Battery management device and battery inspection system including same

Abstract

A battery management device according to one embodiment disclosed in this document may include a first controller that acquires status data including measured values ​​that measure a state of a battery, a second controller that applies machine learning to at least a portion of the status data to generate prediction data that predicts the state of the battery, and compares the prediction data with the status data of the battery to determine the state of the battery, and a communication unit that transmits the status data to a server based on the result of determining the state of the battery.

Description

[Technical field]

[0001] The embodiments disclosed in this document claim the benefit of priority based on Korean Patent Application No. 10-2021-0107250, filed on August 13, 2021, and all contents disclosed in the documents of this Korean patent application are incorporated as part of this specification.

[0002] SUMMARY OF THE DISCLOSURE The embodiments disclosed herein relate to a battery management device and a battery inspection system including the same. [Background technology]

[0003] Electric vehicles obtain power by charging the battery with electricity from an external source and then driving the motor with the voltage charged in the battery. The battery of an electric vehicle may generate heat due to chemical reactions that occur during the process of charging and discharging electricity, and this heat may damage the performance and lifespan of the battery. Therefore, a Battery Management System (BMS) that monitors the temperature, voltage, and current of the battery is operated to diagnose the battery condition.

[0004] However, when a high-performance GPU or NPU is installed in the battery management device to precisely analyze the battery state in real time, there is a problem that the efficiency of the battery management device is reduced and the cost increases. In addition, when the battery management device transmits all the battery data measured by the battery management device to a cloud server for analysis, there is a problem that infrastructure installation costs and communication costs are incurred, and efficiency is reduced in the process of managing and analyzing a large amount of data. Summary of the Invention [Problem to be solved by the invention]

[0005] One objective of the embodiments disclosed herein is to provide a battery management device and a battery inspection system including the same that can precisely analyze only data of batteries that are determined to be in an abnormal state, thereby improving the efficiency of the battery management device and reducing infrastructure installation costs and communication costs.

[0006] The technical problems of the embodiments disclosed in this document are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the following description. [Means for solving the problem]

[0007] A battery management device according to one embodiment disclosed in this document may include a first controller that acquires status data including measured values ​​that measure a state of a battery, a second controller that applies machine learning to at least a portion of the status data to generate prediction data that predicts the state of the battery, and compares the prediction data with the status data of the battery to determine the state of the battery, and a communication unit that transmits the status data to a server based on the result of determining the state of the battery.

[0008] In one embodiment, when the battery is determined to be in an abnormal state, the second controller can compress the status data acquired during a predetermined time period before and after the point in time when the battery is determined to be in an abnormal state. In one embodiment, the communication unit may transmit the compressed battery status data to the server.

[0009] In one embodiment, the measurements may include cumulatively measured voltage, current, and temperature of the battery, and the status data may include the measurements and a calculated State of Health (SOH) of the battery based on the measurements.

[0010] In one embodiment, the prediction data includes a voltage of the battery, and the second controller can apply past data included in the status data and currently measured current and temperature of the battery included in the status data to the machine learning model to predict the voltage of the battery.

[0011] In one embodiment, the second controller can apply the state data to a Long Short Term Memory (LSTM) algorithm to predict the voltage of the battery.

[0012] In one embodiment, the second controller can apply the status data and the prediction data to an anomaly detection algorithm to generate an anomaly score.

[0013] In one embodiment, the second controller can determine that the battery is in an abnormal state if the proportion of abnormal values ​​generated during a specified time that are equal to or greater than a threshold value is equal to or greater than a certain proportion.

[0014] A battery inspection system according to one embodiment disclosed in this document may include a battery management device that applies machine learning to at least a portion of status data including measurement values ​​that measure the status of a battery to generate prediction data that predicts the status of the battery, compares the prediction data with the status data of the battery to determine the status of the battery, and transmits the status data to a server based on the determination result, and a server that determines whether or not the battery is in an abnormal state based on the status data of the battery.

[0015] In one embodiment, when the battery is determined to be in an abnormal state, the battery management device may compress the status data acquired during a predetermined time before and after the time when the battery is determined to be in an abnormal state and transmit the compressed status data to the server.

[0016] In one embodiment, the measurements may include cumulatively measured voltage, current, and temperature of the battery, and the status data may include the measurements and a calculated State of Health (SOH) of the battery based on the measurements.

[0017] In one embodiment, the battery management device can predict the voltage of the battery by applying the past data included in the status data and the currently measured current and temperature of the battery included in the status data to the Long Short Term Memory (LSTM) algorithm.

[0018] In one embodiment, the battery management device applies the status data and the prediction data to an anomaly detection algorithm to generate an anomaly score, and if the proportion of the anomaly scores that are equal to or greater than a threshold among the anomaly scores generated during a predetermined period of time is equal to or greater than a certain proportion, the battery can be determined to be in an abnormal state.

[0019] In one embodiment, the server can extract a false alarm, which is a case where the battery management device determines that the battery is in an abnormal state even though the battery is not actually in an abnormal state.

[0020] In one embodiment, the server may measure and transmit to the battery management device the SOH value of the battery and the threshold of the anomaly detection algorithm if the frequency of the false alarms is equal to or greater than a threshold frequency.

[0021] In one embodiment, the battery management unit can update a long-term and short-term memory algorithm and anomaly detection algorithm based on the SOH value and the threshold value received from the server. Effect of the Invention

[0022] According to an embodiment of a battery management device and a battery inspection system including the same disclosed in this document, it is possible to precisely analyze only data of batteries that are determined to be in an abnormal state, thereby improving the efficiency of the battery management device and reducing infrastructure installation costs and communication costs. [Brief description of the drawings]

[0023] [Figure 1] FIG. 1 is a diagram showing the configuration of a battery pack according to one embodiment disclosed in this document. [Diagram 2] 1 is a block diagram showing a configuration of a battery management device according to an embodiment disclosed in this document. [Diagram 3] FIG. 1 is a diagram generally illustrating a battery management device according to an embodiment disclosed herein. [Figure 4] FIG. 1 is a block diagram showing a configuration of a battery inspection system according to an embodiment disclosed in this document. [Diagram 5] 1 is a flow chart illustrating a method of operation of a battery testing system according to one embodiment disclosed herein. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] Some embodiments disclosed in this document will be described in detail below with reference to exemplary drawings. When referring to components in each drawing, it should be noted that the same components are referred to by the same reference numerals as much as possible when they are displayed in other drawings. In addition, when describing the embodiments disclosed in this document, if a detailed description of related known configurations or functions is deemed to hinder understanding of the embodiments disclosed in this document, the detailed description will be omitted.

[0025] In describing the components of the embodiments disclosed in this document, terms such as first, second, A, B, (a), (b) and the like may be used. Such terms are merely used to distinguish the components from other components, and do not limit the essence, order, or sequence of the components. In addition, unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments disclosed in this document belong. Terms as defined in commonly used dictionaries should be interpreted as having a meaning consistent with the contextual meaning of the relevant art, and should not be interpreted in an ideal or overly formal sense unless expressly defined in this document.

[0026] FIG. 1 is a diagram illustrating a battery pack according to one embodiment disclosed herein. Referring to FIG. 1, a battery pack 100 according to one embodiment disclosed herein may include a battery module 110, a battery management unit 120, and a relay 130.

[0027] The battery module 110 may include a first battery cell 111, a second battery cell 112, a third battery cell 113, and a fourth battery cell 114. Although the number of battery cells is illustrated as four in FIG. 1, the number is not limited thereto, and the battery module 110 may include n (n is a natural number equal to or greater than 2) battery cells.

[0028] The battery module 110 may supply power to a target device (not shown). To this end, the battery module 110 may be electrically connected to the target device. Here, the target device may include an electrical, electronic, or mechanical device that operates by receiving power from a battery pack 100 including a plurality of battery cells 111, 112, 113, and 114, and may be, for example, but is not limited to, an electric vehicle (EV).

[0029] The plurality of battery cells 111, 112, 113, 114 may be, but are not limited to, lithium ion (Li-ion) batteries, lithium ion polymer (Li-ion polymer) batteries, nickel cadmium (Ni-Cd) batteries, nickel metal hydride (Ni-MH) batteries, etc. Meanwhile, although FIG. 1 shows one battery module 110, depending on the embodiment, a plurality of battery modules 110 may be configured.

[0030] A battery management system (BMS) 120 can manage and / or control the state and / or operation of the battery module 110. For example, the battery management system 120 can manage and / or control the state and / or operation of the multiple battery cells 111, 112, 113, 114 included in the battery module 110. The battery management system 120 can manage charging and / or discharging of the battery module 110.

[0031] Furthermore, the battery management device 120 can monitor the voltage, current, temperature, etc. of the battery module 110 and / or each of the multiple battery cells 111, 112, 113, 114 included in the battery module 110. For monitoring via the battery management device 120, sensors and various measurement modules (not shown) can be further provided in the battery module 110, a charge / discharge path, or any position on the battery module 110. The battery management device 120 can calculate parameters indicating the state of the battery module 110, such as SOC (State of Charge) and SOH (State of Health), based on the measured values ​​of the monitored voltage, current, temperature, etc.

[0032] The battery management unit 120 can control the operation of the relay 130. For example, the battery management unit 120 can short-circuit the relay 130 to supply power to a target device. In addition, the battery management unit 120 can short-circuit the relay 130 when a charging device is connected to the battery pack 100.

[0033] The battery management unit 120 can calculate the cell balancing time for each of the multiple battery cells 111, 112, 113, and 114. Here, the cell balancing time can be defined as the time required for balancing the battery cells. For example, the battery management unit 120 can calculate the cell balancing time based on the SOC (State of Charge), battery capacity, and balancing efficiency of each of the multiple battery cells 111, 112, 113, and 114.

[0034] The configuration and operation method of the above-mentioned battery management device 120 will be specifically described below with reference to FIGS. Fig. 2 is a block diagram showing the configuration of a battery management device according to an embodiment disclosed in the present document. Fig. 3 is a diagram for generally explaining a battery management device according to an embodiment disclosed in the present document.

[0035] Referring to FIG. 2, the battery management device 120 may include a first controller 121, a second controller 122, and a communication unit 123.

[0036] In FIG. 2, the battery module 110 can include a plurality of battery cells 111, 112, 113, and 114, but the following description will be given taking the first battery cell 111 as an example.

[0037] 3, the first controller 121 may obtain status data including measurement values ​​measuring the status of the first battery cell 111. Here, the measurement values ​​may include cumulatively measured voltage, current, and temperature of the first battery cell 111. For example, the status data may include measurement values ​​including the measured voltage, current, and temperature of the first battery cell 111, and a State of Health (SOH) of the first battery cell 111 calculated based on the measurement values.

[0038] For example, the first controller 121 can obtain measurement values ​​measuring the voltage, current, and temperature of the first battery cell 111 from time tN to time t-1, and can obtain the SOH of the first battery cell 111 from time tN to time t-1 based on the measurement values, where N can be defined as any constant value indicating the amount of change in the variable t, which means time.

[0039] For example, the first controller 121 may calculate the voltage V of the first battery cell 111 at time tN. t-N , current I t-N , temperature T t-N The SOH value of the first battery cell 111 at time tN can be obtained, and the voltage V of the first battery cell 111 at time t-(N-1) can be obtained. t-(N-1) , current I t-(N-1) , temperature T t-(N-1) and obtain the SOH value of the first battery cell 111 at time t-(N-1). That is, the first controller 121 can obtain the past state data of the first battery cell 111 from time tN to time t-1. In addition, the first controller 121 calculates the current I of the first battery cell 111 measured at the current time t. t and temperature T t Measurements of can be obtained.

[0040] The second controller 122 can apply machine learning to at least a part of the state data to generate prediction data that predicts the state of the first battery cell 111. Here, the prediction data is the voltage V' of the first battery cell 111. t may include.

[0041] Specifically, the second controller 122 may apply the past data included in the status data and the current and temperature of the first battery cell 111 measured at the present time included in the status data to a machine learning model of a time series analysis structure to predict the voltage of the first battery cell 111. A time series may be defined as a sequence arranged at regular time intervals along the flow of time. The machine learning model of the time series analysis structure may include a long short-term memory (LSTM) algorithm. The long short-term memory algorithm may be defined as a structure in which a recurrent neural network (RNN), which is one of the artificial neural networks, is improved to reflect long-term features. The long short-term memory algorithm is mainly applied to problems of time series data that show phenomena that change over time. The long short-term memory algorithm may predict what the n+1th data is for measured time series data of an object having a length of n.

[0042] The second controller 122 applies the voltage, current, temperature, and SOH data of the first battery cell 111 from time tN to time t−1 and the current and temperature data of the first battery cell 111 measured at the current time t to a long-term and short-term memory algorithm to obtain a voltage V′ of the first battery cell 111 at the current time t. t That is, the second controller 122 can apply the state data of the first battery cell 111 from time tN to time t−1 to the long-term and short-term memory algorithm to generate prediction data predicting the voltage of the first battery cell 111 at the current time t.

[0043] The second controller 122 can compare the predicted data with the status data to determine the status of the first battery cell 111. Specifically, the second controller 122 can obtain a predicted voltage V′ of the first battery cell 111 at time t. t and the measured voltage V of the first battery cell 111 at time t tThe anomaly detection algorithm can be applied to generate an anomaly score to analyze whether the battery is in an abnormal state. Here, the anomaly detection algorithm can be defined as an artificial intelligence model that analyzes the anomaly score of an individual that shows different patterns and characteristics from data defined as normal.

[0044] That is, the second controller 122 calculates the predicted voltage V′ of the first battery cell 111 at time t. t and the measured voltage V of the first battery cell 111 at time t t It is possible to generate an outlier that is the difference between

[0045] The second controller 122 can determine that the first battery cell 111 is in an abnormal state when the ratio of abnormal values ​​that are equal to or greater than a threshold among the abnormal values ​​generated during a predetermined time is equal to or greater than a certain ratio. That is, the second controller 122 can determine that the first battery cell 111 is in an abnormal state when the predicted voltage V' of the first battery cell 111 generated during a predetermined time is equal to or greater than a certain ratio. t and the measured voltage V of the first battery cell 111. t When the ratio of the abnormal value exceeding the threshold value among the abnormal values, which are the differences between the first battery cell 111 and the battery cell 112, is maintained for several seconds to several tens of seconds, the first battery cell 111 can be determined to be in an abnormal state.

[0046] When the first battery cell 111 is determined to be in an abnormal state, the second controller 122 can compress the state data acquired during a predetermined time before and after the point in time when the first battery cell 111 is determined to be in an abnormal state.

[0047] The communication unit 123 can transmit status data to the server 200 based on the result of determining the status of the first battery cell 111 of the second controller 122. That is, when it is determined that the first battery cell 111 is in an abnormal state, the communication unit 123 can transmit the compressed status data of the first battery cell 111 to the server 200 using an OTA (Over The Air) technology. Here, the OTA technology is a firmware update method, and can be defined as a technology for updating firmware wirelessly using Wi-Fi (registered trademark) or the like without connecting to a computer.

[0048] As described above, the battery management device 120 and battery inspection system 1000 according to one embodiment disclosed in this document can transmit battery data to a server when a battery is determined to be in an abnormal state, thereby improving the diagnostic efficiency of the battery management device 120 and preventing unnecessary power supply from the battery management device 120.

[0049] In addition, the battery management device 120 and the battery inspection system 1000 first analyze battery data using a voltage prediction model designed in the battery management device, and if an abnormality is detected, the battery data is transmitted to a server with abundant computing resources for precise analysis, thereby precisely analyzing the abnormality in the battery. At the same time, the data transmitted to the server is drastically reduced, thereby reducing communication resources and alleviating the storage and computing power of the cloud computer included in the server.

[0050] Furthermore, the battery management device 120 and the battery inspection system 1000 transmit the minimum necessary battery data to the server without data loss, thereby reducing infrastructure investment costs and network costs for the server.

[0051] Fig. 4 is a block diagram showing a configuration of a battery inspection system according to an embodiment disclosed in the present document, and Fig. 5 is a flowchart showing an operation method of the battery inspection system according to an embodiment disclosed in the present document.

[0052] The configuration and operation of the battery inspection system 1000 will be described below with reference to FIGS. Referring to FIG. 4, a battery inspection system 1000 according to one embodiment disclosed in this document may include a battery management device 120 and a server 200.

[0053] The battery management device 120 may be substantially the same as the battery management device 120 described with reference to FIGS. 1 to 3, and therefore will be described below in brief to avoid duplication of description.

[0054] 5, the operation method of the battery inspection system includes a step (S101) in which the battery management unit 120 acquires status data including a measurement value obtained by measuring the status of the first battery cell 111, a step (S102) in which the battery management unit 120 applies the status data to a long-term and short-term memory algorithm to generate prediction data, a step (S103) in which the battery management unit 120 inputs the prediction data and the status data into an anomaly detection algorithm to determine the status of the first battery cell 111, a step (S104) in which the battery management unit 120 determines whether the first battery cell 111 is in an abnormal state, and if the first battery cell 111 is in an abnormal state, the battery management unit 120 The method may include a step of transmitting status data of the cell 111 to the server 200 (S105), a step of the server 200 extracting false alarms (S106), a step of the server 200 determining whether the frequency of false alarms is equal to or greater than a threshold frequency (S107), and if the frequency of false alarms is equal to or greater than the threshold frequency, a step of the server 200 measuring the SOH value of the first battery cell 111 and the threshold of the anomaly detection algorithm and transmitting them to the battery management device 120 (S108), and a step of the battery management device 120 applying the SOH value and threshold received from the server 200 to the long-term and short-term memory algorithm and the anomaly detection algorithm (S109).

[0055] Steps S101 to S109 will be specifically described below. In step S101, the battery management unit 120 can acquire status data including measurement values ​​obtained by measuring the status of the first battery cell 111. Here, the status data can include measurement values ​​including the measured voltage, current, and temperature of the first battery cell 111, and a State of Health (SOH) of the first battery cell 111 calculated based on the measurement values. That is, the battery management unit 120 can acquire past status data of the first battery cell 111 from time tN to time t-1. In addition, the battery management unit 120 can acquire measurement values ​​of the current and temperature of the first battery cell 111 measured at the current time t.

[0056] In step S102, the battery management unit 120 can apply machine learning to at least a part of the state data of the first battery cell 111, and generate prediction data that predicts the state of the first battery cell 111.

[0057] For example, in step S102, the battery management unit 120 applies the voltage, current, temperature, and SOH data of the first battery cell 111 from time tN to time t-1 and the current and temperature data of the first battery cell 111 measured at the current time t to a long-term and short-term memory algorithm to obtain a voltage V' of the first battery cell 111 at the current time t. t can be predicted.

[0058] In step S103, the battery management unit 120 compares the predicted data with the state data to determine the state of the first battery cell 111. Specifically, the battery management unit 120 calculates the predicted voltage V′ of the first battery cell 111 at time t by t and the measured voltage V of the first battery cell 111 at time t t The above is applied to the anomaly detection algorithm to generate an anomaly score, which can be used to analyze whether the battery is in an abnormal state.

[0059] In step S104, the battery management unit 120 can determine that the first battery cell 111 is in an abnormal state if the rate of abnormal values ​​that are equal to or greater than a threshold among the abnormal values ​​generated during a predetermined time period is equal to or greater than a certain rate.

[0060] In step S105, when the first battery cell 111 is determined to be in an abnormal state, the battery management device 120 can compress status data acquired during a predetermined time before and after the point at which the first battery cell 111 is determined to be in an abnormal state, and transmit the compressed status data to the server 200.

[0061] The server 200 may be defined as a determination device that receives the status data of the first battery cell 111 and determines whether or not the first battery cell 111 is defective. The server 200 may include, for example, cloud computing technology.

[0062] The server 200 may include an artificial intelligence model that extracts a false alarm, which is a case where the first battery cell 111 is determined to be in an abnormal state when in fact it is not, based on the status data of the first battery cell 111 determined by the battery management device 120.

[0063] In step S106, if the battery management unit 120 compresses and transmits to the server 200 the status data of the first battery cell 111 acquired during a predetermined time before and after the time when the battery management unit 120 determines that the battery cell is in an abnormal state, the server 200 can receive the compressed status data transmitted by the battery management unit 120.

[0064] In step S106, the server 200 analyzes the compressed data received from the battery management device 120, that is, the status data acquired during a predetermined time period before and after the point at which the abnormal state is determined, and can extract a false alarm, which is a case where the first battery cell 111 is determined to be in an abnormal state even though it is not actually in an abnormal state.

[0065] In step S107, the server 200 can determine whether the frequency of false alarms is equal to or greater than a threshold frequency. In step S108, if the frequency of false alarms is greater than or equal to the threshold frequency, the server 200 may recalculate the SOH value of the first battery cell 111 included in the battery management unit 120, reset the threshold of the abnormality detection algorithm, and transmit it to the battery management unit 120.

[0066] That is, in step S108, the server 200 can analyze the cause of the false alarm for the first battery cell 111 based on the SOH value of the first battery cell 111 input into the long-term and short-term memory algorithm of the battery management unit 120 and the threshold value of the abnormality detection algorithm.

[0067] The long and short term memory algorithm of the battery management device 120 receives input of measured values ​​of the voltage, current, and temperature of the first battery cell 111 as past data included in the status data of the first battery cell 111, and the SOH value of the first battery cell 111 based on the measured values. Therefore, if an error occurs in the process in which the battery management device 120 calculates the SOH value of the first battery cell 111 based on the voltage, current, and temperature of the first battery cell 111, an error may occur in the long and short term memory algorithm, and such an error may generate a false alarm in the battery inspection system 1000. Therefore, the server 200 may recalculate the SOH value of the first battery cell 111 input to the battery management device 120 and transmit it to the battery management device 120.

[0068] In addition, in step S108, the server 200 can analyze the cause of the false alarm for the first battery cell 111 based on the threshold value of the abnormality detection algorithm of the battery management unit 120.

[0069] The battery management unit 120 uses an abnormality detection algorithm to calculate the predicted voltage V′ of the first battery cell 111 at time t. t and the measured voltage V of the first battery cell 111 at time t tand generates an abnormal value which is a difference between the abnormal value and the threshold value, and if the ratio of the abnormal values ​​which are equal to or greater than the threshold value is equal to or greater than a certain ratio, the first battery cell 111 can be determined to be in an abnormal state. Therefore, if the threshold value for classifying the abnormal value as an abnormal value by the battery management unit 120 is not appropriate, an error may occur in the battery management unit 120's determination of the abnormal state of the first battery cell 111, and such an error may generate a false alarm in the battery inspection system 1000. Therefore, the server 200 may re-set the threshold value of the abnormality detection algorithm input to the battery management unit 120 and transmit the re-set threshold value to the battery management unit 120.

[0070] In step S109, when the battery management unit 120 receives the SOH value of the first battery cell 111 and the threshold value of the abnormality detection algorithm from the server 200, it can update the long-term and short-term memory algorithm and the abnormality detection algorithm.

[0071] The above description is merely an illustrative example of the technical ideas of the present disclosure, and various modifications and variations can be made by a person having ordinary knowledge in the technical field to which the present disclosure pertains without departing from the essential characteristics of the present disclosure.

[0072] Therefore, the embodiments disclosed in this disclosure are intended to explain the technical idea of ​​the disclosure, not to limit it, and the scope of the technical idea of ​​the disclosure is not limited by such embodiments. The scope of protection of the disclosure should be interpreted according to the claims below, and all technical ideas within the scope equivalent thereto should be interpreted as being included in the scope of rights of the disclosure.

Claims

1. A first controller that acquires status data including measurements that measure a state of the battery; a second controller that applies machine learning to at least a portion of the state data to generate prediction data that predicts a state of the battery, and compares the prediction data with the state data of the battery to determine the state of the battery; a communication unit that transmits the status data to a server based on a result of determining the state of the battery; A battery management device comprising:

2. The battery management device according to claim 1 , wherein, when the battery is determined to be in an abnormal state, the second controller compresses the state data acquired during a predetermined time period before and after the time when the battery is determined to be in an abnormal state.

3. The battery management device according to claim 2 , wherein the communication unit transmits the compressed battery status data to the server.

4. 4. The battery management device according to claim 1, wherein the measurement values ​​include cumulatively measured voltage, current, and temperature of the battery, and the state data includes the measurement values ​​and a SOH (State of Health) of the battery calculated based on the measurement values.

5. the prediction data includes a voltage of the battery; The battery management device of claim 4, wherein the second controller applies past data included in the status data and the current and temperature of the battery measured at a current time included in the status data to the machine learning model to predict the voltage of the battery.

6. The battery management device of claim 5 , wherein the second controller applies the status data to a long-term and short-term memory algorithm to predict a voltage of the battery.

7. The battery management device of claim 4 , wherein the second controller applies the status data and the prediction data to an anomaly detection algorithm to generate an anomaly value.

8. The battery management device according to claim 7 , wherein the second controller determines that the battery is in an abnormal state when a proportion of the abnormal values ​​generated during a predetermined time that are equal to or greater than a threshold is equal to or greater than a certain proportion.

9. a battery management device that applies machine learning to at least a portion of status data including measurement values ​​that measure a state of the battery to generate prediction data that predicts a state of the battery, compares the prediction data with the status data of the battery to determine the state of the battery, and transmits the status data to a server based on the result of the determination; A server that determines whether or not the battery is in an abnormal state based on the battery status data; A battery inspection system comprising:

10. 10. The battery inspection system of claim 9, wherein, when the battery is determined to be in an abnormal state, the battery management device compresses the status data acquired during a predetermined time before and after the time when the battery is determined to be in an abnormal state and transmits the compressed status data to the server.

11. 11. The battery inspection system according to claim 9, wherein the measurement values ​​include cumulatively measured voltage, current, and temperature of the battery, and the status data includes the measurement values ​​and a State of Health (SOH) of the battery calculated based on the measurement values.

12. The battery inspection system of claim 11, wherein the battery management device applies past data included in the status data and the current and temperature of the battery measured at the current time included in the status data to a long-term and short-term memory algorithm to predict the voltage of the battery.

13. The battery inspection system of claim 11, wherein the battery management device applies the status data and the prediction data to an anomaly detection algorithm to generate an abnormal value, and if a proportion of the abnormal values ​​generated within a specified time that are equal to or greater than a threshold value is equal to or greater than a certain proportion, the battery is judged to be in an abnormal state.

14. The battery inspection system of claim 13, wherein the server analyzes the process of calculating the status data based on the measurement values ​​obtained during a predetermined time before and after the time when the battery is determined to be in an abnormal state, and extracts a false alarm when the battery management device determines that the battery is in an abnormal state when in fact it is not.

15. The battery inspection system of claim 14 , wherein the server measures the SOH value of the battery and the threshold of the anomaly detection algorithm and transmits them to the battery management device when the frequency of the false alarm is equal to or greater than a threshold frequency.

16. The battery inspection system of claim 15 , wherein the battery management device updates a long-term and short-term memory algorithm and anomaly detection algorithm based on the SOH value and the threshold value received from the server.

Images