Apparatus and method for estimating battery capacity

Abstract

A battery capacity estimation device according to an embodiment of the present invention may include a voltage measurement unit that measures a voltage of a battery cell; a filtering unit that determines voltage data of the battery cell as first data when a logging pattern of the voltage data of the battery cell is outside a predetermined reference range; a statistical analysis unit that determines second data by statistically analyzing the voltage data of the battery cell; and a capacity estimation unit that estimates a capacity of a battery cell to be measured by applying data classified by the filtering unit or the statistical analysis unit to a capacity estimation model generated by classifying the first data and the second data.

Description

[Technical field]

[0001] [CROSS REFERENCE TO RELATED APPLICATIONS] This application claims the benefit of priority based on Korean Patent Application No. 10-2020-0101866, filed on August 13, 2020, and all contents disclosed in the documents of that Korean patent application are incorporated herein by reference.

[0002] The present invention relates to an apparatus and method for estimating battery capacity by constructing a model that separates data that is easy to analyze from data that is not easy to analyze. [Background technology]

[0003] Recently, research and development into secondary batteries has been actively conducted. Here, secondary batteries are batteries that can be charged and discharged, and include both conventional Ni / Cd batteries, Ni / MH batteries, and the latest lithium-ion batteries. Among secondary batteries, lithium-ion batteries have the advantage of having a much higher energy density than conventional Ni / Cd batteries, Ni / MH batteries, and the like. In addition, lithium-ion batteries can be manufactured to be small and lightweight, and are used as power sources for mobile devices. The range of use of lithium-ion batteries has also been expanded to include power sources for electric vehicles, and they are attracting attention as a next-generation energy storage medium.

[0004] In addition, the secondary battery is generally used as a battery pack including a battery module in which a plurality of battery cells are connected in series and / or parallel, and the state and operation of the battery pack are managed and controlled by a battery management system.

[0005] In order to judge the quality of each cell during production, these secondary batteries undergo a Lot Assembly Test (LAT) during the production process. Generally, one sample is selected per lot, which is a unit of multiple cells, and LAT is carried out. LAT accelerates degradation of the secondary battery and checks whether the capacity remains at a certain level even after 300 cycles.

[0006] Through such LAT, full charge and discharge are performed under the same experimental conditions for 1 to 300 cycles, and the pass / fail status of the cell is determined based on the SOH (state of health) life at 300 cycles. However, since the data logged in the LAT includes both data that is easy to analyze and data that is not easy to analyze, there is a problem that it is difficult to analyze the data values ​​and the accuracy is reduced. Summary of the Invention [Problem to be solved by the invention]

[0007] The present invention has been devised to solve the above problems, and aims to provide a battery capacity estimation device and method that can accurately and efficiently estimate the capacity of a battery cell by classifying battery cell logging data into data that is easy to analyze and data that is not easy to analyze, and applying a capacity estimation model to each data. [Means for solving the problem]

[0008] A battery capacity estimation device according to an embodiment of the present invention may include a voltage measurement unit that measures a voltage of a battery cell, a filtering unit that determines voltage data of the battery cell as first data when a logging pattern of the voltage data of the battery cell is outside a preset reference range, a statistical analysis unit that determines second data by statistically analyzing the voltage data of the battery cell, and a capacity estimation unit that estimates capacity by applying data classified by the filtering unit or the statistical analysis unit to a capacity estimation model generated by classifying the first data and the second data.

[0009] A battery capacity estimation method according to an embodiment of the present invention may include a step of measuring a voltage of a battery cell; determining, when a logging pattern of voltage data of the battery cell is outside a preset reference range, the voltage data as first data; determining second data by statistically analyzing the voltage data of the battery cell; and estimating a capacity of the battery cell by applying the first data and the second data to a capacity estimation model, respectively. Effect of the Invention

[0010] According to the battery capacity estimation device and method of the present invention, the battery cell capacity can be estimated accurately and efficiently by dividing the logging data of the battery cell into data that is easy to analyze and data that is not easy to analyze, and applying a capacity estimation model to each data. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 is a block diagram showing the configuration of a typical battery rack. [Diagram 2] 1 is a block diagram showing a configuration of a battery capacity estimation device according to an embodiment of the present invention; [Diagram 3] 4 is a flowchart illustrating an operation of classifying data by the battery capacity estimation device according to an embodiment of the present invention. [Figure 4] FIG. 13 illustrates the correlation between logging data and capacity of a battery cell. [Diagram 5] FIG. 2 is a diagram showing the classification of logging data of battery cells into multiple clusters by a battery capacity estimation device according to an embodiment of the present invention. [Figure 6] FIG. 11 is a diagram showing how logging data of all battery cells is classified into multiple clusters by a battery capacity estimation device according to an embodiment of the present invention. [Figure 7] FIG. 2 is a diagram illustrating an example of a capacity estimation model (LSTM) of a battery capacity estimation device according to an embodiment of the present invention. [Figure 8] 2 is a flowchart illustrating a battery capacity estimation method according to an embodiment of the present invention. [Figure 9] 1 is a block diagram showing a hardware configuration of a battery capacity estimation device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this document, the same reference numerals are used to refer to the same elements in the drawings, and duplicated descriptions of the same elements will be omitted.

[0013] Specific structural or functional descriptions of the various embodiments of the present invention disclosed in this document are merely exemplary for purposes of describing the embodiments of the present invention, and the various embodiments of the present invention may be embodied in various forms and should not be construed as being limited to the embodiments described in this document.

[0014] The terms "first," "second," "primary," or "second" used in the various embodiments may modify various components regardless of order and / or importance, and do not limit the components. For example, a first component may be named a second component, and similarly, a second component may be named instead of a first component, without departing from the scope of the present invention.

[0015] The terms used in this document are merely used to describe a particular embodiment and are not intended to limit the scope of other embodiments. A singular expression may include a plural expression unless a clearly different meaning is present in the context.

[0016] All terms used herein, including technical and scientific terms, may have the same meaning as commonly understood by a person of ordinary skill in the art of the present invention. Terms defined in commonly used dictionaries may be interpreted as having the same or similar meaning as they have in the context of the relevant art, and unless expressly defined in this document, they are not interpreted as ideal or overly formal. In some cases, even if a term is defined in this document, it should not be interpreted as excluding an embodiment of the present invention.

[0017] FIG. 1 is a block diagram showing the configuration of a typical battery rack.

[0018] Referring to FIG. 1, there is shown a schematic diagram of a battery control system including a battery rack 1 and a host controller 2 included in the host system according to an embodiment of the present invention.

[0019] 1, the battery rack 1 includes a chargeable and dischargeable battery module 10 consisting of one or more battery cells, a switching unit 14 connected in series to the (+) terminal side or (-) terminal side of the battery module 10 to control the flow of charge and discharge current of the battery module 10, and a battery management system 20 (e.g., RBMS) that monitors the voltage, current, temperature, etc. of the battery rack 1 and controls and manages to prevent overcharging and overdischarging, etc. In this case, the battery rack 1 may include a plurality of battery modules 10, sensors 12, switching units 14, and battery management systems 20.

[0020] Here, the switching unit 14 is an element for controlling the flow of current for charging or discharging the multiple battery modules 10, and may be, for example, at least one relay, magnetic contactor, etc. depending on the specifications of the battery rack 1.

[0021] The battery management system 20 is an interface to which measured values ​​of the various parameters described above are input, and may include a plurality of terminals and a circuit connected to the terminals for processing the input values. The battery management system 20 may control the ON / OFF of the switching unit 14, for example, a relay or a contactor, and may be connected to the battery modules 10 to monitor the status of each battery module 10.

[0022] Meanwhile, in the battery management system 20 of the present invention, as described below, a regression analysis on the voltage of the battery cell can be performed through a separate program. Also, the calculated regression equation can be used to classify the abnormality type of the battery cell.

[0023] The upper controller 2 can transmit a control signal for the battery module 10 to the battery management system 20. As a result, the operation of the battery management system 20 can be controlled based on the signal applied from the upper controller 2. Meanwhile, the battery cell of the present invention may be configured to be included in a battery module 10 used in an ESS (Energy Storage System). In this case, the upper controller 2 may be a battery bank controller (BBMS) including multiple racks or an ESS controller that controls the entire ESS including multiple banks. However, the battery rack 1 is not limited to such an application.

[0024] The configuration of the battery rack 1 and the configuration of the battery management system 20 are well known, so a more detailed description thereof will be omitted.

[0025] FIG. 2 is a block diagram showing a configuration of a battery capacity estimation device according to an embodiment of the present invention.

[0026] 2, the battery capacity estimation device 200 according to an embodiment of the present invention may include a voltage measurement unit 210, a filtering unit 220, a statistical analysis unit 230, and a capacity estimation unit 240. For example, the battery capacity estimation device 200 according to an embodiment of the present invention may be a device that analyzes charge / discharge data of a battery in real time to estimate the capacity of a battery cell, or may be a device that samples and tests for defects before mounting the battery cells in a module during manufacture.

[0027] The voltage measurement unit 210 may measure the voltage of the battery cell. In this case, the voltage measurement unit 210 may measure the voltage of the battery cell at a certain time interval. For example, the voltage measurement unit 210 may measure the voltage for an idle period after charging or discharging the battery cell. In addition, the voltage measurement unit 210 may measure a current flowing through the battery cell.

[0028] When a logging pattern of the voltage data of the battery cell falls outside a preset reference range, the filtering unit 220 may determine the voltage data as the first data. Specifically, when at least one of a logging time and a logging number of the voltage data of the battery cell falls outside a reference range, the filtering unit 220 may determine the voltage data as the first data.

[0029] For example, the logging pattern of the battery cell voltage data may be such that four voltage data are extracted in 5-minute intervals. In this case, there may be an error in the logging time, such as when the logging time of the battery cell voltage data is (59, 10, 20, 5 minutes), or there may be a lack of certain logging data, such as when the logging time is (5, 10, 20 minutes) or (10, 15, 20 minutes). In such a case, the filtering unit 220 may classify the data as the first data.

[0030] The statistical analysis unit 230 may determine the second data by statistically analyzing the voltage data of the battery cell. In this case, the statistical analysis unit 230 may perform the statistical analysis on data other than the data determined as the first data by the filtering unit 220. In addition, the statistical analysis unit 230 may calculate differential data (dV / dt) of the voltage with respect to time of the battery cell. As described below, the differential data of the voltage with respect to time shows a tendency more clearly than logged voltage data, so the statistical analysis unit 230 may use the differential data of the voltage with respect to time.

[0031] Specifically, the statistical analysis unit 230 may extract principal component data for the voltage data by performing a principal component analysis (PCA) on the differential data.

[0032] In addition, the statistical analysis unit 230 may calculate multiple clusters for the principal component data through k-means clustering, and if the differential data is included in a specific cluster, it may determine it as the second data, and if the differential data is not included in the cluster, it may determine it as the first data.

[0033] For example, the first data may be data in which voltage data of a battery cell is represented in a discontinuous form, and the second data may be data in which voltage data of the battery cell is represented in a continuous form, i.e., the first data may be data that is difficult to analyze, and the second data may be data that is easy to analyze.

[0034] In this manner, the battery capacity estimation device 200 according to an embodiment of the present invention can improve accuracy by applying a capacity estimation model by classifying the first data and the second data having different shapes, compared to a case where data in which the first and second data are mixed is used. For example, according to the battery capacity estimation device 200 according to an embodiment of the present invention, in predicting the capacity of the 300th cycle using data for the initial 100 cycles during charging and discharging of a battery, the data can be classified and applied to the capacity estimation model to improve accuracy.

[0035] The capacity estimation unit 240 may estimate the capacity by applying the data classified by the filtering unit 220 or the statistical analysis unit 230 for the battery cell to be measured to a capacity estimation model generated by classifying the first data and the second data. For example, the capacity estimation model may be a Long Short Term Memory Networks (LSTM) model. In this case, the capacity estimation models generated for the first data and the second data may include different parameter values.

[0036] Meanwhile, although not shown in Fig. 2, the battery capacity estimation device 200 according to an embodiment of the present invention may include a storage unit. In this case, the storage unit may store various data such as voltage data measured by the voltage measurement unit 210, voltage data classified by the filtering unit 220 and the statistical analysis unit 230, and a capacity estimation program. Also, the battery capacity estimation device 200 according to an embodiment of the present invention may operate in a manner of communicating with an external server via a communication unit (not shown) to transmit and receive the above-mentioned data, instead of including the storage unit.

[0037] As described above, according to the battery capacity estimation device 200 according to an embodiment of the present invention, the battery cell logging data is divided into data that is easy to analyze and data that is not easy to analyze, and a capacity estimation model is applied to each data, thereby making it possible to accurately and efficiently estimate the capacity of the battery cell.

[0038] FIG. 3 is a flowchart showing an operation of classifying data by the battery capacity estimation device according to an embodiment of the present invention.

[0039] 3, the battery capacity estimation device according to an embodiment of the present invention first determines whether the battery cell is currently in an idle state (i.e., current=0) (S310). If the battery cell is in an idle state, it is checked whether voltage data has been logged according to a preset method (S320). For example, the logging method may be to log four pieces of voltage data at five-minute intervals.

[0040] If the voltage data does not comply with the prescribed logging method (NO), the data is classified as first data (S330). On the other hand, if the voltage data is normally logged according to the prescribed method (YES), the differential data (dV / dt) of the voltage data with respect to time is calculated (S340).

[0041] Then, characteristic (principal component) data is extracted from the voltage differential data through principal component analysis (PCA). At this time, by performing principal component analysis on the differential data, the dimension of the data can be reduced. Then, the data is classified into multiple clusters by applying k-means clustering to the principal component data (S360).

[0042] At this time, if the classified data is not included in a specific cluster (e.g., cluster 2) (NO), it is classified as the first data (S370). On the other hand, if the data is included in a specific cluster (YES), it is classified as the second data (S380).

[0043] In this manner, the battery capacity estimation device according to an embodiment of the present invention can check the time unit of the logging data and classify the voltage data through the principal component analysis and the k-means clustering algorithm, thereby classifying data with similar characteristics, thereby making it possible to estimate the battery capacity more efficiently and accurately.

[0044] FIG. 4 is a diagram showing the correlation between the logging data and the capacity of a battery cell.

[0045] The graph on the left side of FIG. 4 shows the correlation between the differential data (dV1, dV2, dV3, dV4) (x-axis) of the voltage after charging and discharging over time and the battery capacity value (e.g., SOH) (y-axis), and the table on the right side shows the logging data of the voltage after charging and discharging the battery.

[0046] Referring to FIG. 4, the x-axis represents dV1, dV2, dV3, and dV4, which are voltage differential data after charging, and dV1, dV2, dV3, and dV4, which are voltage differential data after discharging. In this case, in the graph of FIG. 4, the closer the y-axis value is to 1 or -1 and the shorter the vertical bar is, the higher the correlation with the battery capacity value. That is, it can be seen that the data (25 and 29) immediately after charging and discharging have a higher correlation with the battery capacity value than the other data. This is because the change in voltage immediately after charging and discharging is relatively clear compared to the voltage change thereafter.

[0047] FIG. 5 is a diagram showing the logging data of the battery cells classified into a plurality of clusters by the battery capacity estimation device according to an embodiment of the present invention.

[0048] 5, the x-axis represents the number of charge / discharge cycles of the battery cell, and the y-axis represents differential data (dV1 to dV4) of voltage with respect to time in an idle section after charging / discharging the battery cell. Also, cluster 1 represents the first data described above, and cluster 2 represents the second data described above.

[0049] 5, in the case of the first data belonging to cluster 1, the data is relatively discontinuous, that is, the first data may be data that is not easy to use for battery capacity analysis because the characteristics of the first data are irregular.

[0050] Meanwhile, it can be seen that the second data belonging to cluster 2 is shown in the form of a continuous and gentle curve compared to the first data. That is, the second data shows a relatively constant tendency compared to the first data, and may be data that is easy to analyze.

[0051] FIG. 6 is a diagram showing the logging data of all battery cells classified into a plurality of clusters by the battery capacity estimation device according to an embodiment of the present invention.

[0052] Referring to FIG. 6, the x-axis represents the number of charge / discharge cycles of the battery cell, and the y-axis represents differential data (dV2) of voltage with respect to time in an idle section after charging / discharging of all the battery cells.

[0053] As in the case of FIG. 5, when the data of all the battery cells are added up, it can be seen that the data belonging to cluster 1 has many data that are irregularly protruding at the top or bottom. Meanwhile, the data belonging to cluster 2 is shown in a gentler form than cluster 1. Therefore, it is possible to calculate a more accurate result when estimating the capacity of the battery.

[0054] FIG. 7 is a diagram illustrating an example of a capacity estimation model (LSTM) of a battery capacity estimation device according to an embodiment of the present invention.

[0055] The capacity estimation model shown in Figure 7 shows a long short-term memory network model (LSTM). In Figure 7, C is long-term information, h is information at the previous step, σ and tanh are activation functions, W is weighting, and b is noise. The LSTM in Figure 7 is composed of a forget gate layer, a decision layer, an update new state value, and an output value decision stage.

[0056] Specifically, the forget gate layer is a layer with output values ​​between 0 and 1, called h t-1 and x t The input determines what information to keep (1) or discard (0). The decision layer also determines what value to update and save the new state. At this time, the input gate layer determines what value to update, and generates a vector C of new candidate values ​​to be added to the cell state through the tanh layer.

[0057] In the new cell state update step, the old cell state (Ct-1) is updated to the new cell state (Ct). In the output value determination step, it is determined what value to output, and the cell state is subjected to a tanh function to extract a value between -1 and 1, and the output value is multiplied by the output value of the forget gate layer.

[0058] In this way, when the above-described time-series voltage data is input to the capacity estimation model shown in Fig. 7, the capacity value (e.g., capacity%) of the battery cell can be output through the above-described process. Meanwhile, since the LSTM model of Fig. 7 has a known configuration, detailed description thereof will be omitted.

[0059] Meanwhile, the long short-term memory network model (LSTM) of FIG. 7 is merely an example, and the capacity estimation model of the present invention is not limited thereto, and various estimation models may be used.

[0060] FIG. 8 is a flow chart illustrating a battery capacity estimation method according to one embodiment of the present invention.

[0061] 8, in the battery capacity estimation method according to an embodiment of the present invention, first, the voltage of the battery cell is measured (S810). At this time, in step S810, the voltage of the battery cell may be measured at a certain time interval. For example, the voltage may be measured during an idle period after charging or discharging the battery cell.

[0062] Then, if a logging pattern of the voltage data of the battery cells falls outside a preset reference range, the voltage data is determined as the first data (S820). Specifically, in step S820, if at least one of the logging time and the number of the voltage data of the battery cells falls outside a reference range, the voltage data may be determined as the first data. For example, the logging pattern of the voltage data of the battery cells may be one that extracts four voltage data every five minutes.

[0063] Next, second data is determined by statistical analysis of the voltage data of the battery cell (S830). In this case, in step S830, differential data of the voltage with respect to time (dV / dt) of the battery cell may be calculated. Specifically, in step S830, principal component analysis (PCA) may be performed on the differential data to extract principal component data for the voltage data. Also, a plurality of clusters may be calculated for the principal component data through k-means clustering, and if the differential data is included in a specific cluster, it may be determined as the second data, and the differential data not included in the cluster may be determined as the first data.

[0064] Finally, the capacity of the battery cell is estimated by applying the first data and the second data extracted from steps S820 and S830 to a capacity estimation model, respectively (S840). For example, the capacity estimation model may be the long short-term memory network (LSTM) model described above. In this case, the capacity estimation models of the first data and the second data may be the same, but the parameter values ​​input to the capacity estimation model may be different.

[0065] As described above, according to the battery capacity estimation method according to an embodiment of the present invention, the battery cell logging data is divided into data that is easy to analyze and data that is not easy to analyze, and a capacity estimation model is applied to each data, thereby making it possible to accurately and efficiently estimate the capacity of the battery cell.

[0066] FIG. 9 is a block diagram showing a hardware configuration of a battery abnormality diagnosis device according to one embodiment of the present invention.

[0067] Referring to FIG. 9, a battery capacity estimation device 900 according to an embodiment of the present invention may include an MCU 910, a memory 920, an input / output I / F 930, and a communication I / F 940.

[0068] The MCU 910 may be a processor that executes various programs (e.g., a battery capacity estimation program, a principal component analysis program, a k-means clustering program, etc.) stored in the memory 920, processes various data for data classification and capacity estimation of battery cells through such programs, and performs the functions of FIG. 2 described above.

[0069] The memory 920 may store various programs related to statistical analysis of the battery cells, capacity estimation, etc. The memory 920 may also store various data such as voltage data of the battery cells and differential data of the battery cell voltages.

[0070] A plurality of such memories 920 may be provided as necessary. The memory 920 may be a volatile memory or a non-volatile memory. As the volatile memory 920, a RAM, a DRAM, an SRAM, or the like may be used. As the non-volatile memory 920, a ROM, a PROM, an EAROM, an EPROM, an EEPROM, a flash memory, or the like may be used. The examples of the memory 920 and the like listed above are merely illustrative and are not limited to these examples.

[0071] The input / output I / F 930 can provide an interface that connects input devices (not shown) such as a keyboard, mouse, or touch panel and output devices such as a display (not shown) to the MCU 910 to transmit and receive data.

[0072] The communication I / F 940 may be any device capable of transmitting and receiving various data to and from a server and supporting wired or wireless communication. For example, a program for statistical analysis and capacity estimation of a battery cell and various data may be transmitted and received from a separately provided external server via the communication I / F 940.

[0073] In this manner, a computer program according to an embodiment of the present invention may be recorded in the memory 920 and processed by the MCU 910 to be embodied as a module that performs each of the functional blocks illustrated in FIG.

[0074] Although it has been described above that all components constituting the embodiments of the present invention are combined into one or operate in combination, the present invention is not necessarily limited to such an embodiment, that is, all components may be selectively combined into one or more and operate within the scope of the present invention.

[0075] In addition, the terms "comprise", "constitute" or "have" described above mean that the relevant element may be contained, unless otherwise specified, and should be interpreted as including other elements, not excluding other elements. All terms, including technical and scientific terms, unless otherwise defined, should be interpreted as having the same meaning as commonly understood by a person of ordinary skill in the art to which the present invention belongs. Terms commonly used, such as terms defined in a dictionary, should be interpreted to be consistent with the contextual meaning of the relevant art, and should not be interpreted as idealized or overly formal, unless expressly defined in the present invention.

[0076] The above description is merely illustrative of the technical idea of ​​the present invention, and various modifications and variations are possible within the scope of the essential characteristics of the present invention if one has ordinary knowledge in the technical field to which the present invention belongs. Therefore, the embodiments disclosed in the present invention are for the purpose of explanation, not for the purpose of limiting the technical idea of ​​the present invention, and the scope of the technical idea of ​​the present invention is not limited by such embodiments. The scope of protection of the present invention should be interpreted according to the following claims, and all technical ideas within the scope equivalent thereto should be interpreted as being included in the scope of the present invention. [Explanation of symbols]

[0077] 1 Battery Rack 2 Upper controller 10 Battery Module 12 Sensors 14 Switching section 20 Battery Management System 200 Battery capacity estimator 210 Voltage measurement section 220 Filtering section 230 Statistical Analysis Department 240 Capacity estimation part 900 Battery Capacity Estimator 910 MCU 920 Memory 930 Input / Output Interface 940 Communication I / F

Claims

1. A voltage measuring unit that measures the voltage of a battery cell; a filtering unit that determines the voltage data as first data when a logging pattern of the voltage data of the battery cell is out of a preset reference range; a statistical analysis unit for determining second data by performing a statistical analysis on voltage data of the battery cell; a capacity estimation unit that estimates a capacity by applying data classified by the filtering unit or the statistical analysis unit to a capacity estimation model generated by classifying the first data and the second data for a battery cell to be measured; comprising The statistical analysis unit performing a statistical analysis on data other than the data determined as the first data; Calculating differential data of the voltage of the battery cell with respect to time; the statistical analysis unit performs a principal component analysis on a plurality of differential data to extract the plurality of differential data as one principal component data; the statistical analysis unit calculates a plurality of clusters for the principal component data through k-means clustering, and when the differential data is included in a preset cluster, determines it as the second data; The statistical analysis unit determines differential data not included in the cluster as the first data; The first data is data in which voltage data of the battery cell is represented in a discontinuous form, The second data is data in which voltage data of the battery cell is represented in a continuous form.

2. The battery capacity estimation device according to claim 1 , wherein the voltage measurement unit measures a voltage during an idle period after charging or discharging the battery cell.

3. 3. The battery capacity estimation device according to claim 1, wherein the filtering unit determines the voltage data as the first data when at least one of a logging time and a number of the voltage data of the battery cells falls outside the reference range due to an error or a missing logging data.

4. measuring a voltage of a battery cell; determining, when a logging pattern of the voltage data of the battery cell is outside a preset reference range, the voltage data as first data; statistically analyzing data other than the data determined as the first data to determine second data; estimating a capacity of the battery cell by applying the first data and the second data to a capacity estimation model, respectively; comprising The statistical analysis step includes: Calculating differential data of the voltage of the battery cell with respect to time; The step of statistically analyzing further includes the step of performing a principal component analysis on the differential data to extract a plurality of differential data as one principal component data, The statistical analysis step includes calculating a plurality of clusters for the principal component data through k-means clustering, and determining the differential data as the second data if the differential data is included in a predetermined cluster; The step of statistically analyzing further includes determining differential data not included in the cluster as the first data; The first data is data in which voltage data of the battery cell is represented in a discontinuous form, The battery capacity estimation method, wherein the second data is voltage data of the battery cell represented in a continuous form.

Images