Method for Analyzing Battery Charge / Discharge Profile and Battery Charge / Discharge Profile Analyzer

Abstract

A method and device for analyzing a battery charge / discharge profile include a step of training a machine learning model using a plurality of training charge / discharge profiles as a training data set, each of the training charge / discharge profiles including training section classification information, which is a data set in which an identification number of one of a plurality of charge / discharge control sections sequentially performed in an activation process is assigned to each time index, a step of inputting a target charge / discharge profile acquired by the activation process for a battery cell into the machine learning model, and a step of acquiring target section classification information for the input target charge / discharge profile from the machine learning model. The target section classification information is a data set in which an identification number of one of the plurality of charge / discharge control sections is assigned to each time index of the target charge / discharge profile.

Description

Technical Field

[0001] The present invention relates to battery charge-discharge profile analysis. More specifically, based on the unique characteristics of each stage of the activation process, the charge-discharge profile monitored from a battery cell during the activation process is analyzed to automatically assign interval classification information corresponding to the stage of the activation process to the charge-discharge profile.

[0002] This application claims priority based on Korean Patent Application No. 10-2021-0136161 filed on October 13, 2021, and all the contents disclosed in the specification and drawings of the said application are incorporated into this application.

Background Art

[0003] Recently, the demand for portable electronic products such as notebook PCs, video cameras, and mobile phones has increased rapidly. As the development of electric vehicles, energy storage batteries, robots, satellites, etc. has been in full swing, research on high-performance batteries capable of repeated charge and discharge has been actively underway.

[0004] Currently, commercially available batteries include nickel-cadmium batteries, nickel-metal hydride batteries, nickel-zinc batteries, lithium batteries, etc. Among these, lithium batteries have attracted attention because they hardly cause a memory effect compared to nickel-based batteries, can be charged and discharged freely, have a very low self-discharge rate, and have a high energy density.

[0005] Battery cells assembled as finished products by a production line will finally be in a state where they can be shipped after going through an activation process. In the activation process, a series of pre-scheduled charge-discharge control intervals are sequentially performed by charge-discharge equipment, whereby the capacity and performance of the battery cell are adjusted to the design specifications.

[0006] In addition, the activation process is also the last process that can obtain battery inspection data used for sorting defective battery cells before the shipment of battery cells. During the progress of the charge-discharge control section of the activation process, by comparing the time-series data of various battery parameters (for example, voltage, current, capacity) obtained from the battery cell with the reference conditions set in advance so as to be related to several charge-discharge control sections, it can be used to check whether the battery cell is defective.

[0007] For this reason, conventional charge-discharge equipment may not normally perform the sectioning function of repeating the operation of assigning the identification number of the charge-discharge control section to the battery parameters during the progress of each charge-discharge control section of the activation process. Even if the identification number is assigned, a higher-precision section setting may be required to determine the quality of the product in the physical identification section.

[0008] Thus, when there is no section information or, even if it exists, appropriate section information for data analysis is not stored, it is difficult to perform feature extraction, diagnosis, and analysis on the battery cell from the time-series data (charge-discharge profile).

Summary of the Invention

Problems to be Solved by the Invention

[0009] The present invention has been made in view of the above problems, and an object of the present invention is to provide a method and an apparatus for automatically assigning an identification number for each charge-discharge control section of the activation process to each sample value (data point) constituting the time-series data of the charge-discharge profile obtained by monitoring the battery parameters of the battery cell during the activation process.

[0010] Other objects and advantages of the present invention can be understood from the following description and will be more clearly understood by the embodiments of the present invention. Further, the objects and advantages of the present invention can be realized by the means and combinations thereof shown in the claims.

Means for Solving the Problems

[0011] According to one aspect of the present invention, a method for analyzing a battery charge-discharge profile includes a step of training a machine learning model using a plurality of charge-discharge profiles for training as a training dataset, wherein each charge-discharge profile for training includes learning interval classification information which is a dataset in which an identification number of any one of a plurality of charge-discharge control intervals sequentially performed in an activation process is assigned to each of time indexes; a step of inputting a target charge-discharge profile obtained by the activation process for a battery cell into the machine learning model; and a step of obtaining target interval classification information for the input target charge-discharge profile from the machine learning model. The target interval classification information is a dataset in which an identification number of any one of the plurality of charge-discharge control intervals is assigned to each of time indexes of the target charge-discharge profile.

[0012] The target charge-discharge profile may include voltage time-series data indicating a change over time in the voltage of the battery cell according to the time index of the target charge-discharge profile, and current time-series data indicating a change over time in the charge-discharge current of the battery cell according to the time index of the target charge-discharge profile.

[0013] The machine learning model may be a decision tree.

[0014] The method for analyzing the battery charge-discharge profile may further include a step of determining whether the target charge-discharge profile is abnormal by comparing identification numbers assigned to the time indexes of the target interval classification information according to the order of the time indexes of the target interval classification information.

[0015] Among any two of the plurality of charge-discharge control intervals, the identification number of the charge-discharge control interval performed earlier may be smaller than the identification number of the charge-discharge control interval performed later.

[0016] In the step of determining whether there is an abnormality in the target charge-discharge profile, if the identification number assigned to the earlier time index among any two time indexes of the target interval classification information is larger than the identification number assigned to the later time index, the target charge-discharge profile may be determined to be abnormal.

[0017] In the step of determining whether there is an abnormality in the target charge-discharge profile, if there exists an identification number having a value between the two identification numbers assigned to two adjacent time indexes of the target interval classification information among the plurality of identification numbers of the plurality of charge-discharge control intervals, the target charge-discharge profile may be determined to be abnormal.

[0018] In the step of determining whether there is an abnormality in the target charge-discharge profile, if the identification number of at least one charge-discharge control interval among the plurality of charge-discharge control intervals is not assigned to any of the time indexes of the target interval classification information, the target charge-discharge profile may be determined to be abnormal.

[0019] A battery charge-discharge profile analysis device according to another aspect of the present invention includes a data acquisition unit configured to store a plurality of learning charge-discharge profiles, each learning charge-discharge profile including learning interval classification information that is a data set in which an identification number of any one of a plurality of charge-discharge control intervals sequentially performed in an activation process is assigned to each of time indexes, and a data processing unit configured to train a machine learning model using the plurality of learning charge-discharge profiles as a learning data set. The data processing unit is configured to input a target charge-discharge profile obtained by the activation process for a battery cell into the machine learning model and acquire target interval classification information for the target charge-discharge profile. The target interval classification information is a data set in which an identification number of any one of the plurality of charge-discharge control intervals is assigned to each of the time indexes of the target charge-discharge profile.

[0020] The machine learning model may be a decision tree.

[0021] The data processing unit may be configured to compare the identification numbers assigned to the time indices of the target interval classification information according to the order of the time indices of the target interval classification information, and determine whether there is an abnormality in the target charge and discharge profile.

[0022] In addition, the battery activation system according to still another aspect of the present invention includes the battery charge and discharge profile analyzer.

Advantages of the Invention

[0023] According to at least one of the embodiments of the present invention, for each sample value (data point) constituting the time-series data of the charge and discharge profile obtained by monitoring the battery parameters of the battery cell during the activation process, an identification number for each charge and discharge control interval of the activation process can be automatically assigned. Thereby, even if the charging and discharging equipment does not have a sectioning function or the sectioning function fails (an error occurs), the overall charge and discharge profile can be sectioned according to the progress order of the charge and discharge control intervals constituting the activation process, so that the efficiency of feature extraction, diagnosis, and analysis for the battery cell increases.

[0024] The effects of the present invention are not limited to the effects described above, and other effects of the present invention not mentioned will be clearly understood by those skilled in the art from the description of the claims.

[0025] The following drawings attached to this specification illustrate preferred embodiments of the present invention, and serve to further understand the technical idea of the present invention together with the detailed description of the invention. Therefore, the present invention should not be construed as being limited only to the matters described in the drawings.

Brief Description of the Drawings

[0026]

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

DETAILED DESCRIPTION OF THE INVENTION

[0027] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Prior to this, the terms and words used in this specification and the claims should not be construed as being limited to their ordinary or dictionary meanings, and the inventors themselves must interpret them in accordance with the technical idea of the present invention in accordance with the principle that they can appropriately define the concept of the terms in order to explain the invention in the best way.

[0028] Therefore, the embodiments described in this specification and the configurations shown in the drawings are merely the most desirable embodiments of the present invention and do not represent all of the technical ideas of the present invention. It should be understood that there may be various equivalents and modifications that can replace them at the time of this application.

[0029] Terms including ordinal numbers such as first, second, etc. are used for the purpose of distinguishing any one of various components from the rest, and the components are not limited by such terms.

[0030] Throughout the specification, when a certain part "includes" a certain component, this means that, unless otherwise stated to the contrary, it does not exclude other components but may further include other components. Also, terms such as "~ unit" described in the specification indicate a unit that processes at least one function or operation, and this can be implemented by hardware, software, or a combination of hardware and software.

[0031] Furthermore, throughout the specification, when a certain part is "connected" to another part, this includes not only the case where it is "directly connected", but also the case where it is "indirectly connected" through other elements in between.

[0032] FIG. 1 is a diagram exemplarily showing a schematic configuration of a battery activation system 1 according to an embodiment of the present invention.

[0033] Referring to FIG. 1, the battery activation system 1 includes a charge / discharge facility 100 and a battery charge / discharge profile analyzer 200.

[0034] The charge / discharge facility 100 is provided to sequentially perform a series of charge / discharge control intervals included in the activation process in accordance with a predetermined activation schedule for the activation process of the battery cell BC that has been assembled through the production line.

[0035] The charge and discharge device 100 has a charging function, a discharging function, and a standby function, and selectively performs one of the charging function, the discharging function, and the standby function according to an activation schedule, and charges and discharges the battery cell BC based on control parameters (for example, charging voltage, charging current, discharging voltage, discharging current, etc.) preset for each charge and discharge control section. The type of the battery cell BC is not particularly limited as long as it can be repeatedly charged and discharged, such as a lithium ion cell.

[0036] The charge and discharge device 100 includes a charger 110 and a process controller 120.

[0037] The charger 110 includes a power supply unit 111 and a charge and discharge unit 112.

[0038] The power supply unit 111 is configured to switch the power supplied from an AC power supply and / or a DC power supply to DC having a predetermined voltage level that matches the input specification of the charge and discharge unit 112. One or a combination of two of an AC-DC converter and a DC-DC converter can be used as the power supply unit 111.

[0039] The charge and discharge unit 112 has a pair of charge and discharge terminals provided therein connected to the positive and negative electrodes of the battery cell BC, and can charge and discharge the battery cell BC or interrupt charging and discharging in response to a command from the process controller 120. One or a combination of two of a constant current circuit and a constant voltage circuit can be used as the charge and discharge unit 112.

[0040] The process controller 120 stores an activation schedule in a memory mounted therein in advance. The process controller 120 starts an activation process in accordance with the activation schedule in response to a user input, and controls the charger 110 so that the charge and discharge control sections are activated in order in accordance with a preset progression order for each.

[0041] The battery charge / discharge profile analyzer 200 includes a data acquisition unit 210 and a data processing unit 220. The battery charge / discharge profile analyzer 200 may further include an information output unit 230. The operation of the battery charge / discharge profile analyzer 200 may be activated in response to the start of the activation process by the charge / discharge facility 100.

[0042] The data acquisition unit 210 is configured to periodically detect each sample value of the battery parameters at regular intervals during the activation process for the battery cell BC. The battery parameters include the voltage and current of the battery cell BC, and the data acquisition unit 210 includes a voltage detector 211 and a current detector 212.

[0043] The voltage detector 211 is connected to the positive and negative electrodes of the battery cell BC, detects the voltage applied across both ends of the battery cell BC, and generates (outputs) a signal indicating the sample value of the detected voltage.

[0044] The current detector 212 is provided in the charge / discharge path connecting the battery cell BC and the charge / discharge unit 112, detects the charge / discharge current flowing through the battery cell BC during the activation process, and generates (outputs) a signal indicating the sample value of the detected current. In one example, a known current detection element such as a shunt resistor and / or a Hall sensor may be used as the current detector 212. The voltage detector 211 and the current detector 212 may be integrated in the form of a single chip.

[0045] The information output unit 230 is provided to output various information related to the activation process in a form recognizable by the user. In one example, the information output unit 230 may include a monitor, a touch screen, a speaker, and / or a vibrator.

[0046] The data processing unit 220 is operably coupled to the data acquisition unit 210, the information output unit 230, and the charge / discharge facility 100. That two components are operably coupled means that the two components are directly or indirectly connected so as to be able to transmit and receive signals in one direction or both directions.

[0047] The data processing unit 220 can be implemented by hardware using at least one of an ASIC (application specific integrated circuit), a DSP (digital signal processor), a DSPD (digital signal processing device), a PLD (programmable logic device), an FPGA (field programmable gate array), a microprocessor, and other electrical units for performing functions.

[0048] The data processing unit 220 may have a built-in memory. The memory may include, for example, at least one type of storage medium such as a flash memory (registered trademark) type, a hard disk type, an SSD type (Solid State Disk type), an SDD type (Silicon Disk Drive type), a multimedia card micro type, a RAM (random access memory), an SRAM (static random access memory), a ROM (read-only memory), an EEPROM (electrically erasable programmable read-only memory), or a PROM (programmable read-only memory). The memory may store data and programs required for the operations described later by the data processing unit 220. The memory may store data indicating the results of operations performed by the data processing unit 220.

[0049] FIG. 2 is a graph showing the relationship between the charge / discharge profile that can be obtained from the battery cell during the activation process and the charge / discharge control section of the activation process.

[0050] Referring to FIG. 2, the charge / discharge profile includes voltage time-series data and current time-series data.

[0051] Curve C1 shows the voltage time-series data. The voltage time-series data is a set of sample values of the voltage applied across both ends of the battery cell BC periodically detected during the activation process, aligned by the time index of the detection time.

[0052] Curve C2 shows current time-series data. The current time-series data is a set in which sample values of the charge and discharge current of the battery cell BC, which are periodically detected during the activation process, are aligned by the time index of the detection time.

[0053] In FIG. 2, it shows that a total of six charge and discharge control intervals, that is, the first to sixth charge and discharge control intervals N1 to N6, are sequentially performed from the start to the end of the activation process.

[0054] The first charge and discharge control interval N1 is a standby interval before charging. Before performing the second charge and discharge control interval N2 on the assembled battery cell BC, in order to achieve electrochemical stabilization inside the battery cell BC, it is an interval where charging and discharging are not performed and the battery is left idle for a predetermined first rest time. In the first charge and discharge control interval N1, the voltage of the battery is maintained constant, and the current of the battery is also constant at 0 A.

[0055] The second charge and discharge control interval N2 is an interval in which the battery cell BC is charged at a constant current by the charger 110, which starts immediately after the end of the first charge and discharge control interval N1. The second charge and discharge control interval N2 can end when the voltage of the battery cell BC reaches a predetermined cut-off voltage or when a predetermined time has elapsed from the start time of the second charge and discharge control interval N2. In the second charge and discharge control interval N2, the voltage of the battery cell BC continues to increase, and the charging current of the battery cell BC is constant as a predetermined current rate.

[0056] The third charge / discharge control section N3 is a section in which the battery cell BC is charged at a constant voltage by the charger 110, which starts immediately after the end of the second charge / discharge control section N2. The voltage level of the constant current charge may be the same as the cut-off voltage in the second charge / discharge control section N2. The third charge / discharge control section N3 may end when the charging current of the battery cell BC reaches a predetermined cut-off current or when a predetermined time has elapsed since the start time of the third charge / discharge control section N3. In the third charge / discharge control section N3, the voltage of the battery cell BC increases later than in the second charge / discharge control section N2. In the third charge / discharge control section N3, as the difference between the voltage level of the constant voltage charge and the voltage of the battery cell BC gradually decreases, the charging current of the battery cell BC naturally decreases toward 0A.

[0057] The fourth charge / discharge control section N4 is a standby section after charging, and is a section that prevents charge / discharge from being performed during a predetermined second rest time in order to relieve polarization due to charging over the second and third charge / discharge control sections. In the fourth charge / discharge control section N4, the voltage of the battery cell BC gradually stabilizes while slowly decreasing from the cut-off voltage by a slightly lower voltage, and the current of the battery cell BC is constant at 0A as in the first charge / discharge control section N1.

[0058] The fifth charge / discharge control section N5 is a section in which the battery cell BC is discharged at a constant current by the charger 110, which starts immediately after the end of the fourth charge / discharge control section N4. The fifth charge / discharge control section N5 may end when the voltage of the battery cell BC reaches a predetermined discharge end voltage or when a predetermined time has elapsed since the start time of the fifth charge / discharge control section N5. In the fifth charge / discharge control section N5, the voltage of the battery cell BC continues to decrease, and the discharge current of the battery cell BC is constant at a predetermined current rate.

[0059] The sixth charge / discharge control interval N6 is a standby interval after discharge. In order to relieve polarization caused by discharge over the fifth charge / discharge control interval N5, it is an interval where charge / discharge is not performed and the battery is left idle for a predetermined third rest time. At the start time of the sixth charge / discharge control interval N6, the voltage of the battery cell BC recovers from the discharge end voltage by the voltage drop due to the current rate of constant current discharge and the internal resistance of the battery cell BC and then gradually stabilizes. The current of the battery cell BC is constant at 0 A as in the first and fourth charge / discharge control intervals N1 and N4.

[0060] The point to note is that in the first to sixth charge / discharge control intervals N1 to N6 described above, in each of these charge / discharge control intervals, sample values of battery parameters that are uniquely distinguished in each of the remaining charge / discharge control intervals are acquired. As an example, the fifth charge / discharge control interval N5 has a feature that is distinguished from the remaining charge / discharge control intervals in that it is an interval in which a discharge current flows through the battery cell BC. In another example, the first, fourth, and sixth charge / discharge control intervals are distinguished from the second, third, and fifth charge / discharge control intervals in that the current is 0 A. Also, although the first, fourth, and sixth charge / discharge control intervals have in common that the current is 0 A, they are distinguished from each other in that the voltages of the battery cell BC in the three intervals are located in three non-overlapping voltage ranges.

[0061] The battery charge / discharge profile analyzer 200 can provide a sectioning function for post - adding section classification information to the charge / discharge profile after the activation process is completed. This is based on the fact that in each charge / discharge control section of the activation process, sample values of battery parameters with unique characteristics distinguishable from other charge / discharge control sections are obtained. The operation assigns the identification number of the charge / discharge control section identified by the sample value (or the combination of sample values of two battery parameters) to be related to the time index of the sample value (or the combination of sample values of two battery parameters). The identification numbers can have an ascending or descending relationship according to the progression order of the charge / discharge control sections. In one example, the identification number (e.g., N1) of the previously performed charge / discharge control section may be smaller than the identification number (e.g., N2) of the subsequently performed charge / discharge control section. In FIG. 2, the difference between the two identification numbers of two adjacent charge / discharge control sections is exemplified as 1. Hereinafter, for convenience of explanation, the same reference numerals are given to each charge / discharge control section and its identification number.

[0062] The section classification information is a data set of identification numbers aligned according to the time index, located within the time range from the start time to the end time of the activation process. The battery charge / discharge profile analyzer 200 can activate the sectioning function by executing a machine - learning model trained using a plurality of learning charge / discharge profiles as a learning data set. Each learning charge / discharge profile includes learning section classification information, which is a data set in which an identification number of any one of the plurality of charge / discharge control sections sequentially performed in the activation process is assigned to each of its time indices. The learning section classification information can be manually input by the user.

[0063] FIG. 3 is a flowchart showing a learning method for a machine learning model, and FIG. 4 is a diagram for explaining a decision tree which is an example of a machine learning model completed learning by the method of FIG. 3. The method of FIG. 3 can be executed by the battery charge / discharge profile analyzer 200. Of course, as long as multi-classification is possible, other models other than decision trees can be applied as machine learning models.

[0064] Referring to FIGS. 3 and 4, in step S310, the data processing unit 220 inputs a plurality of learning charge / discharge profiles to the machine learning model 400.

[0065] In step S320, the data processing unit 220 acquires observation interval classification information corresponding to each input learning charge / discharge profile from the machine learning model 400. Each observation interval classification information is a data set classified by the machine learning model 400, that is, a time series of identification numbers added to the time index of each learning charge / discharge profile.

[0066] In step S330, the data processing unit 220 compares the observation interval classification information and the learning interval classification information corresponding to each other, and determines whether there is an abnormality in at least one observation interval classification information. The value of step S330 being "no" means that the learning is completed because the impurity of the machine learning model is less than the reference value. The value of step S330 being "yes" means that the learning is still necessary because the impurity of the machine learning model is greater than or equal to the reference value. If the value of step S330 is "yes", the process proceeds to step S340.

[0067] In step S340, the data processing unit 220 re-inputs the learning charge / discharge profile corresponding to the abnormal observation interval classification information among the plurality of learning charge / discharge profiles to the machine learning model 400.

[0068] Figure 4 shows a decision tree 400 for which learning has been completed for the first to sixth charge and discharge control intervals N1 to N6. The decision tree 400 includes a root node 410, first to fourth intermediate nodes 421 to 424, and first to sixth terminal nodes. The root node 410 and the intermediate nodes can be collectively referred to as internal nodes, and the internal nodes have unique classification conditions. The first to sixth terminal nodes 431 to 436 are classification results of time index-specific data (voltage sample value + current sample value) and are associated with the identification numbers of the first to sixth charge and discharge control intervals N1 to N6, respectively.

[0069] The root node 410 is a node that provides a classification condition stronger than the first to fourth intermediate modes 421 to 424. In FIG. 4, at the root node 410, the current sample value SC for each time index is compared with the discharge current value H1 associated with the fifth charge and discharge control interval N5 so that the fifth charge and discharge control interval N5 and the remaining charge and discharge control intervals are classified. The discharge current value H1 is determined by the learning process described above. The identification number N5 associated with the fifth terminal node 435 is assigned to the time index-specific data that satisfies the classification condition of the root node 410. The time index-specific data that does not satisfy the classification condition of the root node 410 is input to the first intermediate node 421.

[0070] The first intermediate node 421 is a node that provides a classification condition stronger than the second to fourth intermediate modes 422 to 424. In FIG. 4, at the first intermediate node 421, the current sample value S C for each time index is compared with the charge current value H2 associated with the second charge and discharge control interval N2 so that the second charge and discharge control interval N2 and the remaining charge and discharge control intervals are classified. The charge current value H2 is determined by the learning process described above. The identification number N2 associated with the second terminal node 432 is assigned to the time index-specific data that satisfies the classification condition of the first intermediate node 421. The time index-specific data that does not satisfy the classification condition of the first intermediate node 421 is input to the second intermediate node 422.

[0071] The second intermediate node 422 is a node that provides a classification condition stronger than the third and fourth intermediate modes 423 and 424. In FIG. 4, at the second intermediate node 422, the voltage sample value S by time index is such that the sixth charge-discharge control section N6 and the remaining charge-discharge control sections are classified V and compared with the voltage value H3 related to the sixth charge-discharge control section N6. The voltage value H3 related to the sixth charge-discharge control section N6 is determined by the learning process described above. To the data by time index that satisfies the classification condition of the second intermediate node 422, the identification number N6 related to the sixth terminal node 436 is assigned. The data by time index that does not satisfy the classification condition of the second intermediate node 422 is input to the third intermediate node 423.

[0072] The third intermediate node 423 is a node that provides a classification condition stronger than the fourth intermediate mode 424. In FIG. 4, at the third intermediate node 423, the voltage sample value S by time index is such that the third charge-discharge control section N3 and the remaining charge-discharge control sections are classified V and compared with the voltage value H4 related to the third charge-discharge control section N3. The voltage value H4 related to the third charge-discharge control section N3 is determined by the learning process described above. To the data by time index that satisfies the classification condition of the third intermediate node 423, the identification number N3 related to the third terminal node 433 is assigned. The data by time index that does not satisfy the classification condition of the third intermediate node 423 is input to the fourth intermediate node 424.

[0073] The fourth intermediate node 424 is a node that provides the last classification condition. In FIG. 4, at the fourth intermediate node 424, the voltage sample value S by time index is such that the first charge-discharge control section N1 and the fourth charge-discharge control section N4 are classified from each other VIt is compared with the voltage value H5 related to the first charge-discharge control section N1. The voltage value H5 related to the first charge-discharge control section N1 is determined by the learning process described above. To the data for each time index that satisfies the classification condition of the fourth intermediate node 424, the identification number N1 related to the first terminal node 431 is assigned. To the data for each time index that does not satisfy the classification condition of the fourth intermediate node 424, the identification number N4 related to the fourth terminal node 434 is assigned.

[0074] The data processing unit 220 can input a target charge-discharge profile into the machine learning model 400 for which learning has been completed and add target section classification information to the target charge-discharge profile. The target charge-discharge profile refers to a charge-discharge profile obtained by an actual activation process for the assembled battery cell BC. The target section classification information refers to section classification information added to the target charge-discharge profile.

[0075] FIGS. 5 to 8 show the sectioning results for the target charge-discharge profile obtained from the machine learning model 400 for which learning has been completed. The data processing unit 220 can compare the identification numbers assigned to the time indices of the target section classification information in the order of the time indices of the target section classification information to determine the presence or absence of abnormality in the target charge-discharge profile. In FIGS. 5 to 8, the symbols K, L, M, N, and O used for the time indices are shown as natural numbers having the relationship of 1 < K < L < M < N < O.

[0076] First, FIG. 5 is an example of normal target section classification information added to the target charge and discharge profile. Referring to FIG. 3, the identification number of the first charge and discharge control section N1 is assigned to each time index from time t1 to before time t2, the identification number of the second charge and discharge control section N2 is assigned to each time index from time t2 to before time t3, the identification number of the third charge and discharge control section N3 is assigned to each time index from time t3 to before time t4, the identification number of the fourth charge and discharge control section N4 is assigned to each time index from time t4 to before time t5, the identification number of the fifth charge and discharge control section N5 is assigned to each time index from time t5 to before time t6, and the identification number of the sixth charge and discharge control section N6 is assigned to each time index from time t6 to before time t7.

[0077] Next, FIG. 6 is an example of abnormal target section classification information added to the target charge and discharge profile. Referring to FIG. 6, when comparing with FIG. 5, the identification number of the fourth charge and discharge control section N4 that is performed next to the third charge and discharge control section N3 is assigned to at least one time index among the time indexes from time t3 to before time t4. As shown in FIG. 6, when the identification number (for example, N4) assigned to the previous time index (for example, M-2) is larger than the identification number (for example, N3) assigned to the later time index (for example, M-1) among any two time indexes of the target section classification information, the data processing unit 220 may determine that the target charge and discharge profile is abnormal.

[0078] Next, FIG. 7 shows another example of the abnormal target section classification information added to the target charge and discharge profile. Referring to FIG. 7, when compared with FIG. 5, the identification number of the first charge and discharge control section N1 is normally assigned to the time index from time t1 to before time t2, while the identification number of the fourth charge and discharge control section N4 instead of the second charge and discharge control section N2 is assigned to some time indexes between time t2 and time t3. As shown in FIG. 7, when there exists an identification number (for example, N3) having a value between two identification numbers (for example, N2, N4) assigned to two adjacent time indexes (for example, K, K + 1) of the target section classification information, the data processing unit 220 may determine that the target charge and discharge profile is abnormal.

[0079] Next, FIG. 8 shows still another example of the abnormal target section classification information added to the target charge and discharge profile. Referring to FIG. 8, when compared with FIG. 5, the identification number of the fourth charge and discharge control section N4 is assigned to all of the time indexes from time t5 to before time t6. As shown in FIG. 8, when the identification number of at least one of the plurality of charge and discharge control sections N1 to N6 (for example, N5) is not assigned to any of the time indexes of the target section classification information, the data processing unit 220 may determine that the target charge and discharge profile is abnormal.

[0080] FIG. 9 is a flowchart showing a method for analyzing a battery charge and discharge profile according to an embodiment of the present invention.

[0081] Referring to FIG. 9, in step S910, the data processing unit 220 uses the plurality of charge and discharge profiles for learning stored in the data acquisition unit 210 as a learning data set to train the machine learning model 400 (see FIGS. 3 and 4).

[0082] In step S920, the data processing unit 220 inputs the target charge and discharge profile obtained by the activation process for the battery cell BC into the machine learning model 400.

[0083] In step S930, the data processing unit 220 obtains target interval classification information for the input target charge-discharge profile from the machine learning model 400. The target interval classification information can be recorded in the memory or transmitted to the user by the information output unit 230.

[0084] The method of FIG. 9 may further include step S940. In step S940, the data processing unit 220 compares the identification numbers assigned to the time indexes of the target interval classification information according to the order of the time indexes of the target interval classification information to determine whether there is an abnormality in the target charge-discharge profile (see FIGS. 5 to 8). If the value of step S940 is "yes", it may proceed to step S950.

[0085] In step S950, the data processing unit 220 may output a notification signal including at least one of (i) the abnormal type of the target interval classification information and (ii) the time index to which the identification number is wrongly assigned. The notification signal can be recorded in the memory or transmitted to the user by the information output unit 230.

[0086] The embodiments of the present invention described above are not necessarily implemented through the apparatus and method, but may be implemented through a program that realizes the functions corresponding to the configurations of the embodiments of the present invention or a recording medium on which the program is recorded. Such an implementation should be easily achievable by an expert in the technical field to which the present invention belongs from the description of the above-described embodiments.

[0087] As described above, the present invention has been described with reference to the limited embodiments and drawings, but the present invention is not limited thereto, and it goes without saying that various modifications and variations are possible within the equivalent scope of the technical idea and claims of the present invention by those having ordinary knowledge in the technical field to which the present invention belongs.

[0088] In addition, since the above-described present invention can be variously substituted, modified, and changed by those having ordinary knowledge in the technical field to which the present invention pertains without departing from the technical idea of the present invention, it is not limited by the above-described embodiments and the attached drawings, and all or part of each embodiment can be selectively combined and configured so that various modifications can be made.

Explanation of Reference Numerals

[0089] 1 Battery activation system BC Battery cell 100 Charging and discharging equipment 110 Charger 111 Power supply unit 112 Charging and discharging unit 120 Controller 200 Battery charge and discharge profile analyzer 210 Data acquisition unit 211 Voltage detector 212 Current detector 220 Data processing unit 230 Information output unit

Claims

1. A step of training a machine learning model using a plurality of charging and discharging profiles for training as a training dataset, wherein each charging and discharging profile for training includes learning interval classification information which is a dataset in which an identification number of any one of a plurality of charging and discharging control intervals sequentially performed in an activation process is assigned to each time index; A step of inputting a target charging and discharging profile obtained by the activation process for a battery cell into the machine learning model; A step of obtaining target interval classification information for the input target charging and discharging profile from the machine learning model, wherein the target interval classification information is a dataset in which an identification number of any one of the plurality of charging and discharging control intervals is assigned to each time index of the target charging and discharging profile; Among any two of the plurality of charging and discharging control intervals, the identification number of the charging and discharging control interval performed earlier is smaller than the identification number of the charging and discharging control interval performed later; When, among any two time indexes of the target interval classification information, the identification number assigned to the earlier time index is larger than the identification number assigned to the later time index, determining that the target charging and discharging profile is abnormal; A method for analyzing a battery charging and discharging profile.

2. A step of training a machine learning model using a plurality of charging and discharging profiles for training as a training dataset, wherein each charging and discharging profile for training includes learning interval classification information which is a dataset in which an identification number of any one of a plurality of charging and discharging control intervals sequentially performed in an activation process is assigned to each time index; A step of inputting a target charging and discharging profile obtained by the activation process for a battery cell into the machine learning model; A step of obtaining target interval classification information for the input target charging and discharging profile from the machine learning model, wherein the target interval classification information is a dataset in which an identification number of any one of the plurality of charging and discharging control intervals is assigned to each time index of the target charging and discharging profile; If there exists an identification number having a value between two identification numbers assigned to two adjacent time indexes of the target section classification information among the plurality of identification numbers of the plurality of charge and discharge control sections, determine that the target charge and discharge profile is abnormal. Method for analyzing battery charge and discharge profile.

3. A step of training a machine learning model using a plurality of training charge and discharge profiles as a training dataset, wherein each training charge and discharge profile includes training section classification information which is a dataset in which an identification number of any one of a plurality of charge and discharge control sections sequentially performed in an activation process is assigned to each of time indexes. A step of inputting the target charge and discharge profile obtained by the activation process for the battery cell into the machine learning model. A step of obtaining target section classification information for the input target charge and discharge profile from the machine learning model, wherein the target section classification information is a dataset in which an identification number of any one of the plurality of charge and discharge control sections is assigned to each of the time indexes of the target charge and discharge profile. If the identification number of at least one of the plurality of charge and discharge control sections is not assigned to any of the time indexes of the target section classification information, determine that the target charge and discharge profile is abnormal. Method for analyzing battery charge and discharge profile.

4. The target charge and discharge profile is voltage time series data indicating the change over time of the voltage of the battery cell according to the time index of the target charge and discharge profile, and current time series data indicating the change over time of the charge and discharge current of the battery cell according to the time index of the target charge and discharge profile, and the method for analyzing a battery charge and discharge profile according to any one of claims 1 to 3.

5. The method for analyzing a battery charge and discharge profile according to any one of claims 1 to 3, wherein the machine learning model is a decision tree.

6. configured to store a plurality of learning charge / discharge profiles, each learning charge / discharge profile including learning section classification information which is a data set in which an identification number of any one of a plurality of charge / discharge control sections sequentially performed in an activation process is assigned to each time index, and a data acquisition unit; a data processing unit configured to train a machine learning model using the plurality of learning charge / discharge profiles as a learning data set; and the data processing unit is configured to input a target charge / discharge profile obtained by the activation process for a battery cell into the machine learning model and acquire target section classification information for the target charge / discharge profile; the target section classification information is a data set in which an identification number of any one of the plurality of charge / discharge control sections is assigned to each time index of the target charge / discharge profile; among any two of the plurality of charge / discharge control sections, the identification number of the charge / discharge control section performed earlier is smaller than the identification number of the charge / discharge control section performed later; when, among any two time indexes of the target section classification information, the identification number assigned to the earlier time index is larger than the identification number assigned to the later time index, the target charge / discharge profile is determined to be abnormal; a battery charge / discharge profile analysis device. **Claim 7**: A data acquisition unit configured to store a plurality of learning charge / discharge profiles, each learning charge / discharge profile including learning section classification information which is a data set in which an identification number of any one of a plurality of charge / discharge control sections sequentially performed in an activation process is assigned to each time index; and a data processing unit configured to train a machine learning model using the plurality of learning charge / discharge profiles as a learning data set; and the data processing unit is configured to input a target charge / discharge profile obtained by the activation process for a battery cell into the machine learning model and acquire target section classification information for the target charge / discharge profile; The target interval classification information is a dataset in which an identification number of any one of the plurality of charge and discharge control intervals is assigned to each time index of the target charge and discharge profile. When there exists an identification number having a value between two identification numbers assigned to two adjacent time indices of the target interval classification information among the plurality of identification numbers of the plurality of charge and discharge control intervals, determine that the target charge and discharge profile is abnormal. Battery charge and discharge profile analysis device.

8. A data acquisition unit configured to store a plurality of learning charge and discharge profiles, each learning charge and discharge profile including learning interval classification information which is a dataset in which an identification number of any one of the plurality of charge and discharge control intervals sequentially performed in an activation process is assigned to each time index. A data processing unit configured to train a machine learning model using the plurality of learning charge and discharge profiles as a training dataset. The data processing unit is configured to input a target charge and discharge profile acquired by the activation process for the battery cell into the machine learning model and acquire target interval classification information for the target charge and discharge profile. The target interval classification information is a dataset in which an identification number of any one of the plurality of charge and discharge control intervals is assigned to each time index of the target charge and discharge profile. When the identification number of at least one of the plurality of charge and discharge control intervals is not assigned to any of the time indices of the target interval classification information, determine that the target charge and discharge profile is abnormal. Battery charge and discharge profile analysis device.

9. The battery charge and discharge profile analysis device according to any one of claims 6 to 8, wherein the machine learning model is a decision tree.

10. A battery activation system including the battery charge and discharge profile analysis device according to any one of claims 6 to 8.

Images