Apparatus and method for inspecting defect of battery

Abstract

A battery defect diagnosis device according to an embodiment of the present invention can obtain a defect determination model based on the inspection accuracy of each of a plurality of defect determination candidate models to which at least one different condition is applied, and can determine whether at least one battery among a plurality of batteries in which an activation process has been performed is defective by using the defect determination model.

Description

Technology Field

[0001] The present invention relates to a battery defect diagnosis device and method, and more specifically, to a battery defect diagnosis device and method for providing a defect determination model with high inspection accuracy when determining defects in a battery during an activation process. Background Technology

[0002] As the price of energy sources rises due to the depletion of fossil fuels and concerns about environmental pollution intensify, the demand for secondary batteries as an eco-friendly alternative energy source is rapidly increasing.

[0003] Due to their ability to be repeatedly charged and regenerated, these secondary batteries are being applied in a wide range of sectors—from small devices such as mobile phones and laptops to large-scale industries like automobiles, robots, and energy storage devices—as a response to today's environmental regulations and high oil prices.

[0004] Meanwhile, among secondary batteries, those used in various fields undergo performance testing before shipment at the final stage of the manufacturing process.

[0005] Conventional battery performance testing involved fully charging each battery cell and measuring its capacity until it was fully discharged.

[0006] However, the inspection method of fully charging and discharging battery cells requires a significant amount of process time, reduces productivity due to increased usage frequency of charging and discharging equipment, and presents difficulties in 100% measurement. Prior art literature

[0007] Korean Registered Patent No. 10-2064459 The problem to be solved

[0008] The objective of the present invention to solve the above-mentioned problems is to provide a battery defect diagnosis device.

[0009] Another objective of the present invention to solve the above-mentioned problems is to provide a battery defect diagnosis method using a battery defect diagnosis device. means of solving the problem

[0010] A battery defect diagnosis device according to an embodiment of the present invention for achieving the above objective comprises at least one processor and a memory storing at least one instruction performed through the processor, wherein the at least one instruction comprises a instruction to obtain a defect determination model based on the inspection accuracy of each of a plurality of defect determination candidate models to which at least one different condition is applied, and a instruction to determine whether at least one battery among a plurality of batteries in which an activation process has been performed is defective using the defect determination model.

[0011] Here, the command to obtain at least one defect determination model may include a command to individually input voltage data of a plurality of batteries that have undergone aging into the plurality of defect determination candidate models to verify the inspection accuracy of each of the defect determination candidate models, and a command to determine the defect determination model using at least one target model with high inspection accuracy among the defect determination candidate models.

[0012] In addition, the above-mentioned plurality of defect judgment candidate models each include: a time point for measuring voltage change amount data of the plurality of batteries on which an activation process has been performed; a method for calculating representative voltage change amount data of the plurality of batteries; and a standard deviation application criterion between the voltage change amount data of the plurality of batteries. And detailed criteria based on at least one condition regarding whether to filter noise for the voltage change amount data of the plurality of batteries may be applied differently.

[0013] Meanwhile, a command to verify the inspection accuracy of each of the above-mentioned defect judgment candidate models may include a command to individually calculate voltage change amount data for the plurality of batteries stored for a predetermined period at a predetermined reference temperature, a command to pre-diagnose whether a defect has occurred in each of the plurality of batteries based on the voltage change amount data, and a command to apply each of the above-mentioned defect judgment candidate models to the plurality of batteries, calculate a defect index according to each of the above-mentioned defect judgment candidate models based on the pre-diagnosis results, and verify the inspection accuracy.

[0014] At this time, the defect index can be calculated by considering the defect detection results of the plurality of batteries to which the defect judgment candidate models are individually applied, using at least one battery among the plurality of batteries that is defective, over-detected, and not detected.

[0015] More specifically, the defect index can be calculated by adding the number of at least one undetected battery to the sum of the defect rate and the over-detection rate of at least one battery.

[0016] Meanwhile, according to one embodiment, a command to determine the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models may include a command to obtain at least one target model with high inspection accuracy among the plurality of defect judgment candidate models, and a command to determine the target model as the defect judgment model when the target model is one.

[0017] Additionally, according to another embodiment, a command to determine the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models may include a command to obtain the at least one target model that has been calculated to have high inspection accuracy among the plurality of defect judgment candidate models; a command to check the detailed criteria for configuration conditions from each of the target models when there are multiple target models; and a command to extract the detailed criteria most frequently applied to the target models and to determine the defect judgment candidate model to which the extracted detailed criteria are applied as the defect judgment model.

[0019] A battery defect diagnosis method according to another embodiment of the present invention for achieving the above objective comprises the steps of: obtaining a defect determination model based on the inspection accuracy of each of a plurality of defect determination candidate models to which at least one different condition is applied; and determining whether at least one battery among a plurality of batteries in which an activation process has been performed is defective using the defect determination model.

[0020] Here, the step of obtaining at least one defect determination model may include the step of individually inputting voltage data of a plurality of batteries that have undergone aging into the plurality of defect determination candidate models to verify the inspection accuracy of each of the defect determination candidate models, and the step of determining the defect determination model using at least one target model with high inspection accuracy among the defect determination candidate models.

[0021] In addition, the above-mentioned plurality of defect judgment candidate models each include: a time point for measuring voltage change amount data of the plurality of batteries on which an activation process has been performed; a method for calculating representative voltage change amount data of the plurality of batteries; and a standard deviation application criterion between the voltage change amount data of the plurality of batteries. And detailed criteria based on at least one condition regarding whether to filter noise for the voltage change amount data of the plurality of batteries may be applied differently.

[0022] Meanwhile, a command to verify the inspection accuracy of each of the above-mentioned defect judgment candidate models may include the steps of: individually calculating voltage change amount data for the plurality of batteries stored for a predetermined period at a predetermined reference temperature; pre-diagnosing whether a defect has occurred in each of the plurality of batteries based on the voltage change amount data; and applying each of the above-mentioned defect judgment candidate models to the plurality of batteries, calculating a defect index according to each of the above-mentioned defect judgment candidate models based on the pre-diagnosis results, and verifying the inspection accuracy.

[0023] At this time, the defect index can be calculated by considering the defect detection results of the plurality of batteries to which the defect judgment candidate models are individually applied, using at least one battery among the plurality of batteries that is defective, over-detected, and not detected.

[0024] More specifically, the defect index can be calculated by adding the number of at least one undetected battery to the sum of the defect rate and the over-detection rate of at least one battery.

[0025] Meanwhile, according to one embodiment, the step of determining the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models may include the step of obtaining at least one target model with high inspection accuracy among the plurality of defect judgment candidate models, and, if there is only one target model, the step of determining the target model as the defect judgment model.

[0026] Additionally, according to another embodiment, the step of determining the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models may include: obtaining the at least one target model among the plurality of defect judgment candidate models that has a high inspection accuracy; if there are multiple target models, verifying detailed criteria for configuration conditions from each of the target models; and extracting the detailed criteria most frequently applied to the target models and determining the defect judgment candidate model to which the extracted detailed criteria are applied as the defect judgment model. Effects of the invention

[0027] The battery defect diagnosis device and method according to an embodiment of the present invention can enable highly reliable defect diagnosis by examining whether a defect has occurred in a battery based on the voltage change amount of a plurality of batteries in which an activation process has been performed for a predetermined period. Brief explanation of the drawing

[0028] FIG. 1 is a block diagram of a battery failure diagnosis device according to an embodiment of the present invention. FIG. 2 is a flowchart for explaining a battery failure diagnosis method operated by a processor in a battery failure diagnosis device according to an embodiment of the present invention. FIG. 3 is a flowchart illustrating a method for obtaining at least one defect determination model among battery defect diagnosis methods according to an embodiment of the present invention. FIG. 4 is a table for explaining the detailed criteria for each condition of defect judgment candidate models according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating a method for obtaining a defect judgment model by extracting at least one target model with high inspection accuracy among battery defect diagnosis methods according to an embodiment of the present invention. FIG. 6 is a scatter plot graph showing the first and second judgment results regarding whether a defect occurred in each of the plurality of batteries according to an embodiment of the present invention. FIG. 7 is a flowchart illustrating a method for obtaining a defect judgment model when there are multiple target models among the battery defect diagnosis methods according to an embodiment of the present invention. FIG. 8 is a table showing the defect indices of a plurality of defect judgment candidate models in the battery defect diagnosis method according to the experimental example of the present invention. FIG. 9 is a table counting detailed criteria for each condition extracted from a plurality of target models in a battery defect diagnosis method according to an embodiment of the present invention. Specific details for implementing the invention

[0029] The present invention is susceptible to various modifications and may have various embodiments; specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the invention to specific embodiments, and it should be understood that the invention includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention. Similar reference numerals have been used for similar components in the description of each drawing.

[0030] Terms such as first, second, A, B, etc., may be used to describe various components, but said components shall not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component. The term "and / or" includes a combination of a plurality of related described items or any of a plurality of related described items.

[0031] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0032] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, terms such as "comprising" or "having" are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0033] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0035] FIG. 1 is a block diagram of a battery failure diagnosis device according to an embodiment of the present invention.

[0036] Referring to FIG. 1, a battery defect diagnosis device (100) according to an embodiment of the present invention may be a device for diagnosing whether the battery is defective before shipment.

[0037] More specifically, the battery defect diagnosis device (100) can diagnose whether a defect has occurred in each of the plurality of batteries by applying a defect determination model based on voltage data of the plurality of batteries in which an activation process has been performed.

[0038] To describe the battery fault diagnosis device (100) in more detail according to its configuration, the battery fault diagnosis device (100) may include a memory (110), a processor (120), a transmission / reception device (130), an input interface device (140), an output interface device (150), and a storage device (160).

[0039] According to the embodiment, each component (110, 120, 130, 140, 150, 160) included in the battery failure diagnosis device (100) can communicate with each other by being connected by a bus (170).

[0040] Among the above configurations (110, 120, 130, 140, 150, 160), the memory (110) and the storage device (160) may be configured as at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory (110) and the storage device (160) may be configured as at least one of a read-only memory (ROM) and a random access memory (RAM).

[0041] Among these, the memory (110) may include at least one instruction executed by the processor (120). According to an embodiment, the at least one instruction includes a instruction to obtain a defect determination model based on the inspection accuracy of each of a plurality of defect determination candidate models to which at least one different condition is applied, and a instruction to determine whether at least one battery among a plurality of batteries in which an activation process has been performed is defective using the defect determination model.

[0042] Here, the command to obtain at least one defect determination model may include a command to individually input voltage data of a plurality of batteries that have undergone aging into the plurality of defect determination candidate models to verify the inspection accuracy of each of the defect determination candidate models, and a command to determine the defect determination model using at least one target model with high inspection accuracy among the defect determination candidate models.

[0043] In addition, the above-mentioned plurality of defect judgment candidate models each include: a time point for measuring voltage change amount data of the plurality of batteries on which an activation process has been performed; a method for calculating representative voltage change amount data of the plurality of batteries; and a standard deviation application criterion between the voltage change amount data of the plurality of batteries. And detailed criteria based on at least one condition regarding whether to filter noise for the voltage change amount data of the plurality of batteries may be applied differently.

[0044] Meanwhile, a command to verify the inspection accuracy of each of the above-mentioned defect judgment candidate models may include a command to individually calculate voltage change amount data for the plurality of batteries stored for a predetermined period at a predetermined reference temperature, a command to pre-diagnose whether a defect has occurred in each of the plurality of batteries based on the voltage change amount data, and a command to apply each of the above-mentioned defect judgment candidate models to the plurality of batteries, calculate a defect index according to each of the above-mentioned defect judgment candidate models based on the pre-diagnosis results, and verify the inspection accuracy.

[0045] At this time, the defect index can be calculated by considering the defect detection results of the plurality of batteries to which the defect judgment candidate models are individually applied, using at least one battery among the plurality of batteries that is defective, over-detected, and not detected.

[0046] More specifically, the defect index can be calculated by adding the number of at least one undetected battery to the sum of the defect rate and the over-detection rate of at least one battery.

[0047] Meanwhile, according to one embodiment, a command to determine the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models may include a command to obtain at least one target model with high inspection accuracy among the plurality of defect judgment candidate models, and a command to determine the target model as the defect judgment model when the target model is one.

[0048] Additionally, according to another embodiment, a command to determine the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models may include a command to obtain the at least one target model that has been calculated to have high inspection accuracy among the plurality of defect judgment candidate models; a command to check the detailed criteria for configuration conditions from each of the target models when there are multiple target models; and a command to extract the detailed criteria most frequently applied to the target models and to determine the defect judgment candidate model to which the extracted detailed criteria are applied as the defect judgment model.

[0049] Meanwhile, the processor (120) may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed.

[0050] As previously described, the processor (120) can execute at least one program command stored in memory (110).

[0051] Hereinafter, a battery defect diagnosis method operated by a processor in a battery defect diagnosis device according to an embodiment of the present invention will be described.

[0053] FIG. 2 is a flowchart for explaining a battery failure diagnosis method operated by a processor in a battery failure diagnosis device according to an embodiment of the present invention.

[0054] Referring to FIG. 2, a processor (120) in a battery defect diagnosis device (100) according to an embodiment of the present invention can obtain a defect determination model (S210). Here, the defect determination model may be a diagnosis model for determining whether a plurality of batteries that have undergone an activation process are defective.

[0055] The processor (120) can determine whether at least one target battery among a plurality of batteries in which an activation process has been performed is defective by using the above defect determination model (S220).

[0056] To explain more specifically according to the embodiment, the processor (120) can determine whether a defect has occurred in each of the plurality of batteries by inputting voltage data of each of the plurality of batteries in which the activation process has been performed into the defect determination model. Here, the voltage data may be voltage change amount data of each of the plurality of batteries measured before and after aging in the activation process.

[0058] FIG. 3 is a flowchart illustrating a method for obtaining at least one defect determination model among battery defect diagnosis methods according to an embodiment of the present invention.

[0059] Referring to FIG. 3, the processor (120) can obtain a plurality of different defect judgment candidate models to obtain the defect judgment model (S310).

[0060] To explain in more detail according to the embodiment, the processor (120) can obtain a plurality of different defect determination candidate models by applying controllable conditions differently in a basic formula model for determining defects in a target battery, such as [Equation 1] below.

[0062]

[0063] Here, △OCV target may be voltage change data of the target battery stored at a predetermined temperature for a predetermined aging period, and △OCV tray may be representative voltage change data of multiple batteries stored in the same tray as the target battery at the same temperature and for the same aging period. In addition, σ traymay be the standard deviation between the voltage change amount data of the plurality of batteries.

[0064] In other words, the above-mentioned controllable conditions include the timing of measurement of voltage change data of a plurality of batteries on which an activation process has been performed, the method of calculating representative voltage change data of the plurality of batteries, and the criteria for applying standard deviation between voltage change data of the plurality of batteries. and may include whether to filter noise on the voltage change amount data of the plurality of batteries.

[0065] Afterwards, the processor (120) can determine the defect judgment model by extracting at least one target model with high inspection accuracy among the plurality of defect judgment candidate models (S320).

[0067] FIG. 4 is a table for explaining the detailed criteria for each condition of defect judgment candidate models according to an embodiment of the present invention.

[0068] Referring to FIG. 4, the processor (120) can obtain a plurality of different defect judgment candidate models by setting at least one of the controllable conditional detailed criteria based on the basic formula model differently in step S310.

[0069] In summary, based on the above basic formula model, the detailed criteria for controllable conditions can be predefined as 6 for the first condition, 3 for the second condition, 3 for the third condition, and 2 for the fourth condition.

[0070] Accordingly, the processor (120) can obtain a total of 108 defect judgment candidate models by combining detailed criteria for each condition.

[0071] To explain more specifically according to the embodiment, among the above conditions, the first condition for controlling the timing of measuring voltage change data of a plurality of batteries may include six detailed calculation criteria.

[0072] For example, the first detailed calculation standard may be a standard for calculating the voltage change amount before and after aging for multiple batteries aged for 24 hours.

[0073] In addition, the second detailed calculation standard may be a standard for calculating voltage change data before and after aging for multiple batteries aged for 96 hours.

[0074] In addition, the third detailed calculation standard may be a standard for calculating voltage change data between the first point in time, 24 hours after aging is performed, and the second point in time, 96 hours after aging is performed.

[0075] Additionally, the fourth detailed calculation criterion may be a criterion that utilizes both the first voltage change amount data measured at the first point in time 24 hours after the aging is performed and the second voltage change amount data measured at the second point in time 72 hours after the first point in time. More specifically, the processor (120) individually inputs the first voltage change amount data and the second voltage change amount data calculated according to the fourth detailed calculation criterion into a defect judgment candidate model to which the second to fourth conditions are applied equally, and determines whether a defect has occurred in the target battery, and if at least one of the judgment results is a defect, it may be determined that a defect has occurred in the target battery.

[0076] Additionally, the fifth detailed calculation criterion may be a criterion that utilizes both the first voltage change amount data measured at the first point in time 24 hours after aging is performed and the second voltage change amount data measured at the second point in time 72 hours after the first point in time, just like the fourth detailed calculation criterion. More specifically, the processor (120) individually inputs the first voltage change amount data and the second voltage change amount data calculated according to the fifth detailed calculation criterion into a defect judgment candidate model to which the second to fourth conditions are applied equally, and determines whether a defect has occurred in the target battery, and if a defect is determined in both cases as a result of the determination, it may be determined that a defect has occurred in the target battery.

[0077] In addition, the 6th detailed calculation criterion may be a criterion for calculating voltage change data from before the start of the activation process to one week after the completion of the process.

[0078] Meanwhile, among the above conditions, the second condition controlling the method of calculating representative voltage change data of a plurality of batteries can be calculated in three ways (first to third calculation methods).

[0079] For example, the first calculation method is the mode value (△OCV) of the voltage change amount data of each of the multiple batteries. Mode It may be a method of calculating ) and applying it as representative voltage change amount data of the plurality of batteries.

[0080] In addition, the second calculation method is the average value (△OCV) of the voltage change amount data of each of the multiple batteries. Mean It may be a method of calculating ) and applying it as representative voltage change amount data of the plurality of batteries.

[0081] In addition, the third calculation method is the median value (△OCV) of the voltage change amount data of each of the multiple batteries. MedianIt may be a method of calculating ) and applying it as representative voltage change amount data of the plurality of batteries.

[0082] Meanwhile, among the above conditions, the third condition for setting the standard deviation of voltage change amount data between multiple batteries can be applied with three setting values ​​(first to third setting values). For example, the first setting value can be applied as 1, and the second setting value can be applied as the value of the outlier detection range (Interquartile Range, IQR). Here, the outlier detection range (IQR) may be a range set for detecting outliers in the distribution of voltage change amount data for the multiple batteries.

[0083] More specifically, the outlier detection area uses the quartiles obtained by dividing the distribution of voltage change data into four equal parts (Q1-Q4), Q2 (the 50th percentile, in other words, the representative voltage change data (△OCV) in the present invention tray Based on (corresponding to ), it may be the value of Q3 (the 75th percentile among the voltage change data of multiple batteries) - Q1 (the 25th percentile among the voltage change data of multiple batteries).

[0084] In addition, the third setting value can be applied by calculating the standard deviation value of the voltage change amount data between multiple batteries.

[0085] Meanwhile, among the above conditions, the fourth condition regarding whether to filter noise from voltage change amount data may include a first detailed criterion (raw data) that is not filtered and a second detailed criterion (excluding outliers) in which noise is filtered.

[0086] However, the battery defect diagnosis method of the present invention is not limited to the first to fourth conditions disclosed above, and at least one condition and detailed criteria according thereto may be additionally applied to improve the accuracy of defect determination of a plurality of batteries in which an activation process has been performed, such as the number of measurements.

[0088] FIG. 5 is a flowchart illustrating a method for obtaining a defect judgment model by extracting at least one target model with high inspection accuracy among battery defect diagnosis methods according to an embodiment of the present invention, and FIG. 6 is a scatter plot graph showing first and second judgment results regarding whether a defect occurs in each of a plurality of batteries according to an embodiment of the present invention.

[0089] Referring to FIG. 5, the processor (120) can calculate voltage change data for a plurality of batteries stored for a predetermined reference temperature and a predetermined period. Here, the predetermined reference temperature may be room temperature (20-25℃) and the predetermined period may be 8 weeks.

[0090] Afterwards, the processor (120) can relatively compare the voltage change data of the plurality of batteries to pre-check whether a defect has occurred in each of the plurality of batteries (S510).

[0091] According to an embodiment, the processor (120) can obtain a normal distribution curve for the voltage change amount data of the plurality of batteries. Subsequently, the processor (120) can identify the section to which the voltage change amount data of each of the plurality of batteries belongs in the normal distribution curve, and pre-determine whether the battery is defective or good.

[0092] Additionally, the processor (120) can individually apply the plurality of defect determination candidate models to the plurality of batteries for which defect occurrence has been pre-confirmed, and diagnose whether defects have occurred in each of the plurality of batteries for each of the plurality of defect determination candidate models (S520).

[0093] Afterwards, the processor (120) can individually calculate the detection accuracy of different plurality of defect judgment candidate models by considering whether each of the plurality of batteries is defective (S530).

[0094] According to an embodiment, the processor (120) can individually calculate the detection accuracy of each of the defect judgment candidate models using a scatter plot graph as shown in FIG. 6.

[0095] Referring to FIG. 6, the scatter plot graph can be represented by a y-axis representing the voltage change data value of each of the plurality of batteries to which the defect status has been pre-confirmed, and an x-axis representing the voltage change data value of each of the plurality of batteries to which one of the plurality of defect determination candidate models has been applied.

[0096] Accordingly, the processor (120) can display each location on the scatter plot graph plane by considering the voltage change data values ​​of the x-axis and y-axis for the plurality of batteries.

[0097] Subsequently, the processor (120) can divide the scatter plot graph plane into first to fourth regions based on the distribution of the plurality of batteries. Accordingly, the processor (120) can analyze the defect type of at least one battery among the plurality of batteries that has a defect.

[0098] More specifically, in the first area, a target battery detected as defective by the corresponding defect judgment candidate model among at least one battery identified as defective in step S510 may be located. In other words, it can be confirmed that at least one target battery located in the first area is defective. Additionally, in the second area, a target battery detected as good by the corresponding defect judgment candidate model among at least one battery identified as defective in step S510 may be located. In other words, it can be confirmed that at least one target battery located in the second area is a defective battery that was not detected by the corresponding defect judgment candidate model. Additionally, in the third area, a target battery detected as good by the corresponding defect judgment candidate model among at least one battery identified as good in step S510 may be located. In other words, it can be confirmed that at least one target battery located in the third area is good. Finally, in the fourth area, a target battery detected as defective by the corresponding defect judgment candidate model among at least one battery identified as good in step S510 may be located. In other words, at least one target battery located in the fourth region may be a good battery that has been over-detected by the corresponding defect judgment candidate model. Accordingly, the processor (120) may classify at least one target battery located in the first region as 'defective', classify at least one target battery located in the second region as 'undetected defect', and classify a target battery located in the fourth region as 'over-detected defect'.

[0099] Meanwhile, not limited to what is disclosed, the processor (120) may classify the defect types of the plurality of batteries into genuine defects or boundary defects based on the respective positions of the plurality of batteries in the scatter plot graph, and then individually analyze the defect types (defects / over-inspections / uninspections) of the plurality of batteries determined to be genuine defects and the plurality of batteries determined to be boundary defects. Here, a genuine defect may be a defect type corresponding to a battery located within a first region because the voltage change amount data value is large, and a boundary defect may be a defect type corresponding to a battery located within a predetermined range from the boundary between the first region and another adjacent region.

[0100] Meanwhile, the processor (120) can calculate the detection accuracy for each defect determination candidate model according to the defect type of at least one target battery located in the first to fourth regions.

[0101] According to an embodiment, the processor (120) can calculate a defect index for each defect judgment candidate model and calculate the detection accuracy of the defect judgment candidate model.

[0102] Here, the defect index is calculated using at least one target battery among the plurality of batteries in which defects, over-detection defects, and undetected defects occur, and may have an inverse relationship with the detection accuracy. In other words, among the plurality of defect judgment candidate models, the detection accuracy of at least one target model with the lowest defect index may be the highest. The defect index can be calculated as shown in [Equation 2] below.

[0104]

[0106] Here, the defect rate may be the ratio of the number of at least one target battery diagnosed as defective by the corresponding defect judgment candidate model among at least one battery pre-diagnosed as defective in step S510, relative to the number of the plurality of batteries. In other words, the defect rate may be calculated as the number of at least one target battery located in the first region relative to the number of the plurality of batteries.

[0107] Additionally, the over-inspection rate may be the ratio of the number of at least one target battery diagnosed as defective by the corresponding defect judgment candidate model among at least one battery pre-diagnosed as good in step S510, relative to the number of the plurality of batteries. In other words, the over-inspection rate may be calculated as the number of at least one target battery located in the fourth region relative to the number of the plurality of batteries.

[0108] Meanwhile, the defect index can be calculated by applying the number rather than the ratio for batteries of the undetected defect type that are diagnosed as good by the corresponding defect judgment candidate model among the target batteries pre-diagnosed as defective in step S510.

[0109] Generally, when a battery is shipped, the reliability of the diagnosis can be critically reduced in the case of non-detection, where a defective battery is diagnosed as normal, compared to over-detection, where a normal battery is detected as defective. Accordingly, the battery defect diagnosis device according to an embodiment of the present invention can detect a defect judgment model with high diagnostic reliability by reflecting the number rather than the ratio for batteries where non-detection occurs when calculating the defect index.

[0110] Afterwards, the processor (120) can extract at least one target model with high detection accuracy among a plurality of defect judgment candidate models and determine a defect judgment model based thereon (S540).

[0111] According to one embodiment, the processor (120) may determine the extracted target model as the defect determination model if there is only one extracted target model.

[0112] Meanwhile, according to another embodiment, the processor (120) can determine the defect judgment model by checking at least one condition individually applied to the plurality of target models when there are multiple extracted target models.

[0114] FIG. 7 is a flowchart for explaining a method for obtaining a defect judgment model when there are multiple target models in a battery defect diagnosis method according to an embodiment of the present invention, FIG. 8 is a table showing defect indices of multiple defect judgment candidate models in a battery defect diagnosis method according to an experimental example of the present invention, and FIG. 9 is a table counting detailed criteria by condition extracted from multiple target models in a battery defect diagnosis method according to an embodiment of the present invention.

[0115] Referring to FIG. 7, the processor (120) can extract at least one target model with the highest detection accuracy among the plurality of defect judgment candidate models (S710). In other words, the processor (120) can extract at least one target model with the lowest defect index among the plurality of defect judgment candidate models.

[0116] For example, the processor (120) can determine the defect index of a plurality of batteries diagnosed by the defect judgment candidate models based on 108 defect judgment candidate models obtained according to the combination of detailed criteria of each of the first to fourth conditions, as shown in FIG. 8.

[0117] Afterwards, the processor (120) can extract multiple defect judgment candidate models with the lowest defect index among 108 defect judgment candidate models as target models.

[0118] Meanwhile, the processor (120) can count the detailed criteria for each condition applied to the target models when there are multiple extracted target models (S720).

[0119] Afterward, the processor (120) can extract a specific detailed criterion with the highest counting number for each of the above conditions (S730). According to an embodiment, the processor (120) can check the first to fourth conditions individually applied to the target models among the 108 defect judgment candidate models, as shown in FIG. 9, and count the detailed criteria for each of the above conditions.

[0120] Subsequently, the processor (120) may determine a defect judgment candidate model to which the extracted condition-specific detailed criteria are applied among a plurality of defect judgment candidate models as a defect judgment model (S740). At this time, in the case of the 6th detailed calculation criterion in the 1st condition, since it is a result determined after the activation process is completed, it may be excluded from the counting ranking.

[0122] The battery defect diagnosis device and method according to the embodiments of the present invention have been described above.

[0123] The battery defect diagnosis device and method according to an embodiment of the present invention can improve inspection reliability by providing a defect determination model with high inspection accuracy that considers a defect index when inspecting for defects in a plurality of batteries that have undergone an activation process.

[0125] The operation of the method according to the embodiments and experimental examples of the present invention can be implemented as a computer-readable program or code on a computer-readable recording medium. A computer-readable recording medium includes all types of recording devices in which data that can be read by a computer system is stored. Additionally, the computer-readable recording medium may be distributed across networked computer systems, allowing computer-readable programs or code to be stored and executed in a distributed manner.

[0126] In addition, computer-readable recording media may include hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Program instructions may include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0127] Some aspects of the invention have been described in the context of a device, but may also be described according to a corresponding method, wherein a block or device corresponds to a method step or a feature of a method step. Similarly, aspects described in the context of a method may also be described according to a corresponding block or item or a feature of a corresponding device. Some or all of the method steps may be performed by (or using) a hardware device, such as, for example, a microprocessor, a programmable computer, or an electronic circuit. In some embodiments, one or more of the most important method steps may be performed by such a device.

[0128] Although the present invention has been described with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims. Explanation of the symbols

[0129] 100: Battery Fault Diagnostic Device 110: Memory 120: Processor 130: Transmitter / Receiver 140: Input interface device 150: Output interface device 160: Storage device 170: Bus

Claims

Claim 1 A battery defect diagnosis device comprising: at least one processor; and a memory for storing at least one instruction performed through the processor, wherein the at least one instruction comprises: an instruction to obtain a defect determination model based on the inspection accuracy of each of a plurality of defect determination candidate models to which at least one different condition is applied; and an instruction to determine whether at least one battery among a plurality of batteries in which an activation process has been performed is defective using the defect determination model. Claim 2 A battery defect diagnosis device according to claim 1, wherein the command for obtaining at least one defect determination model comprises: a command for individually inputting voltage data of a plurality of batteries that have undergone aging into the plurality of defect determination candidate models to verify the inspection accuracy of each of the defect determination candidate models; and a command for determining the defect determination model using at least one target model among the defect determination candidate models that has high inspection accuracy. Claim 3 In claim 1, the plurality of defect judgment candidate models each comprise: a time point for measuring voltage change amount data of the plurality of batteries in which an activation process has been performed; a method for calculating representative voltage change amount data of the plurality of batteries; and a standard deviation application criterion between voltage change amount data of the plurality of batteries. A battery defect diagnosis device in which detailed criteria are applied differently according to at least one condition among whether to noise filter the voltage change amount data of the plurality of batteries. Claim 4 A battery defect diagnosis device according to claim 2, wherein a command to verify the inspection accuracy of each of the defect judgment candidate models comprises: a command to individually calculate voltage change amount data for the plurality of batteries stored for a predetermined period at a predetermined reference temperature; a command to pre-diagnose whether a defect has occurred in each of the plurality of batteries based on the voltage change amount data; and a command to apply each of the defect judgment candidate models to the plurality of batteries, and to calculate a defect index according to each of the defect judgment candidate models based on the pre-diagnosis result, thereby verifying the inspection accuracy. Claim 5 A battery defect diagnosis device according to claim 4, wherein the defect index is calculated using at least one battery among the plurality of batteries that is defective, over-detected, and not detected, by considering the defect detection results of the plurality of batteries to which the prior diagnosis results and the defect judgment candidate models are individually applied. Claim 6 A battery defect diagnosis device according to claim 5, wherein the defect index is calculated by adding the number of at least one undetected battery to the value obtained by adding the defect rate and over-detection rate of at least one battery. Claim 7 A battery defect diagnosis device according to claim 2, wherein the command to determine the defect determination model using at least one target model with high inspection accuracy among the defect determination candidate models comprises: a command to obtain at least one target model with high inspection accuracy among the plurality of defect determination candidate models; and, if the target model is one, a command to determine the target model as the defect determination model. Claim 8 A battery defect diagnosis device according to claim 2, wherein a command to determine the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models comprises: a command to obtain the at least one target model among the plurality of defect judgment candidate models that has a high inspection accuracy; a command to check detailed criteria for configuration-specific conditions from each of the target models when there are multiple target models; and a command to extract the detailed criteria most frequently applied to the target models and to determine the defect judgment candidate model to which the extracted detailed criteria are applied as the defect judgment model. Claim 9 A battery defect diagnosis method comprising: a step of obtaining a defect determination model based on the inspection accuracy of each of a plurality of defect determination candidate models to which at least one different condition is applied; and a step of determining whether at least one battery among a plurality of batteries in which an activation process has been performed is defective using the defect determination model. Claim 10 A battery defect diagnosis method according to claim 9, wherein the step of obtaining at least one defect determination model comprises: a step of individually inputting voltage data of a plurality of batteries that have undergone aging into the plurality of defect determination candidate models to verify the inspection accuracy of each of the defect determination candidate models; and a step of determining the defect determination model using at least one target model among the defect determination candidate models that has high inspection accuracy. Claim 11 In claim 9, the plurality of defect judgment candidate models each comprise: a time point for measuring voltage change amount data of the plurality of batteries in which an activation process has been performed; a method for calculating representative voltage change amount data of the plurality of batteries; and a standard deviation application criterion between voltage change amount data of the plurality of batteries. A battery defect diagnosis method in which detailed criteria are applied differently according to at least one condition among whether to noise filter the voltage change amount data of the plurality of batteries. Claim 12 A battery defect diagnosis method according to claim 10, wherein the command to verify the inspection accuracy of each of the defect determination candidate models comprises: a step of individually calculating voltage change amount data for the plurality of batteries stored for a predetermined period at a predetermined reference temperature; a step of pre-diagnosing whether a defect has occurred in each of the plurality of batteries based on the voltage change amount data; and a step of applying each of the defect determination candidate models to the plurality of batteries, and based on the pre-diagnosis result, calculating a defect index according to each of the defect determination candidate models to verify the inspection accuracy. Claim 13 A battery defect diagnosis method according to claim 12, wherein the defect index is calculated using at least one battery among the plurality of batteries that is defective, over-detected, and not detected, by considering the defect detection results of the plurality of batteries to which the prior diagnosis results and the defect judgment candidate models are individually applied. Claim 14 A battery defect diagnosis method according to claim 13, wherein the defect index is calculated by adding the number of at least one undetected battery to the value obtained by adding the defect rate and over-detection rate of at least one battery. Claim 15 A battery defect diagnosis method according to claim 10, wherein the step of determining the defect determination model using at least one target model with high inspection accuracy among the defect determination candidate models comprises: a step of obtaining at least one target model with high inspection accuracy among the plurality of defect determination candidate models; and a step of determining the target model as the defect determination model when the target model is one. Claim 16 A battery defect diagnosis method according to claim 10, wherein the step of determining the defect judgment model using at least one target model with high inspection accuracy among the defect judgment candidate models comprises: obtaining the at least one target model among the plurality of defect judgment candidate models that has a high inspection accuracy; if there are a plurality of target models, confirming detailed criteria for configuration-specific conditions from each of the target models; and extracting the detailed criteria most frequently applied to the target models and determining the defect judgment candidate model to which the extracted detailed criteria are applied as the defect judgment model.

Images