Weld management apparatus, weld management method, and weld management system

Abstract

According to some embodiments, a welding management device includes: an interface configured to acquire electrical data collected during the welding process of a battery; and a controller configured to calculate a first local reachability density (lrd) of the electrical data based on training data and the electrical data stored in a local outlier factor (LOF) model, calculate an LOF score of the electrical data based on a second lrd and a first lrd of neighboring data within a reference distance of the electrical data in the training data, and diagnose the LOF state of the battery based on the LOF score.

Description

[0001] Cross-references to related applications

[0002] This application claims priority to Korean Patent Application No. 10-2024-0077282, filed on June 14, 2024, the disclosure of which is incorporated herein by reference. Technical Field

[0003] The embodiments disclosed herein relate to a welding management device, a welding management method, and a welding management system. Background Technology

[0004] In recent years, research and development of rechargeable batteries have been actively pursued. Here, a rechargeable battery is a battery capable of being recharged and discharged, and can be interpreted as including conventional Ni / Cd batteries, Ni / MH batteries, and the more recent lithium-ion batteries. Among rechargeable batteries, lithium-ion batteries can achieve higher energy densities than traditional Ni / Cd and Ni / MH batteries, and can be manufactured in smaller and lighter sizes, thus offering high availability as a power source for mobile devices. In recent years, lithium-ion batteries have expanded their applications to power electric vehicles, making them a promising next-generation energy storage medium.

[0005] Welding processes can be performed to manufacture battery cells. For example, resistance welding can be performed to form the negative electrode terminal of a cylindrical battery cell. During the welding process, electrical data from the welding equipment, such as voltage, current, temperature, and resistance, can be collected, and the welding condition of the battery cell can be determined based on this data. The accuracy of determining the welding condition can vary depending on the methods used to collect and process the electrical data. Summary of the Invention

[0006] Technical issues

[0007] The embodiments disclosed herein are intended to provide welding management equipment, welding management methods, and welding management systems that can improve the methods for collecting and processing electrical data, thereby enhancing the performance of determining the welding status.

[0008] The technical objectives of the embodiments disclosed herein are not limited to those mentioned above, and other technical objectives not mentioned will be clearly understood by those skilled in the art from the following description.

[0009] Technical solution

[0010] According to some embodiments, a welding management device includes: an interface configured to acquire electrical data collected during the welding process of a battery; and a controller configured to calculate a first local reachability density (lrd) of the electrical data based on training data and the electrical data stored in a local outlier factor (LOF) model, calculate an LOF score of the electrical data based on a second lrd and a first lrd of neighboring data within a reference distance of the electrical data in the training data, and diagnose the welding status of the battery based on the LOF score.

[0011] According to some embodiments, the electrical data may include dynamic resistance values ​​collected at a first measurement frequency during periods when the welding voltage exceeds a reference voltage during the welding process, and the LOF model may be configured to calculate an LOF score based on the dynamic resistance values.

[0012] According to some embodiments, the controller can be configured to calculate the LOF score by inputting the value of the dynamic resistance value after a first time has elapsed from the start time of the period exceeding the reference voltage into the LOF model.

[0013] According to some embodiments, the controller can be configured to: determine the number of data points for the dynamic resistance value based on the time period during which the welding voltage exceeds the reference voltage and a first measurement frequency, and calculate the number of neighboring data points within the reference distance based on the Euclidean distance in terms of the number of data points.

[0014] According to some embodiments, the controller can be configured to diagnose the welding condition of the battery as defective when the LOF score exceeds a threshold score, and the threshold score can be determined based on tensile strength test data of the test battery manufactured through the welding process.

[0015] According to some embodiments, the controller can be configured to adjust the threshold score when consumables of the welding equipment performing the welding process are replaced.

[0016] According to some embodiments, the electrical data may include a first dataset less than the cutoff frequency and one or more second datasets greater than or equal to the cutoff frequency, and the controller may be configured to detect the time point in the welding process where the defect occurs based on the inflection point of one or more second datasets when the welding condition is diagnosed as defective based on the first dataset.

[0017] According to some embodiments, a welding management method includes: acquiring electrical data collected during the welding process of a battery; calculating a first local reachability density (lrd) of the electrical data based on training data and the electrical data stored in a local outlier factor (LOF) model; calculating an LOF score of the electrical data based on a second lrd and a first lrd of neighboring data within a reference distance of the electrical data in the training data; and diagnosing the welding status of the battery based on the LOF score.

[0018] According to some embodiments, the electrical data may include dynamic resistance values ​​collected at a first measurement frequency during periods when the welding voltage exceeds a reference voltage during the welding process, and the LOF model may be configured to calculate an LOF score based on the dynamic resistance values.

[0019] According to some embodiments, calculating the LOF score may include inputting the value of the dynamic resistance value after a first time has elapsed from the start time of the period exceeding the reference voltage into the LOF model.

[0020] According to some embodiments, calculating the LOF score may include determining the number of data points for the dynamic resistance value based on the period during which the welding voltage exceeds the reference voltage and a first measurement frequency, and calculating the number of neighboring data points within the reference distance based on the Euclidean distance in terms of the number of data points.

[0021] According to some embodiments, diagnosing the weld condition may include diagnosing the weld condition of the battery as defective when the LOF score exceeds a threshold score, and determining the threshold score based on tensile strength test data of the test battery manufactured through the welding process.

[0022] According to some embodiments, the welding management method may also include adjusting a threshold score when consumables of the welding equipment performing the welding process are replaced.

[0023] According to some embodiments, the electrical data may include a first dataset less than the cutoff frequency and one or more second datasets greater than or equal to the cutoff frequency, and the welding management method may further include detecting the time point in the welding process where the defect occurs based on the inflection point of one or more second datasets when the welding condition is diagnosed as defective based on the first dataset.

[0024] According to some embodiments, a welding management system includes: a battery; welding equipment configured to perform a welding process on the battery; and welding management equipment configured to acquire electrical data collected during the welding process; calculate a first lrd of the electrical data based on training data and the electrical data stored in an LOF model; calculate an LOF score of the electrical data based on a second lrd and the first lrd of neighboring data within a reference distance of the electrical data in the training data; and diagnose the welding status of the battery based on the LOF score.

[0025] Beneficial effects

[0026] According to the embodiments disclosed herein, welding management equipment, welding management methods, and welding management systems can be provided to improve the methods for collecting and processing electrical data, thereby enhancing the performance of determining the welding condition. Specifically, defects such as under-welding and over-welding that occur during resistance welding of the negative electrode connector can be detected using electrical data models such as the Local Outlier Factor (LOF) model.

[0027] The technical effects of the embodiments disclosed herein are not limited to those mentioned above, and those skilled in the art will clearly understand from the disclosure herein other effects not mentioned. Attached Figure Description

[0028] Figure 1 The components constituting a welding management system according to some embodiments are shown.

[0029] Figure 2 The components constituting a welding management device are shown according to some embodiments.

[0030] Figures 3 to 5 A method for collecting electrical data according to some embodiments is shown.

[0031] Figure 6 A method for calculating the Local Outlier Factor (LOF) score according to some embodiments is shown.

[0032] Figure 7 A method for diagnosing weld condition based on LOF score, according to some embodiments, is shown.

[0033] Figure 8 A method for classifying weld strength based on LOF scores, according to some embodiments, is shown.

[0034] Figure 9 The process of setting a threshold score based on the relationship between tensile strength and LOF score according to some embodiments is shown.

[0035] Figure 10 A method for adjusting a threshold fraction when replacing consumables in welding equipment, according to some embodiments, is shown.

[0036] Figure 11 The steps for constructing a welding management method according to some embodiments are shown. Detailed Implementation

[0037] In the following description, embodiments disclosed herein will be illustrated with reference to the accompanying drawings. However, this is not intended to limit the disclosure herein to the specific embodiments, and it should be construed as including various modifications, equivalents, and / or substitutions to the embodiments disclosed herein.

[0038] It should be understood that the embodiments and terminology used herein are not intended to limit the technical features set forth herein to the specific embodiments, and include various changes, equivalents, or substitutions to the corresponding embodiments. Regarding the description of the drawings, similar or related reference numerals may be used to refer to similar or related elements. It should be understood that the singular form of the noun corresponding to an item may include one or more things, unless the relevant context clearly indicates otherwise.

[0039] As used herein, each of the phrases such as “A or B,” “at least one of A and B,” “at least one of A or B,” “A, B or C,” “at least one of A, B and C,” and “at least one of A, B or C” may include any one or all possible combinations of items enumerated together in the corresponding phrase. Unless otherwise specifically stated, terms such as “first” and “second,” “first,” “second,” “A,” “B,” “(a),” or “(b)” may be used simply to distinguish corresponding components from one other component and do not otherwise limit the components (e.g., in terms of importance or order).

[0040] In this specification, it should be understood that if an element (e.g., a first element) is referred to as "connected to," "coupled to," or "in contact with" another element (e.g., a second element) with or without the terms "operably" or "communically," it means that the element can be connected to the other element directly (e.g., via wired or wireless) or indirectly (e.g., via a third element).

[0041] Methods according to various embodiments disclosed herein may be included and provided in a computer program product. The computer program product may be traded as a product between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., an optical disc read-only memory and a CD-ROM), or distributed online via an app store (e.g., downloaded or uploaded), or directly between two operator devices. If distributed online, at least a portion of the computer program product may be temporarily generated or at least temporarily stored in the memory of a machine-readable storage medium, such as a manufacturer's server, an app store's server, or a relay server.

[0042] According to the embodiments disclosed herein, each of the above components (e.g., a module or program) may include a single entity or multiple entities, and some of the multiple entities may be set separately from the other components. According to the embodiments disclosed herein, one or more of the above components or operations may be omitted, or one or more other components or operations may be added. Alternatively or additionally, multiple components (e.g., modules or programs) may be integrated into a single component. In this case, the integrated component can still perform one or more functions of each of the multiple components in the same or similar manner as they were performed by a corresponding component of the multiple components prior to integration. According to the embodiments disclosed herein, operations performed by a module, program, or other component may be performed sequentially, in parallel, repeatedly, or heuristically, or one or more operations may be performed in a different order or omitted, or one or more other operations may be added.

[0043] Figure 1 The components constituting a welding management system according to some embodiments are shown.

[0044] refer to Figure 1 The welding management system 100 may include welding equipment 120 for performing the welding process for the battery 110 and welding management equipment 130 for managing the welding process. However, the welding management system is not limited to this, and some components may be omitted from the welding management system 100, or other common components may be further included in the welding management system 100.

[0045] Battery 110 can be formed through a welding process. For example, battery 110 may include battery cells, battery modules, battery packs, etc. Battery cells may include cylindrical battery cells. Cylindrical battery cells may include rivets, current collectors, terminals, flanges, etc. formed on the positive or negative electrode, and the detailed structure of such cylindrical battery cells can be formed through welding.

[0046] The welding equipment 120 can perform a welding process to form a battery 110. The welding process may include a resistance welding process. During the resistance welding process, voltage and current can be input to the welding equipment 120, the welding equipment 120 can output voltage and current to the welding target, and the temperature, resistance, etc., of the welding target can be measured. The quality of the welding process of the battery 110 can be determined based on whether the electrical data described above are within appropriate ranges during the welding process. Therefore, a model for estimating the welding state based on electrical data can be utilized.

[0047] Welding management equipment 130 can manage the welding process of battery 110 performed by welding equipment 120. Welding management equipment 130 can use a condition diagnostic system to diagnose the welding condition of battery cells 110. For example, welding management equipment 130 can determine whether the condition diagnostic model used to diagnose the welding condition is operating correctly or whether the condition diagnostic model needs correction. According to embodiments, the condition diagnostic system may include unsupervised learning-based models, such as Local Outlier (LOF) models, One-Class Support Vector Machine (OC-SVM) models, Deep Support Vector Data Technology (Deep SVDD), etc., or supervised learning-based models, such as XGBoost. According to embodiments disclosed herein, defects occurring during resistance welding of the negative terminal piece, such as under-welding and over-welding, can be detected using an electrical data model such as an LOF model.

[0048] Figure 2 The components constituting a welding management device are shown according to some embodiments.

[0049] refer to Figure 2 The welding management device 130 may include an interface 131 and a controller 132. However, the welding management device is not limited to this; some components may be omitted from the welding management device 130, or other general-purpose components may be included in the welding management device 130.

[0050] Interface 131 can be configured to acquire electrical data collected during the welding process of battery 110. For example, interface 131 may include sensors configured to measure electrical data, and the sensors may include voltage sensors, current sensors, temperature sensors, resistance sensors, etc. Alternatively, the sensors configured to measure electrical data may be provided externally to the welding management device 130, and interface 131 may include a communication unit configured to receive electrical data from external sensors in a wired and / or wireless manner.

[0051] The controller 132 may include memory and a processor. The processor of the controller 132 may be implemented as an array of multiple logic gates for processing various operations or as a general-purpose microprocessor, and may consist of a single processor or multiple processors. For example, the processor may be implemented as at least one of a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), and an application processor (AP).

[0052] The memory of controller 132 can store various data, commands, mobile applications, computer programs, etc. The processor can process various operations by executing commands stored in the memory. The memory can be implemented as a non-volatile device such as read-only memory (ROM), programmable ROM (PROM), erasable programmable ROM (EPROM), electrically erasable PROM (EEPROM), flash memory, parallel random access memory (PRAM), magnetoresistive RAM (MRAM), resistive RAM (RRAM), ferroelectric RAM (FRAM), or a volatile device such as dynamic RAM (DRAM), static RAM (SRAM), synchronous dynamic RAM (SDRAM), parallel RAM (PRAM), or it can be implemented as a hard disk drive (HDD), solid-state drive (SSD), secure digital storage (SD), micro SD, or a combination thereof.

[0053] The controller 132 can be configured to calculate a first local reachability density (lrd) of electrical data based on training data and electrical data stored in the LOF model. The LOF model may include a model trained to diagnose welding states corresponding to electrical data using the LOF algorithm. The LOF model can be implemented as an artificial intelligence model trained using various machine learning techniques. The LOF model stores training data for model training, and the training data may include a set of electrical data measured during the welding process. The LOF model can calculate the first lrd of the electrical data collected from the battery 110. The lrd value of the LOF algorithm can represent the density of the probability that new data will reach existing data. The specific calculation process of the first lrd can be found in [reference needed]. Figure 6 This will be described below.

[0054] The controller 132 can be configured to calculate the LOF score of the electrical data based on a second lrd and a first lrd of neighboring data within a reference distance in the training data. The training data and the electrical data can be n-dimensional data. For example, the training data and the electrical data can be n-dimensional dynamic resistance data. Alternatively, other types of electrical data—such as welding voltage, welding current, transformer voltage, and welding temperature—can be used instead of dynamic resistance. The dimension n of the training data and the electrical data can correspond to the number of collected values. For example, 140-dimensional dynamic resistance data can contain 140 resistance values ​​measured at regular intervals during the welding process. The distance to the electrical data can be a distance in n-dimensional space. In an embodiment, the reference distance can refer to the k training data points closest to the electrical data. When k is 5, five second lrd values ​​can be calculated for the five training data points within the reference distance. The LOF model can calculate the LOF score of the electrical data based on the first lrd and the k second lrds.

[0055] The controller 132 can be configured to diagnose the welding condition of the battery 110 based on the LOF score. The LOF score can be used as an indicator of the welding condition of the battery 110. The LOF score can replace conventional indicators, such as tensile strength. Tensile strength may require separate measuring equipment and procedures, and the battery sampled for tensile strength measurement may be damaged. In contrast, LOF score-based diagnostics can perform a non-destructive, comprehensive inspection.

[0056] According to an embodiment, the electrical data may include dynamic resistance values ​​collected at a first measurement frequency during periods when the welding voltage exceeds a reference voltage during the welding process, and the LOF model may be configured to calculate the LOF score based on the dynamic resistance values. For example, the electrical data may be collected at the first measurement frequency during periods when the welding voltage exceeds 0.3 V, and in addition to dynamic resistance, the electrical data may also be voltage, current, temperature, etc. For example, the first measurement frequency may be 50 kHz. The specific values ​​of the reference voltage and the first measurement frequency may be changed as needed.

[0057] According to an embodiment, controller 132 can be configured to calculate the LOF score by inputting the dynamic resistance value, after a first time has elapsed since the start of the period exceeding the reference voltage, into the LOF model. For example, the first time could be the time elapsed for the first 30 values, and could be determined based on a first measurement frequency. Numbers other than 30 can be used as needed. By excluding the first 30 values ​​in this way, transient data can be excluded, thereby improving the diagnostic performance of the LOF model.

[0058] According to an embodiment, the controller 132 can be configured to determine the number of data points for the dynamic resistance value based on the period during which the welding voltage exceeds a reference voltage and a first measurement frequency, and to calculate the number of neighboring data points within a reference distance based on the Euclidean distance in the dimension of the number of data points. For example, when 140 data values ​​are collected, the 140-dimensional Euclidean distance can be used for distance calculation using the LOF algorithm. According to an embodiment, a reference distance can be determined such that the number of neighboring data points is a preset k.

[0059] According to an embodiment, controller 132 can be configured to diagnose the welding condition of battery 110 as defective when the LOF score exceeds a threshold score, and the threshold score can be determined based on tensile strength test data of a test battery manufactured through a welding process. For example, the test battery can be manufactured in a large-scale production line through a welding process, and the tensile strength value of the test battery can be tested. Based on the tensile strength test data, a threshold score can be set, and the defect detection performance and false alarm performance against normal welding can be verified based on the threshold score.

[0060] According to an embodiment, controller 132 can be configured to adjust the threshold score when consumables of welding equipment 120 performing the welding process are replaced. When consumables are replaced, the dispersion of electrical data collected from battery 110 can remain constant, but the average or center position of the electrical data may change. In this case, false alarms may occur when normal welding is classified as defective when the threshold score is kept at the same value, and therefore the threshold score can be adjusted based on the change in the average or center position.

[0061] According to an embodiment, the electrical data may include a first dataset below the cutoff frequency and one or more second datasets greater than or equal to the cutoff frequency, and the controller 132 may be configured to detect the time point of defect occurrence during the welding process based on the inflection points of the one or more second datasets when the welding condition is diagnosed as defective based on the first dataset. In conventional electrical data measurement, only the first dataset below the cutoff frequency is collected; however, in contrast, in the welding management system 100, data can also be collected for a frequency range greater than or equal to the cutoff frequency. That is, in the welding management system 100, the cutoff frequency can be significantly increased compared to the conventional cutoff frequency. As described below... Figure 5 As shown, in the case of one or more second datasets with a cutoff frequency greater than or equal to the cutoff frequency, the data inflection point can be shown more clearly than in the case of the first dataset with a cutoff frequency less than the cutoff frequency, thus enabling the detection of the time point at which the cause of the welding defect can be detected.

[0062] Figures 3 to 5 A method for collecting electrical data according to some embodiments is shown.

[0063] refer to Figure 3 The graph 310 shows a conventional method for measuring welding voltage and the graph 320 shows an improved method.

[0064] The conventional method in graph 310 collects data during periods when the welding current exceeds a specific value, while the improved method in graph 320 collects data during periods when the welding voltage exceeds a specific value. In the case of the improved method, since data collection is based on voltage rather than current, the welding execution segment and the data collection segment can overlap.

[0065] In the conventional method of graph 310, the same number of data points are collected regardless of the duration of the data collection segment. However, in the improved method of graph 320, a greater number of data points can be collected as the duration of the data collection segment becomes longer.

[0066] refer to Figure 4 Graph 410 can be shown to represent a conventional method for measuring dynamic resistance, and graph 420 can be shown to represent an improved method. As... Figure 3 The situation is the same as the welding voltage in the case of welding. Figure 4 In the case of dynamic resistance, data collection can be performed more appropriately by improving the method.

[0067] refer to Figure 5 The graph 510 shows the dynamic resistance values ​​measured in the range below the cutoff frequency and the graph 520 shows the dynamic resistance values ​​measured in the range above or equal to the cutoff frequency.

[0068] Graph 510 can represent the dynamic resistance value measured at a frequency of 1 kHz. For example, the cutoff frequency can be in the range of 1 kHz to 5 kHz. When the frequency of the welding voltage and / or welding current provided by the welding equipment 120 changes, the dynamic resistance value measured in the battery 110, which is the welding target, can also change. The dynamic resistance value can include an impedance value. Graph 520 can represent the dynamic resistance value measured in a frequency range exceeding the cutoff frequency. As shown, at frequencies such as 5 kHz, 10 kHz, 20 kHz, etc., the inflection point 521 of the change in the increasing / decreasing pattern of the dynamic resistance value can be shown more clearly. According to an embodiment, when the welding condition is diagnosed as defective based on the LOF score in the range below the cutoff frequency of graph 510, the inflection point 521, which is estimated to be the cause of the welding defect, can be detected in the range above or equal to the cutoff frequency of graph 520.

[0069] Figure 6 A method for calculating the Local Outlier Factor (LOF) score according to some embodiments is shown.

[0070] refer to Figure 6 The distribution of training data and electrical data used to calculate the LOF score can be shown in 610, and equations 620 to 650 can be used to calculate the LOF score.

[0071] In distribution 610, electrical data p and neighboring data o within a reference distance of electrical data p can be shown. According to an embodiment, the number k of neighboring data points o can be set to 5, and the reference distance to electrical data p can be set to a distance value that makes k 5. The LOF score of electrical data p can be calculated based on a first lrd value of electrical data p and 5 second lrd values ​​of the 5 neighboring data points o.

[0072] In equation 620, k, representing the number of neighboring data points o, can be 5. This can be based on the set of neighboring data points o. The second lrd value lrd k (o), and the first lrd value lrd k (p) to calculate the LOF score. k (p). |It can be The size. In Equation 630, it can be based on the set of neighboring data o. and reachability distance To calculate the first lrd value lrd k (p). The second lrd value can also be calculated in a similar manner. k (o).

[0073] In Equation 640, the set N of neighboring data o can be calculated based on the distance d(p,q) between the training data and the electrical data p and the distance k-distance(p) from the k-th nearest training data point to the electrical data p. k (p). In Equation 650, the reachability distance can be calculated based on the distance k-distance(p) and the distance d(p,q) from the k-th nearest training data point to the electrical data p. When the number of data values ​​of electrical data p is n, the distance d(p, q) can be calculated as an n-dimensional Euclidean distance.

[0074] Figure 7 A method for diagnosing weld condition based on LOF score, according to some embodiments, is shown.

[0075] refer to Figure 7 The diagram shows the training data 710 used to calculate the LOF score and the new data 720 displayed together with the training data 710. Although the data is represented in a simplified two-dimensional manner, LOF models can typically operate on higher dimensions, such as n=140.

[0076] The LOF model can be trained based on training data 710. Each data point in training data 710 can be electrical data measured during the welding process. For example, training data 710 and new data 720 can represent any of the data on dynamic resistance, welding voltage, welding current, and welding temperature.

[0077] The new data 720 may include new data 1 located in a distribution close to the training data 710 and new data 2 located in a distribution far from the training data 710. This can be compared with... Figure 6 The LOF scores for new data 1 and new data 2 are calculated in the same way. A LOF score for new data 1 with a value close to 1 can be calculated, which can be classified as normal welding. On the other hand, a LOF score for new data 2 can be calculated with a value higher than 1, which can have a high probability of being classified as defective welding.

[0078] Figure 8 A method for classifying weld strength based on LOF scores, according to some embodiments, is shown.

[0079] refer to Figure 8 The graph 810 can be shown to represent a conventional method for monitoring welding status based on electrical data, and the graph 820 can be shown to represent an improved method.

[0080] In the conventional method of curve 810, the model score may not show a significant difference even when the weld strength changes. On the other hand, in the improved method of curve 820 according to the welding management system 100, over-welding 821, normal welding 822, and under-welding 823 can be classified into scores that are distinguishable from each other.

[0081] Figure 9 The process of setting a threshold score based on the relationship between tensile strength and LOF score according to some embodiments is shown.

[0082] refer to Figure 9 The graph 900 can be shown, which illustrates the relationship between tensile strength and LOF score measured for multiple test cells.

[0083] As shown in graph 900, the tensile strength and LOF score of the tested battery can be significantly correlated with each other. Quantitatively, the tensile strength and LOF score in graph 900 can represent R at approximately 80% level. 2 The correlation level of 80% can be observed not only across the entire range, but also in the region of interest near the tensile strength specification value of battery 110.

[0084] The threshold score 910 can be set in Figure 900. When the tensile strength specification of the negative terminal of the battery is approximately 0.9 kgf, the 50% level of the average tensile strength test data can be approximately 1.51 kgf, and test batteries with this as the control limit can be produced as samples. In this case, the LOF score corresponding to 1.51 kgf can be set as the threshold score 910, and even in this case, it can be confirmed that false alarms where normal welds are incorrectly classified as defective do not increase. In this way, the threshold score 910 can be set based on the tensile strength test data of the test battery.

[0085] Figure 10 A method for adjusting a threshold fraction when replacing consumables in welding equipment, according to some embodiments, is shown.

[0086] refer to Figure 10 A graph 1000 can be shown, which illustrates a method for adjusting the threshold fraction when replacing consumables in welding equipment 120.

[0087] When consumables are replaced, the average position of the LOF score may shift. In this case, if the threshold score remains unchanged, a large number of false alarms may occur, where normal welds are incorrectly diagnosed as defective. To address this issue, the shift amount can be calculated by collecting 50 data points after replacing the consumables of the welding equipment 120, and the threshold score can be adjusted by the same amount.

[0088] Figure 11 The steps for constructing a welding management method according to some embodiments are shown.

[0089] refer to Figure 11 The welding management method 1100 may include steps 1110 to 1140. However, the welding management method is not limited to this, and some steps may be omitted or other general steps may be added, and the steps of the welding management method 1100 may be performed in a different order than that shown.

[0090] Welding management method 1100 may include time-series processing steps in welding management equipment 130. Therefore, even if the following description is omitted, the content described above for welding management equipment 130 can be equivalently applied to welding management method 1100.

[0091] Steps 1110 to 1140 of the welding management method 1100 can be executed by the interface 131 and controller 132 of the welding management device 130.

[0092] In step 1110, the welding management device 130 may perform the step of acquiring electrical data collected during the welding process of the battery.

[0093] In step 1120, the welding management device 130 may perform a first lrd step of calculating electrical data based on training data and electrical data stored in the LOF model.

[0094] In step 1130, the welding management device 130 may perform the step of calculating the LOF score of the electrical data based on the second lrd and the first lrd of the neighboring data within the reference distance of the electrical data in the training data.

[0095] In step 1140, the welding management device 130 can perform a step of diagnosing the welding status of the battery based on the LOF score.

[0096] According to an embodiment, the welding management method 1100 can be implemented as a computer program stored in a computer-readable storage medium. That is, the computer program can include commands for implementing the welding management method 1100, and the commands of the program can be stored in a computer-readable storage medium. The computer program can include a mobile application.

[0097] In embodiments, computer-readable storage media may include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floppy disks; and hardware devices specifically configured to store and execute computer program instructions, such as ROMs, RAMs, and flash memory. Computer program instructions may include machine language code generated by a compiler and high-level language code that can be executed by a computer using an interpreter or similar tool.

[0098] Unless otherwise stated, terms such as “comprising,” “including,” or “having” above mean that the corresponding component may be present, and therefore should be understood as potentially including, rather than excluding, other components. Unless otherwise defined, all terms including technical or scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments disclosed herein pertain. Terms of ordinary use, such as those defined in dictionaries, should be interpreted as having a meaning consistent with their meaning in the context of the relevant field and will not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0099] The above description is merely an example of the technical concepts disclosed herein, and those skilled in the art to which the embodiments disclosed herein pertain can make various modifications and variations without departing from the basic characteristics of the embodiments disclosed herein. Therefore, the embodiments disclosed herein are not intended to limit the technical concepts of the embodiments disclosed herein, but rather to explain them, and the scope of the technical concepts disclosed herein is not limited by these embodiments. The scope of protection disclosed herein should be interpreted by the appended claims, and all technical ideas within the same scope should be interpreted as being included within the scope of the claims herein.

[0100] (List of reference numerals in the attached image)

[0101] 100: Welding Management System 110: Battery

[0102] 120: Welding equipment; 130: Welding management equipment

[0103] 131: Interface 132: Controller

Claims

1. A welding management device, comprising: An interface configured to acquire electrical data collected during the battery welding process; as well as The controller is configured to: The first local reachability density (lrd) of the electrical data is calculated based on the training data stored in the Local Outlier Factor (LOF) model and the electrical data. The LOF score of the electrical data is calculated based on the second lrd and the first lrd of the neighboring data within the reference distance of the electrical data in the training data; and The welding condition of the battery is diagnosed based on the LOF score.

2. The welding management equipment according to claim 1, wherein, The electrical data includes dynamic resistance values ​​collected at a first measurement frequency during periods when the welding voltage exceeds a reference voltage during the welding process, and The LOF model is configured to calculate the LOF score based on the dynamic resistance value.

3. The welding management equipment according to claim 2, wherein, The controller is configured to calculate the LOF score by inputting the value of the dynamic resistance value after a first time has elapsed from the start time of the period exceeding the reference voltage into the LOF model.

4. The welding management equipment according to claim 2, wherein, The controller is configured to: Based on the time period during which the welding voltage exceeds the reference voltage and the first measurement frequency, the number of data points for the dynamic resistance value is determined; and The number of neighboring data points within the reference distance is calculated based on the Euclidean distance, with the number of data points as the dimension.

5. The welding management equipment according to claim 1, wherein, The controller is configured to diagnose the welding condition of the battery as defective when the LOF score exceeds a threshold score, and The threshold score is determined based on the tensile strength test data of the test cells manufactured through the welding process.

6. The welding management equipment according to claim 5, wherein, The controller is configured to adjust the threshold score when consumables of the welding equipment performing the welding process are replaced.

7. The welding management equipment according to claim 1, wherein, The electrical data includes a first dataset less than the cutoff frequency and one or more second datasets greater than or equal to the cutoff frequency, and The controller is configured to detect the time point at which a defect occurs during the welding process based on the inflection points of the one or more second datasets when the welding state is diagnosed as defective based on the first dataset.

8. A welding management method, comprising: Acquire electrical data collected during the battery welding process; The first local reachability density (lrd) of the electrical data is calculated based on the training data and electrical data stored in the Local Outlier Factor (LOF) model. The LOF score of the electrical data is calculated based on the second lrd and the first lrd of the neighboring data within the reference distance of the electrical data in the training data; and The welding condition of the battery is diagnosed based on the LOF score.

9. The welding management method according to claim 8, wherein, The electrical data includes dynamic resistance values ​​collected at a first measurement frequency during periods when the welding voltage exceeds a reference voltage during the welding process, and The LOF model is configured to calculate the LOF score based on the dynamic resistance value.

10. The welding management method according to claim 9, wherein, Calculating the LOF score involves inputting the value of the dynamic resistance value after a first time has elapsed from the start time of the period exceeding the reference voltage into the LOF model.

11. The welding management method according to claim 9, wherein, Calculating the LOF score includes: Based on the time period during which the welding voltage exceeds the reference voltage and the first measurement frequency, the number of data points for the dynamic resistance value is determined; and The number of neighboring data points within the reference distance is calculated based on the Euclidean distance, with the number of data points as the dimension.

12. The welding management method according to claim 8, wherein, The diagnosis of the welding condition includes identifying the welding condition of the battery as defective when the LOF score exceeds a threshold score, and The threshold score is determined based on the tensile strength test data of the test cells manufactured through the welding process.

13. The welding management method according to claim 12, further comprising: The threshold score is adjusted when the consumables of the welding equipment performing the welding process are replaced.

14. The welding management method according to claim 8, wherein, The electrical data includes a first dataset less than the cutoff frequency and one or more second datasets greater than or equal to the cutoff frequency, and The welding management method further includes: when the welding state is diagnosed as defective based on the first dataset, detecting the time point at which the defect occurs during the welding process based on the inflection points of the one or more second datasets.

15. A welding management system, comprising: Battery; A welding device configured to perform a welding process on the battery; as well as A welding management device is configured to acquire electrical data collected during the welding process, calculate a first local reachability density (lrd) of the electrical data based on training data and electrical data stored in a Local Outlier Factor (LOF) model, calculate an LOF score of the electrical data based on a second lrd and the first lrd of neighboring data within a reference distance of the electrical data in the training data, and diagnose the welding status of the battery based on the LOF score.

Images