[0004] Conventional deep learning models for estimating the OCV-SoC relationship are based on NCM battery cells where the change in OCV due to degradation is not significant, and there is a problem that model performance may be degraded by real-time SoC estimation errors when using SoC as input data.

[0005] Therefore, a new method is needed to estimate the OCV-SoC relationship.

[0006] According to the disclosed embodiments, the OCV-SoC relationship due to degeneration can be predicted through CCV data based on a deep learning model.

[0007] The technical problems to be solved by the embodiments of the present disclosure are not limited to those described above, and other technical problems can be inferred from the following embodiments.

[0008] An electronic device according to one embodiment of the present disclosure comprises one or more processors; and one or more memories for storing one or more instructions, wherein the one or more processors may be configured to input at least one first closed circuit voltage (CCV) data corresponding to a first charging cycle of a battery cell into an encoder by executing one or more instructions, obtain a first latent vector data associated with at least one first CCV data from the encoder, input the first latent vector data and at least one first CCV data into a learned deep learning model, obtain a second latent vector data associated with OCV data corresponding to at least one first CCV data from the deep learning model, input the second latent vector data into a decoder, and obtain OCV data according to the SoC from the decoder.

[0009] In an electronic device according to one embodiment of the present disclosure, the deep learning model may be a deep learning model configured to learn the correlation between the input data set and the output data set by inputting a plurality of CCV data sets corresponding to each of a plurality of charging cycles of a battery cell and a plurality of CCV data sets related to a plurality of CCV data obtained by inputting the plurality of CCV data sets to an encoder, and a plurality of OCV data sets related to a plurality of CCV data as an output data set.

[0010] In an electronic device according to one embodiment of the present disclosure, the encoder is an encoder of an auto encoder learned by using OCV data according to the SoC as input and output, and the decoder may be a decoder of the auto encoder.

[0011] In an electronic device according to one embodiment of the present disclosure, the battery cell may be a battery cell comprising a lithium-rich manganese oxide as a positive active material.

[0012] An electronic device according to one embodiment of the present disclosure may have at least one first CCV data corresponding to a charging start CCV (SV), a charging intermediate CCV (MV), and a charging stop CCV (EV) of a battery cell.

[0013] In an electronic device according to one embodiment of the present disclosure, the autoencoder is based on a 1D Convolutional Neural Network layer and can be trained to minimize the MSE error between OCV data according to the input SoC and OCV data according to the output SoC.

[0014] A method for predicting OCV-SoC according to degradation performed by an electronic device according to one embodiment of the present disclosure may include: inputting at least one first CCV data corresponding to a first charging cycle of a battery cell into an encoder; obtaining a first latent vector data associated with at least one first CCV data from the encoder; inputting the first latent vector data and at least one first CCV data into a learned deep learning model; obtaining a second latent vector data associated with OCV data corresponding to at least one first CCV data from the deep learning model; inputting the second latent vector data into a decoder; and obtaining OCV data according to SoC from the decoder.

[0015] In a method for predicting OCV-SoC according to degradation performed by an electronic device according to one embodiment of the present disclosure, the deep learning model may be a deep learning model configured to learn the correlation between the input data set and the output data set by using a plurality of CCV data sets corresponding to each of a plurality of charging cycles of a battery cell and a plurality of CCV data sets obtained by inputting the plurality of CCV data sets to an encoder as an input data set, and using a plurality of OCV data sets corresponding to the plurality of CCV data sets as an output data set.

[0016] In a method for predicting OCV-SoC according to degradation performed by an electronic device according to one embodiment of the present disclosure, the encoder is an encoder of an autoencoder learned with OCV data according to SoC as input and output, and the decoder may be a decoder of an autoencoder.

[0017] In a method for predicting OCV-SoC according to degradation performed by an electronic device according to one embodiment of the present disclosure, the battery cell may be a battery cell comprising a lithium-rich manganese oxide as a positive active material.

[0018] In a method for predicting OCV-SoC based on degradation performed by an electronic device according to one embodiment of the present disclosure, at least one first CCV data may correspond to a charging start CCV (SV), a charging intermediate CCV (MV), and a charging stop CCV (EV) of a battery cell.

[0019] In a method for predicting OCV-SoC based on degeneration performed by an electronic device according to one embodiment of the present disclosure, the autoencoder is based on a one-dimensional convolutional neural network layer and can be trained to minimize the MSE error between OCV data based on an input SoC and OCV data based on an output SoC.

[0020] In one embodiment of the present disclosure, an electronic device has a computer-readable, non-transient computer-readable storage medium having a program stored on it for executing a method for predicting OCV-SoC according to degradation on a computer, wherein the method for predicting OCV-SoC according to degradation may be characterized by comprising: a step of inputting at least one first CCV data corresponding to a first charging cycle of a battery cell into an encoder; a step of obtaining a first latent vector data associated with at least one first CCV data from the encoder; a step of inputting the first latent vector data and at least one first CCV data into a learned deep learning model; a step of obtaining a second latent vector data associated with OCV data corresponding to at least one first CCV data from the deep learning model; a step of inputting the second latent vector data into a decoder; and a step of obtaining OCV data according to SoC from the decoder.

[0021] According to the embodiments disclosed in this document, the SoC / SoHC estimation algorithm of the BMS used in conventional NCM battery cells can be used, and it can operate robustly against SoC estimation errors.

[0022] The effects of the invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description in the claims.

[0023] FIG. 1 is a block diagram of an electronic device according to one embodiment of the present disclosure.

[0024] FIG. 2 is a drawing showing an autoencoder according to one embodiment of the present disclosure.

[0025] FIG. 3 is a diagram illustrating the operation of a deep learning model according to one embodiment of the present disclosure.

[0026] FIG. 4 is a diagram illustrating the process of generating CCV data according to one embodiment of the present disclosure.

[0027] FIG. 5 is a flowchart of the operation of an electronic device according to a representative embodiment of the present disclosure.

[0028] In describing the embodiments, technical details that are well known in the art to which this disclosure belongs and are not directly related to this disclosure are omitted. This is intended to convey the essence of this disclosure more clearly without obscuring it by omitting unnecessary explanations.

[0029] For the same reason, some components in the attached drawings have been exaggerated, omitted, or schematically depicted. Additionally, the size of each component does not entirely reflect its actual dimensions. Identical or corresponding components in each drawing have been assigned the same reference numbers.

[0030] The advantages and features of the present disclosure and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure is complete and to fully inform those skilled in the art of the scope of the invention, and the present disclosure is defined only by the scope of the claims. Throughout the specification, like reference numerals refer to like components.

[0031] At this time, it will be understood that each block of the process flow diagrams and combinations of the flow diagrams can be executed by computer program instructions. Since these computer program instructions can be loaded into the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing equipment, the instructions executed through the processor of the computer or other programmable data processing equipment create means for performing the functions described in the flow diagram block(s). Since these computer program instructions can also be stored in computer-available or computer-readable memory that can be directed toward the computer or other programmable data processing equipment to implement functions in a specific way, the instructions stored in such computer-available or computer-readable memory can also produce a manufactured item containing means of instruction for performing the functions described in the flow diagram block(s). Since computer program instructions can also be loaded onto a computer or other programmable data processing equipment, the instructions that execute the computer or other programmable data processing equipment by creating a process that is executed by a computer through a series of operation steps performed on the computer or other programmable data processing equipment can also provide steps for executing the functions described in the flow diagram block(s).

[0032] Additionally, each block may represent a module, segment, or part of code containing one or more executable instructions for executing a specified logical function(s). Also, it should be noted that in some alternative embodiments, the functions mentioned in the blocks may occur out of order. For example, two blocks shown in succession may actually be executed substantially simultaneously, or the blocks may be executed in reverse order according to the corresponding function.

[0033] In this embodiment, the term "part" refers to a software or hardware component, such as an FPGA or ASIC, and the "part" performs certain roles. However, the meaning of "part" is not limited to software or hardware. The "part" may be configured to reside in an addressable storage medium or configured to operate one or more processors. Thus, for example, the "part" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functions provided within the components and "parts" may be combined into a smaller number of components and "parts" or further separated into additional components and "parts." Furthermore, the components and "parts" may be implemented to operate one or more CPUs within a device or secure multimedia card.

[0034] The expression “at least one of a, b, and c” described throughout the specification may include ‘a alone’, ‘b alone’, ‘c alone’, ‘a and b’, ‘a and c’, ‘b and c’, or ‘a, b, and c all’.

[0035] The "terminal" mentioned below may be implemented as a computer or portable terminal capable of connecting to a server or other terminal via a network. Here, the computer includes, for example, a notebook, desktop, or laptop equipped with a web browser, and the portable terminal is a wireless communication device that ensures portability and mobility, and may include all types of handheld-based wireless communication devices such as IMT (International Mobile Telecommunication), CDMA (Code Division Multiple Access), W-CDMA (W-Code Division Multiple Access), LTE (Long Term Evolution), communication-based terminals, smartphones, tablet PCs, etc.

[0036] Embodiments of the present disclosure are described below with reference to the attached drawings so that those skilled in the art can easily implement them. However, the present disclosure may be embodied in various different forms and is not limited to the embodiments described herein.

[0037] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the attached drawings.

[0038] FIG. 1 is a block diagram of an electronic device according to one embodiment of the present disclosure.

[0039] Referring to FIG. 1, the electronic device (100) may include at least one memory (110) and at least one processor (120). According to an embodiment, the electronic device (100) illustrated in FIG. 1 may further include at least one component (e.g., a communication interface) other than the components illustrated in FIG. 1.

[0040] According to one embodiment, the memory (110) may include volatile memory and / or non-volatile memory. According to one embodiment, the memory (110) may store data used by at least one component of the electronic device (100) (e.g., processor (120)). For example, the data may include software (or, related instructions), input data, or output data. In one embodiment, the instructions may cause the electronic device (100) to perform operations defined by the instructions when executed by the processor (120).

[0041] According to one embodiment, the processor (120) may be implemented as a computer or a similar device according to hardware, software, or a combination thereof. In hardware, the processor (120) may be implemented in the form of an electronic circuit that processes electrical signals to perform control functions, and in software, it may be implemented in the form of a program that drives the hardware processor (120). According to one embodiment, the processor (120) may be operatively connected to a component (e.g., memory (110)) included in the electronic device (100) to control the connected component.

[0042] An electronic device (100) according to one embodiment may further include a communication interface (not shown). The communication interface may establish a wired or wireless communication channel with an external device (e.g., an OBD (on-board diagnostics) device of a vehicle, a cloud server, a charger / discharger) and transmit and receive various data with the external device. The communication circuit of the electronic device (100) may include at least one communication port for being connected by a wired cable to communicate with the external device via a wire. The communication circuit of the electronic device (100) may include a cellular communication module and be configured to be connected to a cellular network (e.g., 3G, LTE, 5G, Wibro, or Wimax). According to one embodiment, the communication circuit of the electronic device (100) may include a short-range communication module and transmit and receive data with the external device using short-range communication (e.g., Wi-Fi, Bluetooth, Bluetooth Low Energy (BLE), UWB), but is not limited thereto.

[0043] Meanwhile, unless otherwise specifically mentioned in the following description, the operation of the electronic device (100) may be interpreted as being performed under the control of the processor (120).

[0044] According to one embodiment, a processor (120) may input at least one first CCV data corresponding to a first charging cycle of a battery cell into an encoder. The processor (120) may acquire first latent vector data associated with at least one first CCV data from the encoder and input the first latent vector data and at least one first CCV data into a learned deep learning model. The processor (120) may acquire second latent vector data associated with OCV data corresponding to at least one first CCV data from the deep learning model. The processor (120) may be configured to input the second latent vector data into a decoder and acquire OCV data according to the SoC from the decoder. FIG. 2 is a diagram showing an autoencoder according to one embodiment of the present disclosure. Descriptions that overlap with the foregoing may be omitted and may be explained using the configurations of FIG. 1.

[0045] Referring to FIG. 2, an autoencoder is an unsupervised learning technique that converts input data into a signal through an encoder (210) and then generates output data, such as a label, through a decoder (220). It may have a structure that includes an encoder (210) that compresses data and a decoder (220) that restores the compressed data. The autoencoder (200) may be an artificial neural network model for obtaining a small-dimensional representation by compressing input data. The compressed data may be called a latent vector. The decoder (220) receives the compressed data and restores it to the original data input to the encoder (210).

[0046] The processor (120) can train the autoencoder (200) so that the input data and the output data are as similar as possible. In the encoder (210), the dimensionality of the data is reduced so that a latent vector representing the input data can be output. In the decoder (220), the input data corresponding to the latent vector can be restored as output data and output. The processor (120) can train the autoencoder (200) in a way that minimizes the mean squared error (MSE) between the input data and the restored output data.

[0047] When the input and output data of the autoencoder (200) are set into multiple OCV data sets, the process of the processor (120) training the autoencoder (200) can be represented as Equations 1 to 3.

[0048] [Mathematical Formula 1]

[0049]

[0050] [Mathematical Formula 2]

[0051]

[0052] [Mathematical Formula 3]

[0053]

[0054] Loss function in Equations 1 to 3 is the part that reflects the MSE error and the part that reflects nonlinearity It may include. is the input OCV data and output OCV data It may include the MSE error calculated for, is an activation function (rectified linear unit) It can include. An activation function is a function that outputs the input value as is if the input value is positive, and outputs 0 if the input value is negative. Is It refers to a hyperparameter for determining the degree of reflection.

[0055] A loss function is a function that measures the difference between the model's predicted value and the actual value. A processor (120) can train an autoencoder (200) by minimizing the MSE error based on the loss function. Additionally, the processor (120) can include an activation function in the loss function to solve the problem of conventional autoencoders where the output value increases along with the restoration error. As a result, the processor (120) can increase the computational efficiency of the autoencoder (200) and reflect non-linearity during the training process.

[0056] The processor (120) can utilize the Adam (adaptive moment estimation) optimization algorithm during the learning process of the autoencoder (200). Additionally, the processor (120) can highlight relationships between features by utilizing a 1D convolutional neural network (1dCNN) layer (211) instead of a fully connected layer as a layer of the autoencoder (200).

[0057] FIG. 3 is a diagram illustrating the operation of a deep learning model according to an embodiment of the present disclosure. Descriptions that overlap with the foregoing may be omitted and may be explained using the configurations of FIG. 1.

[0058] Referring to FIG. 3, a processor (120) can train an autoencoder (200) by inputting a plurality of open circuit voltage (OCV) data sets (310). The processor (120) can train the autoencoder (200) to output a plurality of OCV data sets (350) that are as similar as possible to the input plurality of OCV data sets (310). The plurality of OCV data sets (310) may be a set of OCV data having different SoCs during one charging cycle. For example, the plurality of OCV data sets (310) may correspond to 1,000 OCV data extracted at 0.1% intervals from SoC 0% to 100% per charging cycle. That is, the plurality of OCV data sets (310) may not contain SoC information. These plurality of OCV data sets (310) may be obtained from an OCV-SoC table representing OCV according to SoC. Multiple OCV data sets (310) may refer to data such as OCV data according to the SoC that is input during the process of the processor (120) training the autoencoder.

[0059] A processor (120) according to one embodiment can obtain a latent vector (340) corresponding to a plurality of OCV data sets (310) during the process of training an autoencoder (200). The latent vector (340) may be two-dimensional data that reflects the features of the plurality of OCV data sets (310).

[0060] A processor (120) according to one embodiment may input at least one closed circuit voltage (CCV) data (320) to a learned encoder (210). The at least one CCV data (320) may include a charging start CCV (SV) corresponding to the charging start time for each charging cycle, a charging intermediate CCV (MV) corresponding to the charging intermediate time, and a charging stop CCV (EV) corresponding to the charging stop time. The processor (120) may obtain a latent vector (330) corresponding to at least one CCV data (320) as the output of the encoder (210). The latent vector (330) may be two-dimensional data that reflects the features of at least one CCV data (320). In the case of an actual vehicle, it is not common to charge the SoC from 0% to 100%, and it is common to start charging at a first SoC (e.g., 30%) and end charging at a second SoC (e.g., 80%). Accordingly, at least one CCV data (320) may include a CCV (SV) corresponding to the start time of charging for a specific charging cycle, a CCV (MV) corresponding to the middle time of charging, and a CCV (EV) corresponding to the end time of charging.

[0061] The processor (120) can train a deep learning model (300) by using the result of associating at least one CCV data (320) and a latent vector (330) as input data and the latent vector (340) as output data. Associating at least one CCV data (320) and a latent vector (330) is intended to distinguish which charging cycle the latent vector (330) originated from, since SoC data is not input during the training process. The latent vector (340) obtained as the output of the deep learning model (300) can be restored into a plurality of OCV data sets (350) through a decoder (220).

[0062] In summary, the processor (120) can obtain multiple OCV data sets (350) through at least one CCV data (320) based on the learned deep learning model (300). As described above, since the multiple OCV data sets (350) are a set of OCV data having different SoCs during one charging cycle, the processor (120) can generate an OCV-SoC table using the multiple OCV data sets (350). The multiple OCV data sets (350) may refer to data such as OCV data according to the SoC that is restored and output during the process of training the autoencoder. The process of the processor (120) obtaining an OCV-SoC table through at least one CCV data (320) can be represented as Equations 4 to 7.

[0063] [Mathematical Formula 4]

[0064]

[0065] [Mathematical Formula 5]

[0066]

[0067] [Mathematical Formula 6]

[0068]

[0069] [Mathematical Formula 7]

[0070]

[0071] In mathematical formulas 4 to 7 These represent the charging start CCV, charging middle CCV, and charging stop CCV data, respectively. means a latent vector (330) obtained by inputting at least one CCV data (320) into a learned encoder (210), and means a latent vector (340) obtained by inputting a plurality of OCV data sets (310) into a learned encoder (210). It means a deep learning model (300) trained to take at least one CCV data (320) and a latent vector (330) as inputs and to take a latent vector (340) as outputs. It means an OCV-SoC table configured by combining a plurality of OCV data sets (350) restored by inputting a latent vector (340) into a learned decoder (220) with a corresponding SoC.

[0072] FIG. 4 is a diagram illustrating the process of generating CCV data according to one embodiment of the present disclosure. Descriptions that overlap with the foregoing may be omitted and may be explained using the configurations of FIG. 1.

[0073] Referring to FIG. 4, the processor (120) may extract at least one CCV data (320) to train a deep learning model (300). The at least one CCV data (320) may include an SV (321), an MV (323), and an EV (325) for each cycle. According to one embodiment, to train a deep learning model (300), a set of SV (321) and EV (325) corresponding to an interval of 60% SoC and corresponding to an SoC range of 10% to 95% may be randomly extracted from a plurality of CCV data corresponding to a plurality of charging cycles. SV (321) may be CCV data corresponding to the point where charging started in a specific charging cycle, and EV (325) may be CCV data corresponding to the point where charging stopped in a specific charging cycle. MV (323) may be CCV data corresponding to a point halfway through the charging time to the extracted SV (321) and EV (325).

[0074] FIG. 5 is a flowchart of the operation of an electronic device according to a representative embodiment of the present disclosure.

[0075] Since the operation method of Fig. 5 can be performed by the electronic device (100) of Fig. 1, descriptions that overlap with the above-mentioned content may be omitted and may be explained using the configurations of Fig. 1.

[0076] The embodiment illustrated in FIG. 5 is merely one embodiment, and the order of operations according to various embodiments of the present disclosure may differ from that illustrated in FIG. 5, and some operations illustrated in FIG. 5 may be omitted, the order of operations may be changed, or operations may be merged.

[0077] In step S510, the processor (120) may input at least one first CCV data corresponding to a first charging cycle of the battery cell into an encoder. The encoder may be an encoder (210) of an autoencoder (200). The at least one first CCV data may include, in a specific charging cycle, a charging start CCV corresponding to a charging start time, a charging intermediate CCV corresponding to a charging intermediate time, and a charging stop CCV corresponding to a charging stop time. The battery cell may be, for example, a battery cell (also known as a manganese-rich cell) comprising a lithium-rich manganese oxide as a positive active material.

[0078] In step S520, the processor (120) can obtain first latent vector data associated with at least one first CCV data from the encoder.

[0079] In step S530, the processor (120) may input the first latent vector data and at least one first CCV data into a learned deep learning model. The deep learning model may be a deep learning model constructed by learning the correlation between the input data set and the output data set, using a plurality of CCV data sets corresponding to each of the plurality of charging cycles of the battery cell and a latent vector data set related to the plurality of CCV data obtained by inputting the plurality of CCV data sets into an encoder as the input data set, and a latent vector data set related to the plurality of OCV data corresponding to the plurality of CCV data as the output data set.

[0080] In step S540, the processor (120) can obtain second latent vector data associated with OCV data corresponding to at least one first CCV data from a deep learning model.

[0081] In step S550, the processor (120) can input the second potential vector data into the decoder. The decoder may be the decoder (220) of the autoencoder (200).

[0082] In step S560, the processor (120) can obtain OCV data corresponding to the SoC from the decoder. OCV data corresponding to the SoC may mean multiple OCV data sets corresponding to different SoCs. The processor (120) can create an OCV-SoC table using multiple OCV data sets.

[0083] The electronic device according to the embodiments described above may include a processor, memory for storing and executing program data, permanent storage such as a disk drive, a communication port for communicating with an external device, and user interface devices such as a touch panel, a key, an icon, etc. Methods implemented as software modules or algorithms may be stored on a computer-readable recording medium as computer-readable code or program instructions executable on the processor. Here, computer-readable recording media include magnetic storage media (e.g., ROM (read-only memory), RAM (random-access memory), floppy disks, hard disks, etc.) and optical reading media (e.g., CD-ROM, DVD (digital versatile disc)). The computer-readable recording medium may be distributed across networked computer systems, allowing computer-readable code to be stored and executed in a distributed manner. The medium may be readable by a computer, stored in memory, and executed by a processor.

[0084] Various embodiments of the present disclosure may be represented by functional block configurations and various processing steps. These functional blocks may be implemented by various numbers of hardware and / or software configurations that execute specific functions. For example, the embodiments may employ integrated circuit configurations such as memory, processing, logic, look-up tables, etc., which can execute various functions by the control of one or more microprocessors or other control devices. Similar to how components may be implemented as software programming or software elements, the embodiments may be implemented in programming or scripting languages ​​such as C, C++, Java, assembler, etc., including various algorithms implemented as combinations of data structures, processes, routines, or other programming configurations. Functional aspects may be implemented as algorithms executed on one or more processors. Additionally, the embodiments may employ prior art for electronic configuration, signal processing, and / or data processing. Terms such as “mechanism,” “element,” “means,” and “configuration” may be used broadly and are not limited to mechanical and physical configurations. The above terms may include the meaning of a series of software processes (routines) in conjunction with processors, etc.

[0085] The aforementioned embodiments are merely examples, and other embodiments may be implemented within the scope of the claims set forth below.

Claims

1. In an electronic device, One or more processors; and It includes one or more memories that store one or more instructions, and The above one or more processors, by executing the above one or more instructions, At least one first CCV (closed circuit voltage) data corresponding to the first charging cycle of the battery cell is input to the encoder, and Obtaining first latent vector data associated with at least one first CCV data from the encoder, and The first latent vector data and the at least one first CCV data are input into a trained deep learning model, and From the deep learning model above, second latent vector data related to OCV data corresponding to at least one first CCV data is obtained, and The above second latent vector data is input into the decoder, and An electronic device configured to acquire OCV data according to the SoC from the above decoder.

2. In Paragraph 1, The above deep learning model is, A plurality of CCV data sets corresponding to each of the plurality of charging cycles of the battery cell and a latent vector data set related to the plurality of CCV data obtained by inputting the plurality of CCV data sets into an encoder are used as input data sets, and Using a latent vector dataset associated with a plurality of OCV data corresponding to the plurality of CCV data as an output dataset, An electronic device, which is a deep learning model configured to learn the correlation between the input data set and the output data set.

3. In Paragraph 1, The above encoder is, It is an encoder of an auto encoder trained using OCV data according to the above SoC as input and output, and The above decoder is, An electronic device that is a decoder of the above-mentioned autoencoder.

4. In Paragraph 1, The above battery cell is, An electronic device, which is a battery cell comprising lithium-rich manganese oxide as a positive electrode active material.

5. In Paragraph 1, The above at least one first CCV data is, An electronic device corresponding to the charge start CCV (SV), charge intermediate CCV (MV), and charge stop CCV (EV) of the battery cell.

6. In Paragraph 3, The above autoencoder is, It is based on 1D Convolutional Neural Network layers, and An electronic device trained to minimize the MSE error between OCV data according to the input SoC and OCV data according to the output SoC.

7. In a method for predicting OCV-SoC based on the degradation of an electronic device, A step of inputting at least one first CCV data corresponding to a first charging cycle of a battery cell into an encoder; A step of obtaining first latent vector data associated with at least one first CCV data from the encoder; A step of inputting the first latent vector data and the at least one first CCV data into a trained deep learning model; A step of obtaining second latent vector data associated with OCV data corresponding to at least one first CCV data from the deep learning model; The step of inputting the above second potential vector data into a decoder; and A method for predicting OCV-SoC based on degradation, comprising the step of obtaining OCV data based on SoC from the above decoder.

8. In Paragraph 7, The above deep learning model is, A plurality of CCV data sets corresponding to each of the plurality of charging cycles of the battery cell and a latent vector data set related to the plurality of CCV data obtained by inputting the plurality of CCV data sets into an encoder are used as input data sets, and Using a latent vector dataset associated with a plurality of OCV data corresponding to the plurality of CCV data as an output dataset, A method for predicting OCV-SoC based on degeneration, which is a deep learning model configured to learn the correlation between the input data set and the output data set.

9. In Paragraph 7, The above encoder is, It is an encoder of an autoencoder trained using OCV data according to the above SoC as input and output, and The above decoder is, A method for predicting OCV-SoC based on degeneration, which is a decoder of the above-mentioned autoencoder.

10. In Paragraph 7, The above battery cell is, A method for predicting OCV-SoC according to degradation of a battery cell containing lithium-rich manganese oxide as a positive electrode active material.

11. In Paragraph 7, The above at least one first CCV data is, A method for predicting OCV-SoC based on degradation, corresponding to the charging start CCV (SV), charging intermediate CCV (MV), and charging stop CCV (EV) of the battery cell.

12. In Paragraph 9, The above autoencoder is, It is based on 1D Convolutional Neural Network layers, and A method for predicting OCV-SoC based on degeneration, which is trained to minimize the MSE error between OCV data based on the input SoC and OCV data based on the output SoC.

13. A computer-readable, non-transient computer-readable storage medium having a program stored on it for executing a computer for a method to predict OCV-SoC based on electronic device degradation, The OCV-SoC prediction method based on the above degradation is: A step of inputting at least one first CCV data corresponding to a first charging cycle of a battery cell into an encoder; A step of obtaining first latent vector data associated with at least one first CCV data from the encoder; A step of inputting the first latent vector data and the at least one first CCV data into a trained deep learning model; A step of obtaining second latent vector data associated with OCV data corresponding to at least one first CCV data from the deep learning model; The step of inputting the above second potential vector data into a decoder; and A non-transient computer-readable storage medium characterized by including the step of acquiring OCV data according to the SoC from the above decoder.

Images