System and method for predicting venting occurrence time of a battery cell

Abstract

Provided is a system for predicting a time of occurrence of gas emission of a battery cell including a platform portion having a sealed portion and disposed on at least one side of a pouch-type battery case and an electrode lead protruding from an end of the platform portion, the system including a storage unit for collecting data on a time of occurrence of gas emission according to a width of a remaining sealed portion; a measurement unit for periodically measuring the width of the remaining sealed portion of the battery cell to be measured; and a determination unit for predicting the time of occurrence of gas emission of the battery case to be measured by comparing the measured width of the remaining sealed portion with the collected data.

Description

Technical Field

[0001] This invention relates to a system and method for predicting the timing of exhaust gas generation in a single battery cell.

[0002] This application claims the benefit of priority based on Korean Patent Application No. 10-2021-0086372, filed on July 1, 2021, the entire contents of which are incorporated herein by reference. Background Technology

[0003] Recently, rechargeable and dischargeable secondary batteries have been widely used as energy sources for wireless mobile devices. Furthermore, secondary batteries have garnered attention as energy sources for electric vehicles, hybrid electric vehicles, and the like, and have been introduced as a solution to air pollution caused by existing gasoline and diesel vehicles using fossil fuels. Therefore, due to the advantages of secondary batteries, the types of applications using them are diversifying, and it is expected that secondary batteries will be used in even more fields and products in the future than they are now.

[0004] Secondary batteries can be classified according to the composition of their electrodes and electrolytes into lithium-ion batteries, lithium-ion polymer batteries, lithium polymer batteries, etc. Among these, the use of lithium-ion polymer batteries, which are less prone to electrolyte leakage and easier to manufacture, is increasing. Generally, secondary batteries are classified according to the shape of their casings into cylindrical or prismatic batteries where the electrode assembly is housed in a cylindrical or rectangular metal can, and pouch batteries where the electrode assembly is housed in a pouch-type casing made of aluminum laminated sheets. The electrode assembly included in the battery casing is a rechargeable and dischargeable power generation device, comprising a positive electrode, a negative electrode, and a separator between the positive and negative electrodes. Electrode assemblies are classified into jelly roll type: a structure formed by inserting a separator between long sheets of positive and negative electrodes coated with active material and winding the resulting structure; and stacked type, in which positive and negative electrodes, each having a certain size, are stacked sequentially with a separator inserted between them.

[0005] Figure 1 This is a schematic diagram illustrating the typical form of a pouch-type battery cell.

[0006] refer to Figure 1 The pouch-type battery cell has the following structure: an electrode assembly 20 is housed within a pouch-type battery casing 10, electrode leads 30 protrude from opposite ends of the battery casing 10, and a sealing portion 11a is formed around the outer periphery of the battery casing 10. In this case, the sealing portion 11a and a gas pocket portion 11b, which serves as the space between the sealing portion 11a and the housing space, are formed in a platform portion 11, which serves as the space between the space housing the electrode assembly and the end of the battery casing. The gas pocket portion 11b is a space where gases generated in the battery due to various reasons accumulate.

[0007] Based on customer requirements, various high-temperature storage tests are performed on the battery cells described above. These tests determine the safety and durability of the battery cells under harsh conditions. In this case, the time to gas venting of the battery cells can be predicted through high-temperature storage tests, thus estimating the durability and performance of the battery cells.

[0008] Specifically, when the amount of gas generated in a single battery cell increases due to various reasons such as high-temperature environmental experiments, the width of the sealing portion 11a gradually decreases and eventually ruptures, resulting in gas discharge, i.e., exhaust phenomenon.

[0009] In the prior art, the timing of venting of the sealing portion 11a is predicted by directly measuring the width of the sealing portion 11a with the naked eye and checking for the presence of venting. Therefore, this process takes a considerable amount of time, and the experiment must be repeated if venting occurs at a time point when no measurement was performed.

[0010] [Related Literature]

[0011] [Patent Literature]

[0012] Korean Patent Registration No. 10-2125238. Summary of the Invention

[0013] [Technical Issues]

[0014] To address the aforementioned problems, the present invention aims to provide a system for predicting the timing of exhaust emissions from individual battery cells. This system can automatically predict the timing of exhaust emissions and improve the accuracy of the prediction.

[0015] [Technical Solution]

[0016] According to the invention, a system for predicting the venting time of a battery cell includes a platform portion having a sealing portion formed on at least one side of a pouch-type battery casing and an electrode lead protruding from an end of the platform portion. The system includes a storage unit configured to collect data on the venting time based on the width of the remaining sealing portion, a measuring unit configured to periodically measure the width of the remaining sealing portion of the battery cell to be measured, and a determining unit configured to predict the venting time of the battery cell to be measured by comparing the measured width of the remaining sealing portion with the collected data.

[0017] The measurement unit may include: a camera configured to capture images or videos of a portion of the platform; and a calculator configured to calculate the width of the remaining sealed portion in the captured images or videos.

[0018] The measuring unit measures the width of the remaining seal by shortening the measurement time interval as the venting time approaches.

[0019] As a specific example, data collection and measurement of the width of the remaining sealed portion can be performed at high temperatures of 60°C or higher.

[0020] As a specific example, the determination of a cell can be achieved by using machine learning or deep learning to predict the timing of the exhaust of the battery cell to be measured.

[0021] Specifically, the determining unit can derive the correlation between the width of the remaining sealing portion and the corresponding venting time from this data.

[0022] Furthermore, the determining unit can predict the venting time of the battery cell based on this correlation, for each measurement time interval of the width of the remaining sealed portion, according to the measured width of the remaining sealed portion.

[0023] As another example, the system for predicting the exhaust time of a battery cell according to the present invention may further include a learning unit configured to learn the result of predicting the exhaust time.

[0024] As a specific example, the learning unit can be configured with training data for predicting the venting time of a battery cell, and the determination unit can derive a new correlation between the width of the remaining sealing portion and the corresponding venting time from the training data, and predict the venting time of the battery cell based on the measured width of the remaining sealing portion from the correlation.

[0025] Specifically, the learning unit can configure the training data by comparing the predicted exhaust occurrence time with the actual exhaust occurrence time to verify the validity of the data, and using the results of verifying the validity of the data to update the data collected in the storage unit.

[0026] In addition, the present invention provides a method for predicting the exhaust time of a battery cell using a system for predicting the exhaust time of a battery cell.

[0027] The method for predicting the venting time of a battery cell according to the present invention includes collecting data on the venting time based on the width of the remaining sealing portion, periodically measuring the width of the remaining sealing portion of the battery cell to be measured, and predicting the venting time of the battery cell to be measured by comparing the measured width of the remaining sealing portion with the collected data.

[0028] As a specific example, periodically measuring the width of the remaining sealing portion may include: capturing images or videos of the platform portion by a camera, and calculating the width of the remaining sealing portion in the captured images or videos.

[0029] As a specific example, predicting the venting time of a battery cell to be measured may include: deriving the correlation between the width of the remaining seal portion and the corresponding venting time, and based on this correlation, periodically predicting the venting time of the battery cell to be measured according to the measured width of the remaining seal portion.

[0030] As a specific example, the method according to the invention may further include: learning the result of predicting the timing of exhaust gas occurrence.

[0031] As a specific example, learning to predict exhaust gas occurrence time can include: validating the data by comparing the predicted exhaust gas occurrence time with the actual exhaust gas occurrence time, and configuring the training data by updating the data using the results of validating the data.

[0032] Predicting the venting time of a battery cell to be measured may include: deriving the correlation between the width of the remaining seal portion and the corresponding venting time from the training data, and predicting the venting time of the battery cell based on the measured width of the remaining seal portion from the correlation.

[0033] [Beneficial Effects]

[0034] According to the present invention, machine learning is used to predict the time of exhaust, thereby automatically predicting the time of exhaust of individual battery cells and improving the accuracy of the prediction. Attached Figure Description

[0035] Figure 1 This is a schematic diagram illustrating the typical form of a pouch-type battery cell.

[0036] Figure 2 This is a block diagram illustrating the configuration of a system for predicting the time of exhaust generation of a battery cell according to an embodiment of the present invention.

[0037] Figure 3 This is a schematic diagram illustrating the process of measuring the width of the remaining sealing portion.

[0038] Figure 4 It is a photograph that displays an image captured by a camera.

[0039] Figure 5 This is a block diagram illustrating the configuration of a system for predicting the time of exhaust generation of a battery cell according to another embodiment of the present invention.

[0040] Figure 6 This is a flowchart of the process of learning to predict results.

[0041] Figure 7 This is a schematic diagram illustrating a deep learning-based learning method. Detailed Implementation

[0042] The present invention will be described in detail below. First, the terms or expressions used in this specification and claims should not be construed as limited to their meanings as defined in common dictionaries, but should be understood in accordance with the meanings and concepts consistent with the technical spirit of the invention, based on the inventors' ability to appropriately define the terms or expressions to best explain the principles of the invention.

[0043] It should be understood that the terms "comprising" and / or "including," as used herein, specify the presence of the described features, values, steps, operations, elements, components, or combinations thereof, but do not preclude the presence or addition of one or more features, values, steps, operations, elements, components, or combinations thereof. It should be understood that when a component such as a layer, membrane, region, plate, etc., is referred to as being "on" another component, that component is "directly" on the other component or other components are inserted between these components. It should be understood that when a component such as a layer, membrane, region, plate, etc., is referred to as being "below" another component, that component is "directly" below the other component or other components are inserted between these components. Furthermore, it should be understood that when a component is "on" another component, that component is on or below the other components.

[0044] The present invention will be described in detail below.

[0045] (First Embodiment)

[0046] Figure 2 This is a block diagram illustrating the configuration of a system for predicting the time of exhaust generation of a battery cell according to an embodiment of the present invention. Figure 3 This is a schematic diagram illustrating the process of measuring the width of the remaining sealing portion.

[0047] refer to Figure 2 The present invention relates to a system 100 for predicting the venting time of a battery cell, the battery cell including a platform portion having a sealing portion formed on at least one side of a pouch-type battery casing, and electrode leads protruding from the end of the platform portion, and the system 100 including: a storage unit 110 for collecting data on the venting time based on the width of the remaining sealing portion; a measurement unit 120 for periodically measuring the width w of the remaining sealing portion of the battery cell to be measured; and a determination unit 130 for predicting the venting time of the battery cell to be measured by comparing the measured width of the remaining sealing portion with the collected data.

[0048] According to the present invention, the exhaust gas occurrence time is predicted based on machine learning, thereby automatically predicting the exhaust gas occurrence time and improving the accuracy of the prediction.

[0049] refer to Figures 1 to 3 The present invention aims to predict the venting time of a pouch-type battery cell. As described above, the pouch-type battery cell has the following structure: an electrode assembly 20 is housed in a pouch-type battery housing 10, electrode leads 30 protrude from opposite ends of the battery housing 10, and a sealing portion 11a is formed around the outer periphery of the battery housing 10. In this case, the sealing portion 11a and an air pocket portion 11b, which serves as the space between the sealing portion 11a and the housing space, are formed in a platform portion 11, which serves as the space between the space housing the electrode assembly and the end of the battery housing.

[0050] The electrode lead 30 includes a positive lead and a negative lead, and the positive lead and the negative lead can be configured as follows: Figure 1 and Figure 3 The electrode assembly and its components are shown protruding from the battery casing in opposite directions, but are not limited thereto. Detailed descriptions of the electrode assembly and its components are well known to those skilled in the art and are therefore omitted.

[0051] Meanwhile, the battery casing 10 is not particularly limited, as long as it can be used as the external material for encapsulating the battery, and cylindrical, prismatic, or pouch-shaped—especially pouch-shaped battery casings—can be used. Pouch-shaped battery casings are typically formed from aluminum laminated sheets and may include an internal sealant layer for sealing, a metal layer to prevent material penetration, and an external resin layer forming the outermost layer of the pouch-shaped battery casing. Details of the battery casing are well known to those skilled in the art, and therefore a description thereof is omitted here.

[0052] In this scenario, when the battery cell operates in a high-temperature environment—for example, a high-temperature storage experiment—a large amount of gas is generated within the battery cell and collected in the gas bag portion 11b. However, when the gas volume increases excessively, the seal of the sealing portion 11a is breached, resulting in an venting phenomenon where internal gas is released to the outside. In this case, the venting phenomenon does not occur instantaneously, but rather gradually decreases in width w of the remaining sealing portion as the internal pressure of the battery cell increases, and the seal is breached when the internal pressure is outside a threshold range, thus causing the venting phenomenon to occur. Here, the width w of the remaining sealing portion refers to the width w of the remaining sealing portion when the sealing portion (the heat-fused portion) between the sealing portions gradually tears as the internal pressure of the battery cell increases. According to the present invention, the venting time of the battery cell based on the measured width of the remaining sealing portion can be predicted from multiple data points without having to wait for the venting phenomenon to actually occur.

[0053] Multiple data points are related to the venting time based on the width of the remaining seal portion, and are stored in storage unit 110. In storage unit 110, during the prediction and measurement of venting times for multiple battery cells, captured images, the width of the remaining seal portion, and the venting time are accumulated to generate large datasets. Here, each venting time can also be defined as the time elapsed after measuring the remaining seal portion until venting occurs. In this case, storage unit 110 may include a database (DB) for storing and managing the aforementioned data, and this data can be used as the basis for configuring training data in the learning unit described below. In this case, the data can be categorized and stored according to the specifications of the battery cells and experimental conditions.

[0054] The measuring unit 120 measures the width of the remaining sealed portion of the battery cell 1 to be measured. Specifically, the measuring unit 120 may include a camera 121 for capturing images or videos of the platform portion 11, and a calculator 122 for calculating the width of the remaining sealed portion in the captured images or videos.

[0055] refer to Figure 1 and Figure 3 Images or videos of the platform portion 11 can be captured while the battery cell 1 is mounted in the clamp 123. At least one battery cell 1 can be mounted in the clamp 123 for fixation, thereby facilitating the capture of images or videos of the platform portion 11. Although Figure 3 The illustration shows a battery cell 1 mounted in a fixture 123, but as... Figure 4 As shown, when two or more battery cells 1 are mounted in the clamp 123, images or videos of two or more battery cells 1 can be captured simultaneously. Furthermore, the battery cells 1 can be mounted in the clamp 123 in a direction perpendicular to the ground, thereby facilitating image capture of the platform portion 11 and easy gas collection in the gas bag portion 11b. Additionally, the shape of the clamp 123 is not particularly limited, but it is preferable to minimize the contact area between the battery cells and the clamp 123 to prevent interference with the venting of the battery cells 1. Figure 4 As shown, the electrode lead 30 is preferably covered to prevent short circuits.

[0056] There are no particular limitations on camera 121, as long as camera 121 can capture images or videos, and for example, a charge-coupled device (CCD) camera can be used.

[0057] Additionally, the measurement section 120 may further include a display device (not shown) that displays the captured image in the form of image data.

[0058] Figure 4It is a photograph that displays an image captured by a camera.

[0059] like Figure 4 As shown, the image captured by camera 121 is displayed visually by a display device. The width w of the remaining sealed portion in the displayed image is measured by calculator 122. Calculator 122 can be a general-purpose computing device.

[0060] In this case, such as Figure 4 As shown, the display device can display a grid-like scale on the image, and the calculator 122 calculates the width of the portion of the image corresponding to the remaining sealed portion by comparing it with the scale. Figure 4 In this case, it can be seen that the width of the remaining sealing portion is calculated to be 8.7mm.

[0061] The above process can be performed periodically until venting occurs in the battery cell, and data regarding the venting occurrence time based on the width of the remaining sealed portion can be stored in the storage unit 110. In this case, due to the gas generated in the battery cell, the width of the remaining sealed portion gradually decreases, and the measuring section 120 measures the width of the remaining sealed portion by shortening the measurement time interval as the venting occurrence time approaches—that is, as the width of the remaining sealed portion decreases. This is because as the width of the remaining sealed portion decreases, the adhesive force between the bags decreases, thereby accelerating separation. Therefore, the measurement time interval can be set shorter to accurately identify the venting occurrence time.

[0062] Meanwhile, when battery cells are placed in high-temperature environments—such as high-temperature storage experiments—venting of the battery cells is likely to occur, and data collection and measurement of the width of the remaining sealed portion can be performed at high temperatures, especially at 60°C or higher.

[0063] The measurement results performed by the measurement section 120 can be sent and stored in the storage unit 110 and configured as part of the data.

[0064] When measuring the width of the remaining sealed portion, the timing of venting of the battery cell is predicted. The determination unit 130 can be operated in a computing device and can automatically predict the timing of venting of the battery cell to be measured through machine learning or deep learning, thereby improving the accuracy of the prediction.

[0065] Specifically, determining unit 130 can derive the correlation between the width of the remaining seal and the corresponding exhaust time from data—that is, data regarding the exhaust time based on the width of the remaining seal. The correlation between the width of the remaining seal and the corresponding exhaust time can be understood as indicating the trend indicated by the exhaust time (time taken to exhaust) relative to the measured width of the remaining seal. For example, the correlation between the width of the remaining seal and the corresponding exhaust time can be expressed as a single equation, and this can be accomplished through regression analysis. Therefore, by setting the width of the remaining seal as the independent variable and the exhaust time as the dependent variable, a relational expression that appropriately reflects the data can be derived. The relational expression can be in various forms such as a linear function, a quadratic function, other polynomial functions, exponential functions, logarithmic functions, etc. For example, when there is a linear functional relationship between the exhaust time and the width of the remaining seal, the relational expression can be expressed as the following equation:

[0066] y = ax + b (1)

[0067] (In equation (1) above, x represents the width of the remaining sealing portion (mm), y represents the time (h) from the measurement of the width of the remaining sealing portion until venting occurs, and a and b are constants.)

[0068] As described above, by expressing the correlation between the width of the remaining seal portion and the time of venting as an equation, the time of venting can be automatically predicted by measuring only the width of the remaining seal portion.

[0069] When the correlation between the width of the remaining sealed portion and the corresponding venting time is derived, the determining unit 130 can predict the venting time of the battery cell based on the measured width of the remaining sealed portion. When the correlation is derived in the form of the above equation, the venting time can be predicted by substituting the measured width of the remaining sealed portion into the above equation (1). Because the measuring unit 120 predicts the width of the remaining sealed portion at specific time intervals, the determining unit 130 can also predict the venting time of the battery cell for each measurement interval of the width of the remaining sealed portion.

[0070] The above process is performed automatically and repeatedly until the venting time of the battery cell occurs. That is, the system according to the invention for predicting the venting time of a battery cell can measure the width of the remaining sealing portion and predict the venting time by comparing the measured width with previously stored data, and repeatedly measure the width of the remaining sealing portion at specific time intervals when venting does not occur. This method improves the accuracy of prediction compared to experiments performed by visual inspection, and accurately identifies the moment of venting occurrence.

[0071] (Second Embodiment)

[0072] Figure 5 This is a block diagram illustrating the configuration of a system for predicting the time of exhaust generation of a battery cell according to another embodiment of the present invention. Figure 6 This is a flowchart of the process of learning to predict results. Figure 7 This is a schematic diagram illustrating a deep learning-based learning method.

[0073] refer to Figure 5 The system 200 for predicting the timing of exhaust emissions from individual battery cells may further include a learning unit 140 for learning the prediction results. In this invention, for ease of description, the function of the learning unit 140 can be analogous to that of a calculator and a determination unit, configured as a computing device similar to a calculator and a determination unit, and operated in the same device as a calculator and a determination unit. According to the invention, the accuracy of the prediction can be further improved by using machine learning or deep learning to reflect whether the prediction result is accurate or inaccurate.

[0074] refer to Figure 5 and Figure 6 The learning unit 140 can be configured with training data for predicting the exhaust gas occurrence time. The training data can be configured by updating previously stored data in the storage unit using the results of new measurements. First, the learning unit 140 verifies the validity of the data by comparing the predicted exhaust gas occurrence time with the actual exhaust gas occurrence time. When the predicted exhaust gas occurrence time matches the actual exhaust gas occurrence time, the data is deemed valid and recorded in the storage unit 110. When the predicted exhaust gas occurrence time differs from the actual exhaust gas occurrence time, the data stored in the storage unit 110 is modified and updated. In this case, the reasons for the inconsistency can be analyzed by considering the experimental conditions (such as the temperature of individual battery cells) input along with the stored data. The learning unit 140 configures the training data by updating the storage unit 110 using the verification results. In this way, the learning unit 140 can derive more accurate data through machine learning.

[0075] Furthermore, when the training and learning data are configured to be executed by the learning unit 140 through deep learning, the learning unit 140 can be configured as a deep neural network (DNN).

[0076] DNN is a type of deep learning (machine learning) model used to classify input data based on learning data, and can be understood as a system or network that forms one or more layers in one or more computers and performs judgments based on multiple data.

[0077] refer to Figure 7A DNN may include an input layer 141, one or more hidden layers 142, and an output layer 143.

[0078] Training data is fed into input layer 141, and weights are updated in reverse by comparing the actual values ​​with the resulting values ​​calculated through one or more hidden layers 142 and output layer 143. After learning is complete, the resulting values ​​can be obtained by inputting the information needed for prediction.

[0079] One or more hidden layers 142 may include convolutional layers, pooling layers, and fully connected layers. Here, the convolutional layers may extract feature maps and perform convolution operations relative to the image input to the input layer. Pooling layers may be connected to the convolutional layers to perform subsampling on the output of the convolutional layers. Fully connected layers may be connected to the pooling layers and learn the subsampled output of the pooling layers according to the category to be output to the output layer 143.

[0080] The structure of the connected layers of a DNN can be formed by appropriately selecting well-known algorithms, and can be, for example, a convolutional neural network (CNN) structure or a recurrent neural network (RNN) structure.

[0081] DNN can be implemented in a single computer or through a network connecting multiple computers.

[0082] The learning unit 140 inputs updated training data into the input layer 141 of the DNN. When the input training data passes through the hidden layer 142, the final output is output from the output layer 143. The learning unit 140 can learn from the newly updated training data obtained by updating the weights based on the results of validating the prediction results.

[0083] Once the data learning is complete, the determination unit 130 derives a new correlation between the width of the remaining sealed portion and the corresponding venting time from the learned data, and based on this correlation, predicts the venting time of the battery cell according to the measured width of the remaining sealed portion. Thereafter, the process of verifying the validity of the prediction results and reflecting the verification results can be repeated to further improve the accuracy of the prediction results.

[0084] In addition, based on the exhaust time of a battery cell as described above, the present invention provides a method for predicting the exhaust time of a battery cell.

[0085] Specifically, a method for predicting the venting time of a battery cell may include collecting data on the venting time based on the width of the remaining seal portion, periodically measuring the width of the remaining seal portion of the battery cell to be measured, and predicting the venting time of the battery cell to be measured by comparing the measured width of the remaining seal portion with the collected data.

[0086] During data collection, in the process of predicting and measuring the venting time of multiple battery cells, the width of the remaining sealed portion, the captured images, and the venting time can be accumulated in the storage cell.

[0087] Periodically measuring the width of the remaining seal portion can include capturing images or videos of the platform portion by a camera and calculating the width of the remaining seal portion in the captured images or videos. Image or video capture can be performed while the battery cell is mounted in a fixture. The width of the remaining seal portion can be measured while the captured images are displayed on a display device.

[0088] The width of the remaining sealed portion can be measured periodically until the venting time of the battery cell occurs, and data regarding the venting time based on the width of the remaining sealed portion can be stored in a storage unit. In this case, the width of the remaining sealed portion can be measured by shortening the measurement time interval as the venting time approaches. This process can be performed at a high temperature of 60°C or higher.

[0089] Simultaneously, machine learning or deep learning can be used to predict the venting time of the battery cell to be measured, and this can include deriving the correlation between the width of the remaining seal portion and the corresponding venting time, and periodically predicting the venting time of the battery cell to be measured based on the width of the remaining seal portion according to this correlation. The derivation of the correlation can be performed through regression analysis as described in detail above.

[0090] The method for predicting the exhaust time of a single battery cell according to the present invention may further include learning the prediction results.

[0091] Learning from the predictions can include validating the data by comparing the predicted exhaust timing with the actual exhaust timing, and configuring the training data by updating the data using the validation results.

[0092] Subsequently, the correlation between the width of the remaining sealed portion and the corresponding venting time can be derived from the training data, and the venting time of the battery cell based on the measured width of the remaining sealed portion can be predicted from this correlation.

[0093] The above description is merely an example of the technical concept of the present invention, and those skilled in the art can make various modifications and alterations without departing from the basic characteristics of the invention. Therefore, the accompanying drawings set forth herein are not intended to limit the technical concept of the invention, but rather to describe it, and the scope of the technical concept of the invention is not limited by the drawings. The scope of protection of the present invention should be interpreted based on the appended claims, and all technical concepts within the same scope as the present invention should be interpreted as being included within the scope of the present invention.

[0094] In this specification, terms such as up, down, left, right, forward, and backward are used for descriptive convenience only, and it will therefore be apparent that these terms may change depending on the position of the object or the observer.

[0095] [Figure Labels]

[0096] 1: Battery cell

[0097] 10: Battery casing

[0098] 11: Platform Section

[0099] 11a: Sealing part

[0100] 11b: Airbag section

[0101] 20: Electrode assembly

[0102] 30: Electrode leads

[0103] 100: System for predicting exhaust timing

[0104] 110: Storage unit

[0105] 120: Measurement Unit

[0106] 121: Camera

[0107] 122: Calculator

[0108] 123: Fixture

[0109] 130: Determine the unit

[0110] 140: Learning Unit

[0111] 141: Input Layer

[0112] 142: Hidden Layer

[0113] 143: Output layer.

Claims

1. A system for predicting the timing of venting in a single battery cell, the battery cell comprising a platform portion having a sealed portion and disposed on at least one side of a pouch-type battery housing and electrode leads protruding from an end of the platform portion, the system comprising: A storage unit configured to collect data on the timing of venting based on the width of the remaining sealing portion; A measuring unit configured to periodically measure the width of the remaining sealed portion of the battery cell to be measured; as well as A determining unit is configured to predict the venting time of the battery cell to be measured by comparing the width of the measured remaining sealed portion with the collected data.

2. The system according to claim 1, wherein, The measurement unit includes: A camera, configured to capture images or videos of the platform portion; and A calculator configured to calculate the width of the remaining sealed portion in a captured image or video.

3. The system according to claim 1, wherein, The measuring unit measures the width of the remaining sealed portion by shortening the measurement time interval as the exhaust time approaches.

4. The system according to claim 1, wherein, The data collection and the measurement of the width of the remaining sealing portion are performed at a high temperature of 60°C or higher.

5. The system according to claim 1, wherein, The determining unit uses machine learning or deep learning to predict the time of exhaust occurrence of the battery cell to be measured.

6. The system according to claim 5, wherein, The determining unit derives the correlation between the width of the remaining sealing portion and the corresponding venting time from the data.

7. The system according to claim 6, wherein, Based on the correlation, the determining unit predicts the venting time of the battery cell for each measurement time interval of the width of the remaining sealing portion, according to the measured width of the remaining sealing portion.

8. The system of claim 1 further includes a learning unit configured to learn the result of predicting the exhaust gas occurrence time.

9. The system according to claim 8, wherein, The learning unit is configured with training data to predict the timing of exhaust gas occurrence, and The determining unit derives a new correlation between the width of the remaining sealing portion and the corresponding venting time from the training data, and predicts the venting time of the battery cell based on the measured width of the remaining sealing portion from the correlation.

10. The system according to claim 8, wherein, The learning unit configures the training data by comparing the predicted exhaust occurrence time with the actual exhaust occurrence time to verify the validity of the data, and uses the results of verifying the validity of the data to update the data collected in the storage unit.

11. A method for predicting the time of exhaust generation of a single battery cell using the system according to claim 1, the method comprising: Collect data on the timing of venting based on the width of the remaining sealed portion; Periodically measure the width of the remaining sealed portion of the battery cell to be measured; as well as The timing of venting of the battery cell to be measured is predicted by comparing the measured width of the remaining sealed portion with the collected data.

12. The method according to claim 11, wherein, Periodically measuring the width of the remaining sealing portion includes: capturing images or videos of the platform portion by a camera, and calculating the width of the remaining sealing portion in the captured images or videos.

13. The method according to claim 11, wherein, Predicting the venting time of the battery cell to be measured includes: deriving the correlation between the width of the remaining seal portion and the corresponding venting time, and based on the correlation, periodically predicting the venting time of the battery cell to be measured according to the measured width of the remaining seal portion.

14. The method of claim 11, further comprising: The results of learning to predict the timing of exhaust gas occurrence.

15. The method according to claim 14, wherein, The results of learning to predict exhaust gas occurrence time include: validating the validity of the data by comparing the predicted exhaust gas occurrence time with the actual exhaust gas occurrence time, and configuring the training data by updating the data using the results of validating the data.

16. The method according to claim 15, wherein, Predicting the venting time of the battery cell to be measured includes: deriving the correlation between the width of the remaining seal portion and the corresponding venting time from the training data, and predicting the venting time of the battery cell to be measured based on the measured width of the remaining seal portion from the correlation.

Images