A method and system for real-time acquisition and monitoring of data of a formation and distribution system

Abstract

The present application relates to the technical field of data monitoring of chemical composition and capacity system, and provides a kind of chemical composition and capacity system data real-time acquisition monitoring method and system, its method includes: the running data of each battery in chemical composition stage or sub-volume stage is collected, and based on running data determination abnormal starting battery;According to abnormal starting battery, stop instruction is issued to corresponding control unit, to stop the charge-discharge operation of abnormal starting battery;The adjacent influence degree of the adjacent battery of abnormal starting battery is obtained;According to adjacent influence degree, stop instruction is issued to adjacent battery that meets influence determination condition, to stop the charge-discharge operation of adjacent battery;For the local area where abnormal starting battery is located, start enhanced heat dissipation measures;According to the current state of unaffected battery, adjust the charge-discharge strategy of unaffected battery.The present application has the effect of improving the production efficiency and safety of tray battery.

Description

Technical Field

[0001] This invention relates to the technical field of data monitoring in a formation and formulation system, specifically to a method and system for real-time data acquisition and monitoring in a formation and formulation system. Background Technology

[0002] In modern battery production, formation and capacity testing are critical steps that determine the initial performance and safety of batteries. However, in high-density production environments, the impact of heat transfer between cells, coupled with the inherent time lag between data acquisition and analysis, often makes it difficult to detect and accurately locate anomalies in a timely manner. This renders traditional monitoring and early warning methods inadequate and unable to meet the high efficiency and safety requirements of production lines.

[0003] Specifically, in a large lithium-ion battery production workshop, newly assembled cells undergo a formation and capacity testing process. These cells are neatly placed on dedicated trays and then fed into a tall formation and capacity testing cabinet by automated equipment. The system performs precise charge and discharge operations on each cell according to a preset program to activate the internal chemical substances and accurately measure its final usable capacity. During this process, the control system continuously collects three key operational data points for each cell: voltage, current, and surface temperature. This information is initially collected by a lower-level controller within the capacity testing cabinet and then uploaded to the workshop's central monitoring server via wired or wireless network. Software on the server analyzes this large amount of data to determine whether each cell is functioning correctly.

[0004] To improve production efficiency and output per unit area, modern production lines generally employ high-density cell arrangement. On the trays of the capacity sorting cabinet, the spacing between cells is compressed to a minimum, typically only a few millimeters. While this compact layout saves space, it also introduces a significant physical phenomenon: the heat transfer effect between cells becomes extremely pronounced. The heat generated by any cell during charging and discharging will inevitably be conducted and radiated to its neighboring cells.

[0005] However, unexpected situations can occur. Suppose that a battery cell located in the center of the charging tray experiences a barely perceptible micro-short circuit due to a tiny impurity introduced during manufacturing. In the initial stages of charging, this micro-short circuit generates additional Joule heat, causing the cell's temperature to rise at a slightly faster rate than normal cells. Due to the existence of data acquisition and transmission links, there is an inherent time lag—potentially hundreds of milliseconds to several seconds—in the entire process from the cell's surface temperature sensor reading to the lower-level computer processing and then uploading to the central server. By the time this abnormal temperature information reaches the central server, the heat generated by the faulty cell has already caused the surface temperatures of several adjacent cells to begin to rise slightly and abnormally through short-distance heat conduction.

[0006] At this point, the monitoring software on the central server receives a picture of multiple battery cells in a region simultaneously showing slightly elevated temperatures. Traditional analysis logic typically relies on the independent threshold of individual cells, checking if any cell's temperature exceeds a preset safety limit. In this scenario, the temperature of the faulty cell may not yet have reached the level to trigger an alarm, while the temperatures of surrounding cells are only slightly elevated, also within safe limits. Faced with such a group of small-scale, collective temperature increases, the analysis software struggles to accurately pinpoint the root cause of the problem immediately. It might misinterpret this as poor ventilation within the capacity cabinet or a collective zero-point drift of temperature sensors in a certain area. The system cannot connect these seemingly isolated but physically related data points to trace the initial source of heat.

[0007] This delay and ambiguity in diagnosis prevents the system from taking the most effective countermeasures. The conventional approach is that when an alarm is finally triggered (usually when the temperature of the faulty cell rises sharply and exceeds a threshold), the system issues a broad instruction, such as stopping the charging and discharging of all cells in the entire tray and initiating powerful cooling. This not only forces the testing of all normal cells in the tray to be interrupted, wasting valuable production time, but also requires manual intervention to check each of the hundreds of cells in the tray one by one to find the real "bad point," making the entire process inefficient. Even more dangerous is that if the micro-short circuit problem worsens very rapidly, this valuable diagnostic window may be missed, and the faulty cell may enter a critical state of thermal runaway. Even if the system eventually takes action, it may be too late, threatening the safety of the equipment and even the entire workshop.

[0008] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0009] This application discloses a real-time data acquisition and monitoring method and system for a formation and capacity testing system, which aims to solve the problem that in existing formation and capacity testing systems, due to the influence of heat transfer between cells and the time difference in data acquisition and analysis, abnormal situations are difficult to detect and accurately locate in a timely manner, and traditional monitoring and early warning methods cannot meet the high requirements of production lines for efficiency and safety.

[0010] The technical solution of this application is as follows:

[0011] In a first aspect, this application discloses a method for real-time data acquisition and monitoring of a formulation and capacity-building system, comprising:

[0012] The system collects operational data from each cell during the formation or capacity testing phase, and identifies the cell that initiates the anomaly based on the operational data. The cell that initiates the anomaly is the cell that first meets the criteria after matching the electrochemical state characteristics with the electrochemical anomaly pattern and combining the results of local cross-validation and confidence scores. This cell is used to characterize the cell that is the initial source of the anomaly.

[0013] Based on the abnormal starting cell, a stop command is sent to the corresponding control unit to stop the charging and discharging operation of the abnormal starting cell;

[0014] Obtain the proximity influence of neighboring cells of the abnormal starting cell; the proximity influence is used to characterize the combined effect of thermal transfer and electromagnetic coupling on neighboring cells of the abnormal starting cell.

[0015] Based on the degree of proximity influence, a stop command is issued to the neighboring cells that meet the influence determination conditions to stop the charging and discharging operations of the neighboring cells.

[0016] For the localized area where the abnormal battery cell is located, enhanced heat dissipation measures will be implemented;

[0017] Adjust the charging and discharging strategies for unaffected cells based on their current status.

[0018] Through this technical solution, this application can collect the operating data of each cell in the formation stage or capacity testing stage in real time, and accurately identify the cell that starts to be abnormal. This allows for timely cessation of the charging and discharging operations of the abnormal cell and its affected neighboring cells, as well as the initiation of local enhanced heat dissipation. At the same time, it optimizes the charging and discharging strategies of unaffected cells. This effectively solves the problem that traditional monitoring methods are difficult to detect and accurately locate abnormalities in a timely manner when faced with high-density cell arrangements, and significantly improves production efficiency and safety.

[0019] Secondly, this application also discloses a real-time data acquisition and monitoring system for a formation and formulation system, used to perform real-time data acquisition and monitoring of the formation and formulation system, including:

[0020] The abnormal cell determination module is used to collect the operating data of each cell during the formation or capacity testing stage, and to determine the cell that initiates the abnormality based on the operating data. The cell that initiates the abnormality is the cell that first meets the determination criteria after matching the electrochemical state characteristics with the electrochemical abnormality mode, and combining the results of local cross-validation and confidence score, within a preset time window. This cell is used to characterize the cell that is the initial source of the abnormal event.

[0021] The stop command issuing module is used to issue a stop command to the corresponding control unit based on the abnormal starting cell, so as to stop the charging and discharging operation of the abnormal starting cell;

[0022] The proximity influence acquisition module is used to acquire the proximity influence degree of the neighboring cells of the abnormal starting cell; the proximity influence degree is used to characterize the combined effect of the thermal transfer and electromagnetic coupling of the abnormal starting cell on the neighboring cells.

[0023] The neighboring cell stop module is used to issue a stop command to neighboring cells that meet the influence determination conditions based on the degree of proximity influence, so as to stop the charging and discharging operation of the neighboring cells.

[0024] The heat dissipation activation module is used to activate enhanced heat dissipation measures for the local area where the abnormal starting cell is located;

[0025] The charge / discharge strategy adjustment module is used to adjust the charge / discharge strategy of unaffected cells based on their current state.

[0026] Through this technical solution, the system provided in this application can realize real-time, accurate monitoring and rapid response to cell anomalies in the formation and capacity testing system. Through modular design, it effectively coordinates various functional units, solves the lag and limitations of traditional systems in anomaly detection, location and handling, and significantly improves the safety, efficiency and intelligence level of the formation and capacity testing process.

[0027] Beneficial Effects: The real-time data acquisition and monitoring method for the formation and capacity testing system disclosed in this application effectively solves the problem of difficulty in timely detection and accurate location of anomalies caused by heat transfer between cells and time differences in data acquisition and analysis in existing technologies. Specifically, this method defines the anomalous initiating cell as the initial source cell determined within a preset time window by combining spectral fluctuation components, internal resistance change trends, electrochemical anomaly mode matching results, local cross-validation results, and confidence scores. This allows for more accurate differentiation between the actual anomaly source cell and neighboring cells affected by heat transfer or electromagnetic coupling in the formation and capacity testing scenario. Based on this, the system can quickly issue a stop command to the corresponding control unit according to the anomalous initiating cell, stopping the charging and discharging operation of the anomalous cell and preventing further deterioration of the anomaly. Simultaneously, this method further acquires the proximity influence degree of neighboring cells of the cell initiating the anomaly. This influence degree comprehensively characterizes the effects of heat transfer and electromagnetic coupling, enabling the system to issue stop commands to neighboring cells that meet the influence judgment conditions based on the proximity influence degree, thereby effectively preventing the spread of the anomaly. Furthermore, for the local area where the cell initiating the anomaly is located, the system can activate enhanced heat dissipation measures to further control the development of the anomaly. Finally, based on the current state of unaffected cells, the system can adjust their charging and discharging strategies to ensure production continuity and efficiency. Through the above technical solutions, this application overcomes the problems of difficulty in tracing the root cause of "group" temperature rise data, diagnostic delays, and low processing efficiency in existing technologies. It achieves early and accurate warning and localized, refined processing of cell anomalies, significantly improving the safety, production efficiency, and intelligence level of the formation and capacity testing process, and avoiding the resource waste and potential safety risks of shutting down the entire tray of cells. Attached Figure Description

[0028] Figure 1 This is a flowchart of a method for real-time data acquisition and monitoring of a formulation and capacity system according to one embodiment of the present invention;

[0029] Figure 2 This is a flowchart of a method for real-time data acquisition and monitoring of a formulation and capacity system according to another embodiment of the present invention;

[0030] Figure 3 This is a system block diagram of a real-time data acquisition and monitoring system for a chemical composition and capacity setting system, as described in another embodiment of the present invention.

[0031] Explanation of reference numerals in the attached figures:

[0032] 1. Real-time data acquisition and monitoring system for the formation and capacity testing system; 11. Abnormal cell determination module; 12. Stop command issuance module; 13. Proximity influence acquisition module; 14. Proximity cell stop module; 15. Heat dissipation measure activation module; 16. Charge and discharge strategy adjustment module. Detailed Implementation

[0033] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0034] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0035] This application proposes a method for real-time data acquisition and monitoring of a formation and dispensing system, combined with... Figure 1 As shown, it includes:

[0036] S1. Collect the operating data of each cell during the formation stage or capacity testing stage, and determine the cell that initiates the abnormality based on the operating data. The cell that initiates the abnormality is the cell that first meets the judgment criteria after matching the electrochemical state characteristics with the electrochemical abnormality mode, and combining the local cross-validation results and confidence scores within a preset time window. This cell is used to characterize the cell that is the initial source of the abnormal event.

[0037] S2, based on the abnormal starting cell, sends a stop command to the corresponding control unit to stop the charging and discharging operation of the abnormal starting cell;

[0038] S3, obtain the proximity influence degree of the neighboring cells of the abnormal starting cell; the proximity influence degree is used to characterize the combined effect of the thermal transfer and electromagnetic coupling of the abnormal starting cell on the neighboring cells.

[0039] S4, based on the degree of proximity influence, issue a stop command to the neighboring cells that meet the influence determination conditions to stop the charging and discharging operations of the neighboring cells;

[0040] S5 activates enhanced heat dissipation measures for the local area where the abnormal starting cell is located;

[0041] S6 adjusts the charging and discharging strategy of the unaffected cells based on their current state.

[0042] To better understand the real-time data acquisition and monitoring method for the formulation and compatibility system proposed in this application, some key terms involved will be explained first.

[0043] "Abnormal initiation cell" refers to the cell that, within a preset time window, first meets the judgment criteria after constructing electrochemical state characteristics based on spectral fluctuation components and internal resistance change trends extracted from operational data, matching these electrochemical state characteristics with electrochemical abnormality patterns, and combining local cross-validation results and confidence scores. The necessity of this concept lies in the fact that, in a formation and capacity testing system, the abnormal source cell and the affected cell may simultaneously exhibit temperature rise, impedance fluctuations, or abnormal charge / discharge states within a short period. If judgment is based solely on a single threshold or a single temporal sequence, the affected cell may easily be misclassified as the abnormal source cell. By combining spectral fluctuation components, internal resistance change trends, pattern matching results, and local cross-validation results for the determination of the abnormal initiation cell, the accuracy of identifying the initial source of the abnormality is improved.

[0044] Furthermore, the determination of the abnormal starting cell in this application is not a general battery management judgment for vehicle operation or energy storage operation scenarios, but rather a process scenario where cells are subject to long-term controlled charging and discharging, high-density array arrangement, significant local heat transfer, and the measurement environment is susceptible to strong electromagnetic interference during the formation or capacity testing stages. In this scenario, the true starting point of an anomaly often manifests only as a weak perturbation at the spectral level and a corrected shift in the internal resistance trend in the early stages, without developing into a significant macroscopic temperature rise or voltage exceedance. Therefore, this application uses a joint characterization of spectral fluctuation components and internal resistance change trends, and suppresses the interference of adjacent affected cells on the identification of the anomaly source through local cross-validation, thereby making the determination of the abnormal starting cell more suitable for the formation and capacity testing process environment.

[0045] "Proximity Influence" is an indicator used to characterize the combined effect of thermal transfer and electromagnetic coupling on neighboring cells during the formation or capacity testing stages of an abnormal cell. Setting this indicator is necessary because during formation and capacity testing, the cells are arranged in a high-density array. The temperature rise, current disturbance, and local electromagnetic field changes generated by the abnormal cell will act on neighboring cells along both the thermal conduction and electromagnetic coupling paths. If judgment is based solely on physical distance or a single temperature change, it is difficult to accurately distinguish the affected range and degree of influence. However, by combining the effects of thermal transfer and electromagnetic coupling to form the proximity influence, it is beneficial to achieve localized and differentiated protection and control.

[0046] In a preferred embodiment, neighboring cells are subject to tiered control based on the magnitude of their proximity impact. For neighboring cells with a high-risk proximity impact level, their charging and discharging operations are directly stopped, and the heat dissipation capacity of the corresponding area is simultaneously increased. For neighboring cells with a medium-risk proximity impact level, their charging and discharging rates are reduced or their charging and discharging cycles are adjusted, while their operating status is continuously monitored. For neighboring cells with a low-risk proximity impact level, their operation is maintained, but their data sampling frequency and anomaly detection sensitivity are increased. Through this tiered control method, a unified shutdown of the entire tray of cells can be avoided after an anomaly occurs, thereby ensuring safety while also meeting the continuous operation requirements of the formation and capacity testing production line.

[0047] The "electrochemical anomaly pattern matching threshold" and "confidence score" are crucial parameters for determining whether a battery cell is abnormal. The electrochemical anomaly pattern matching threshold describes the lower limit of similarity between the cell's operating data and known abnormal electrochemical behavior patterns, while the confidence score characterizes the reliability of the matching result. Setting these two parameters is necessary because battery cells may experience normal fluctuations within a certain range during operation. Relying solely on a single threshold to determine anomalies could lead to false positives or false negatives. By simultaneously setting the pattern matching threshold and the confidence score, both the intensity of abnormal features and the reliability of identification can be considered during anomaly detection, thereby improving the stability of anomaly determination.

[0048] Based on the above terminology, this application provides a method for real-time data acquisition and monitoring of a formulation and capacity control system.

[0049] First, it is necessary to collect operational data for each cell during the formation or capacity testing stages. This operational data can include parameters such as cell voltage, current, temperature, and internal resistance. These parameters directly reflect the changes in the electrochemical state of the cell during the formation and capacity testing process; therefore, continuous data collection is fundamental for subsequent anomaly identification and state analysis. Without continuous acquisition of operational data, it is impossible to determine whether the cell state has changed, and it is also impossible to establish a basis for anomaly identification.

[0050] Specifically, voltage, current, and temperature sensors can be used to monitor each battery cell in real time, and the collected data can be transmitted to a data processing unit for storage and analysis. For example, a high-precision voltage sampling module, a current detection module, and temperature detection elements such as thermocouples or thermistors can be used to achieve continuous monitoring of the operating status of each battery cell. In another implementation, an integrated multi-functional sensor can be used to jointly collect voltage, current, and temperature data, and the data can be transmitted to the monitoring system via wired or wireless communication.

[0051] After acquiring the operational data, it is necessary to determine the cell that initiated the anomaly based on this data. This determination step is essential because when an anomaly occurs in the system, the anomaly often does not appear simultaneously in all cells. Instead, it usually starts in one cell and gradually affects the surrounding cells. If the cell that initially exhibited the anomaly cannot be identified, it is difficult to determine the location of the anomaly source and to take targeted isolation measures.

[0052] Specifically, a baseline model of the normal operating state of a battery cell can be established, and the deviation between the current operating data and the baseline model can be compared in real time. When the deviation exceeds a preset threshold, time series analysis methods can be used to identify the battery cell that first exhibits abnormal characteristics in time, thereby determining the battery cell that initiated the abnormality. In another implementation, a data-driven model can be used to analyze the operating state of the battery cell. By training a sample set containing both normal and abnormal data, the model can identify abnormal electrochemical behavior patterns and determine the battery cell that initiated the abnormality based on confidence scores and time information.

[0053] Once an abnormal cell is identified as the initiating cell, a stop command needs to be issued to the corresponding control unit to halt its charging and discharging operations. This step is necessary because continued participation in the charging and discharging process when a cell is malfunctioning may further worsen the abnormal state, such as causing a sustained increase in temperature or runaway electrochemical reactions. Therefore, by promptly stopping the charging and discharging of the abnormal cell, it can be isolated before the abnormality spreads, thereby reducing the impact on other cells in the system.

[0054] Specifically, the control unit can be a control module located in the formation and capacity testing equipment. Upon receiving a stop command, it cuts off the charging and discharging circuit of the abnormal cell by controlling a relay, power switch, or electronic switch. In another implementation, the control unit can also adjust the cell's charging and discharging control parameters to put it into standby or low-power mode, thereby stopping its participation in the formation and capacity testing process.

[0055] Subsequently, it is necessary to obtain the proximity impact of the adjacent cells of the cell that initiated the abnormality. This step is essential because, in a multi-cell system, although the cell that initiated the abnormality has been isolated, the heat and electromagnetic disturbances it generates may still continue to affect surrounding cells for a short period. Without assessing the impact on adjacent cells, it is difficult to determine which cells require further protection.

[0056] Specifically, by establishing a thermal conduction model and an electromagnetic coupling model, and combining parameters such as the real-time temperature and current change rate of the abnormally initiated cell, as well as its physical distance from neighboring cells, the degree of influence on neighboring cells can be calculated, thus obtaining the proximity influence degree. Alternatively, additional temperature or electromagnetic field sensors can be placed near neighboring cells to obtain actual measurements of heat transfer and electromagnetic coupling strength, and the proximity influence degree can be calculated accordingly.

[0057] After obtaining the proximity influence level, a stop command is issued to neighboring cells that meet the influence determination criteria, thereby halting their charging and discharging operations. This step is necessary because when a neighboring cell is significantly affected by the abnormally initiated cell, continued participation in the charging and discharging process could lead to an abnormal state within a short period. Therefore, by implementing preventative control over neighboring cells based on the proximity influence level, the risk of abnormal spread can be reduced.

[0058] For example, a proximity influence threshold can be set. When the proximity influence of a neighboring cell exceeds this threshold, it is considered to meet the influence determination condition, and a stop command is issued to it. In another implementation, a hierarchical control strategy can be set according to the magnitude of the proximity influence. For example, charging and discharging of cells with high influence can be stopped immediately, while charging and discharging rates can be reduced for cells with low influence.

[0059] Simultaneously, enhanced heat dissipation measures are initiated for the localized area where the abnormal cell originates. This step is necessary because even after charging and discharging operations cease, the abnormal cell may still maintain a high temperature for a short period. If heat dissipation is not addressed promptly, the heat may continue to transfer to surrounding areas, affecting other cells.

[0060] For example, local fans, liquid cooling systems, or spray cooling devices can be activated to force heat dissipation from the abnormal area. Alternatively, the operating status of the overall cooling system of the container can be adjusted to provide higher heat dissipation capacity to the abnormal area.

[0061] Finally, adjust the charging and discharging strategies for the unaffected cells based on their current status. This step is necessary because if the operating strategies for the remaining cells are not adjusted after isolating the abnormal cells and some neighboring cells, the overall production efficiency of the formation and capacity testing system may be affected. Therefore, by appropriately adjusting the operating strategies for the unaffected cells, the continuity of production can be maintained while ensuring system safety.

[0062] For example, the charging and discharging current, voltage, or charging and discharging time of unaffected cells can be dynamically adjusted based on their remaining capacity, health status, and production plans. In another implementation, unaffected cells can be divided into multiple groups, and different charging and discharging tasks can be assigned to different groups, thereby optimizing overall production efficiency.

[0063] In summary, the method of this application, through the processing flow of "operational data acquisition, abnormal initiation cell determination, abnormal cell isolation, proximity impact assessment, proximity cell protection, local heat dissipation control, and unaffected cell strategy adjustment", enables the formation and capacity testing system to promptly identify the source of abnormality and control its impact range when an abnormality occurs, while taking into account system operating efficiency, thereby improving the stability and safety of system operation.

[0064] Optional, combined Figure 2 As shown, the steps of collecting operational data from each cell during the formation or capacity testing phase, and determining the cell at the onset of an abnormality based on the operational data, can be further refined into the following process. This process includes:

[0065] A1 collects the voltage, current and internal resistance of each cell during the formation or capacity testing stage as operating data;

[0066] A2 performs adaptive filtering on the running data to obtain the processed data;

[0067] A3 performs spectral analysis on the voltage and current in the processed data to extract the fluctuation components within a specific frequency range;

[0068] A4, based on the fluctuation components and combined with the changing trend of internal resistance in the processed data, constructs electrochemical state characteristics;

[0069] A5 matches electrochemical state characteristics with electrochemical anomalous patterns in the anomalous pattern library;

[0070] A6 identifies cells that match an electrochemical anomaly pattern and meet threshold conditions as the initiation cells of anomalies.

[0071] The acquisition of voltage, current, and internal resistance of each cell during the formation or capacity testing stages as operational data refers to the continuous acquisition of key electrical parameters reflecting changes in the electrochemical state of the cell during the charging and discharging process using corresponding data acquisition units or sensor units. Voltage, current, and internal resistance are collected together as operational data because a single parameter typically only reflects one aspect of the cell's state and is insufficient to fully characterize the abnormal evolution process. For example, relying solely on voltage changes is easily affected by operating condition switching and sampling disturbances; relying solely on current changes makes it difficult to distinguish between external control changes and internal abnormal changes within the cell; and while internal resistance can reflect the degree of cell aging and changes in internal transmission state, it is difficult to promptly reflect the dynamic fluctuation characteristics when an anomaly occurs when used alone. Therefore, jointly acquiring voltage, current, and internal resistance is beneficial for simultaneously obtaining transient response information and slowly changing state information of the cell, thus providing a more complete data foundation for subsequent anomaly identification. Specifically, voltage is used to reflect the terminal voltage response characteristics of the cell at different charging and discharging stages, current is used to reflect the charging and discharging rate and load changes, and internal resistance is used to characterize the changes in ohmic impedance and polarization impedance inside the cell. The changes in internal resistance are often related to cell aging, lithium plating, local micro-short circuits or interface degradation, so it is necessary to collect them synchronously.

[0072] Furthermore, adaptive filtering is performed on the operating data to obtain processed data. This involves noise suppression and disturbance removal processing of the acquired voltage, current, and internal resistance data to reduce the impact of acquisition link noise, environmental electromagnetic interference, operating condition switching spikes, and random measurement errors on subsequent analysis. Setting up an adaptive filtering step is necessary because the number of cells in the formation and capacity testing system is large, and the sampling environment is complex. The operating data is often mixed with background noise, sampling jitter, channel interference, and instantaneous control fluctuations. If spectrum analysis or anomaly pattern matching is directly performed on the unprocessed operating data, non-abnormal disturbances are easily misidentified as abnormal features, or the weak signal changes corresponding to early cell anomalies are masked, thus affecting the accuracy of determining the cell at the inception of the anomaly. The necessity of using adaptive filtering instead of fixed-parameter filtering lies in the fact that noise characteristics and effective signal characteristics are not constant under different operating conditions. Fixed filtering parameters cannot simultaneously achieve both noise suppression and preservation of weak anomaly features. Adaptive filtering, on the other hand, can dynamically adjust the filtering parameters according to the statistical characteristics of the signal and noise, so that the processed data, while reducing the impact of background noise, retains as much subtle fluctuation information related to electrochemical anomalies as possible, making it more suitable for subsequent anomaly identification and analysis.

[0073] Specifically, performing spectral analysis on the voltage and current in the processed data to extract fluctuation components within a specific frequency range involves converting the voltage and current signals from time-domain representation to frequency-domain representation to identify periodic fluctuations, quasi-periodic fluctuations, or localized anomalous frequency responses occurring within a specific frequency range. Setting up a spectral analysis step is necessary because when an electrochemical anomaly occurs in a battery cell, it doesn't necessarily manifest as a large change in macroscopic amplitude initially; more commonly, it presents as weak but specific frequency-characteristic anomalous fluctuations in the voltage and current signals. If only the original curve is observed in the time domain, some early anomalous signals may be masked by normal charging and discharging fluctuations, making timely differentiation difficult. Spectral analysis, however, can separate the frequency structure hidden within the time-domain signal, making it easier to identify fluctuation components related to the anomaly. Here, fluctuation components can be understood as the set of spectral fluctuations extracted within a preset frequency range, while specific frequency components are subsets of frequency points or bands within these fluctuation components that are strongly correlated with the electrochemical anomaly. For example, local short circuits, lithium plating, interfacial film rupture, or enhanced local side reactions may cause voltage and current to exhibit energy distribution changes distinct from normal operating conditions within certain frequency ranges. Therefore, extracting fluctuation components within specific frequency ranges through spectral analysis is beneficial for improving the ability to detect early anomalies. A specific frequency refers to a frequency position in the frequency domain representation of cell operating data that corresponds to or is correlated with the electrochemical anomaly evolution process. It does not refer to an arbitrary, fixed, single frequency value, but rather to a target frequency region or target frequency point that, after analysis, is considered to have discriminative significance for anomaly identification under specific process stages, local temperature conditions, aging states, and noise backgrounds. The purpose of this setting is to transform weak anomalous disturbances that are not easily identified directly in the time domain into separable and localizable frequency domain features, thereby improving the ability to identify early anomalies. A specific frequency range refers to a frequency domain analysis interval set around the anomaly identification target. It is used to limit the frequency boundaries that are the focus of extraction and observation during spectrum analysis. It is not constant throughout all operating stages, but is dynamically adjusted according to the charging and discharging stage of the battery cell, local temperature, and aging state. This makes the spectrum analysis more focused on the frequency bands where anomalies are more likely to appear, while reducing the interference of irrelevant noise frequency bands on the analysis results. The significance of setting a specific frequency range is to avoid averaging the entire frequency band, which would cause the abnormal frequency bands to be submerged by noise or diluted by invalid frequency bands.

[0074] Based on this, and combining the fluctuation components with the changing trends of internal resistance in the processed data, electrochemical state characteristics are constructed. This involves jointly characterizing the dynamic fluctuation information obtained from spectral analysis with the changing trends of internal resistance over time, thereby forming a set of features to describe the current electrochemical state of the cell. This step is necessary because while using only the fluctuation components of voltage and current can reflect abnormal signs at the dynamic response level of the cell, it lacks sufficient support to determine whether the abnormality corresponds to changes in the cell's internal state. Conversely, using only the changing trends of internal resistance more easily reflects state changes at the slow variable level, but may be insufficiently sensitive to some early abnormal responses. Therefore, combining the fluctuation components with the changing trends of internal resistance allows for the simultaneous consideration of both the dynamic and state characteristics of the abnormality. This ensures that the constructed electrochemical state characteristics reflect both the abnormal disturbance response and the direction of change in the cell's internal electrochemical state, thereby enhancing the robustness of anomaly identification and the completeness of the judgment criteria. Specifically, the trend of internal resistance change can include an upward trend of internal resistance, a phased sudden change trend, or an abnormal fluctuation trend. The electrochemical state characteristics constituted by these trends and fluctuation components can more comprehensively characterize the current actual state of the battery cell.

[0075] Subsequently, the electrochemical state characteristics are matched with electrochemical anomaly patterns in the anomaly pattern library. This involves comparing the current cell's electrochemical state characteristics with the feature templates corresponding to various electrochemical anomaly patterns in the pre-established anomaly pattern library to determine whether the current cell exhibits known anomaly pattern characteristics. Setting up the anomaly pattern library matching step is necessary because although the aforementioned steps have extracted electrochemical state characteristics with discriminative value from the operational data, these characteristics are still data-level representations. Without further establishing a correspondence with known anomaly patterns, it is difficult to clearly determine whether the current characteristic changes belong to the normal fluctuation range or have entered an abnormal evolution state. The anomaly pattern library can pre-store typical electrochemical state characteristic patterns corresponding to lithium plating, internal short circuits, precursors to thermal runaway, interface degradation, or other electrochemical anomalies. By comparing the current cell's electrochemical state characteristics with these anomaly patterns, abstract characteristic changes can be transformed into specific anomaly judgment criteria, thereby providing pattern support for determining the cell initiation of anomalies. The matching process can be implemented using similarity calculation, pattern recognition algorithms, or classification models, with the aim of determining which type of electrochemical anomaly pattern the current cell is more similar to.

[0076] Finally, cells that match electrochemical anomaly patterns and meet threshold conditions are identified as the initiating cells of anomalies. This means that after feature matching is completed, a threshold determination is applied to the matching results. When the electrochemical state characteristics of a cell reach a preset matching degree with electrochemical anomaly patterns in the anomaly pattern library, and the corresponding confidence score or similarity index exceeds a preset threshold, the cell is identified as the initiating cell of anomalies. This final determination step is necessary because the anomaly pattern matching results themselves may be affected by operating condition changes, boundary samples, or local disturbances. Without setting threshold conditions to constrain the matching results, some critical state cells or transient fluctuation cells may be mistakenly identified as anomaly source cells. By introducing threshold conditions, cells with truly high anomaly certainty can be distinguished from ordinary fluctuation cells, thereby improving the reliability of anomaly initiation cell determination. Once identified, this cell serves as the initial source cell of the anomaly event, providing a control object and analysis starting point for subsequent stop command issuance, proximity impact assessment, and the initiation of local enhanced heat dissipation measures.

[0077] Optionally, the steps of collecting the voltage, current, and internal resistance of each cell during the formation or capacity testing phase as operational data include:

[0078] Based on the charging and discharging stage of the battery cell and the local heat dissipation status, the gain and bandwidth are adjusted: the gain and bandwidth of the analog signal conditioning circuit are adjusted to adapt to different signal strengths and noise levels.

[0079] Perform measurement calibration: Inject calibration signals into the measurement loop and measure the response to evaluate and compensate for measurement deviations caused by electromagnetic interference in real time and achieve calibration;

[0080] After completing gain and bandwidth adjustment and measurement calibration, the voltage, current and internal resistance of each cell are collected to obtain raw data;

[0081] Preliminary digital filtering is performed on the raw data to remove high-frequency random noise, resulting in preprocessed data.

[0082] By combining the physical location information of the battery cells with the operating status of surrounding equipment, the preprocessed data is compensated and corrected, and the operating data is output.

[0083] Specifically, gain and bandwidth adjustment refers to dynamically adjusting the gain and bandwidth of the analog signal conditioning circuit based on the charging and discharging stage of the battery cell (e.g., constant current charging, constant voltage charging, constant current discharging, etc.) and its local heat dissipation status (e.g., temperature rise, cooling fan activation, etc.). The purpose is to ensure that the voltage, current, and internal resistance signals of the battery cell can be accurately amplified and captured under different operating conditions, while effectively suppressing noise of different frequencies, thereby adapting to different signal strengths and noise levels.

[0084] The measurement calibration process can be understood as injecting a known calibration signal into the measurement loop and measuring the system's response to that signal. By comparing the actual response with the expected response, measurement deviations caused by factors such as electromagnetic interference can be evaluated and compensated in real time, thereby achieving accurate calibration of the measurement data. Its purpose is to eliminate the influence of the external environment on measurement accuracy and ensure the authenticity of the data.

[0085] In practical applications, after gain and bandwidth adjustment and measurement calibration, the voltage, current, and internal resistance of each cell are collected to obtain raw data. This raw data consists of direct measurements after preliminary analog signal processing and calibration.

[0086] Furthermore, preliminary digital filtering is performed on the raw data to remove unavoidable high-frequency random noise during the acquisition process, such as instantaneous spike noise generated by the sensor itself or the circuit board, thereby obtaining smoother and more reliable preprocessed data.

[0087] Furthermore, by combining the physical location information of the battery cell with the operating status of surrounding equipment, the preprocessed data is compensated and corrected to further improve its accuracy. For example, the physical location of the battery cell may affect the intensity of electromagnetic interference it experiences, while the operating status of surrounding equipment (such as the charging and discharging of adjacent cells, fan speed, etc.) may also cause crosstalk or thermal effects on the measurement data of the target battery cell. Using this information, the preprocessed data can be finely compensated and corrected, ultimately outputting high-quality operational data.

[0088] In the formation and capacity testing line, due to the parallel operation of numerous charge and discharge channels, frequent switching of devices, and the dense arrangement of cells, the measurement link is susceptible to disturbances from stage switching, channel crosstalk, and changes in local electromagnetic fields. Therefore, this application introduces a gain-bandwidth adjustment mechanism during the operational data acquisition stage that matches the charging / discharging stage and local heat dissipation status. Furthermore, it performs real-time closed-loop calibration of measurement deviations by injecting calibration signals into the measurement loop. This is combined with information on the physical location of the cells and the operating status of surrounding equipment to compensate and correct the data. This approach ensures that the operational data not only has high sampling accuracy but also maintains good reliability under strong electromagnetic interference environments, providing a stable data foundation for subsequent abnormal cell initiation identification.

[0089] Optionally, the steps of performing adaptive filtering on the running data to obtain the processed data include:

[0090] The charging and discharging stages of the battery cell, local temperature, and the operating status of adjacent battery cells are obtained to form operating condition information and local environmental noise characteristics;

[0091] Based on operating condition information and local environmental noise characteristics, adjust the parameters of the adaptive filter;

[0092] Analyze the spectral characteristics of the battery cell signal to identify specific frequency components associated with electrochemical anomalies;

[0093] Based on the adjusted adaptive filter parameters, adaptive filtering is performed on the running data;

[0094] During the filtering process, specific frequency components are weighted and protected to retain anomalously correlated components while suppressing background noise;

[0095] Based on specific frequency components and retained abnormal correlation components, combined with the physical location information of the cells, crosstalk between adjacent cells is identified and canceled to obtain processed data.

[0096] Specifically, before performing adaptive filtering on the operating data, it is first necessary to obtain the charging and discharging phase of the battery cell, its local temperature, and the operating status of adjacent cells. This information is then used to form operating condition information and local environmental noise characteristics. The charging and discharging phase reflects the current operating mode of the battery cell, while the local temperature directly affects the electrochemical reaction rate and internal impedance of the cell. The operating status of adjacent cells may introduce crosstalk such as electromagnetic coupling or heat transfer. This operating condition information and local environmental noise characteristics are dynamically changing and are crucial for subsequent adjustment of filtering parameters.

[0097] Furthermore, based on the generated operating condition information and local environmental noise characteristics, the parameters of the adaptive filter are adjusted. For example, when the battery cell is in the high-current charging and discharging stage, it may generate significant electromagnetic noise; in this case, the filter's cutoff frequency or gain can be adjusted accordingly. When the local temperature is high, the internal noise of the battery cell may increase, and the noise suppression capability of the filter needs to be enhanced accordingly. This adaptive adjustment of parameters ensures that the filter is always in its optimal operating state to adapt to the constantly changing operating environment.

[0098] Simultaneously, it is necessary to analyze the spectral characteristics of the cell signals to identify specific frequency components associated with electrochemical anomalies. Electrochemical anomalies, such as lithium dendrite growth and SEI film rupture, often exhibit specific frequency fluctuation patterns in the cell's voltage and current signals. Through spectral analysis, these anomaly-related frequency components can be precisely located, providing guidance for subsequent filtering processes. Specific frequency components can be understood as frequency points or bands in the cell's operating data spectrum that are highly correlated with known or predicted electrochemical anomaly patterns.

[0099] Based on this, adaptive filtering is performed on the running data according to the adjusted adaptive filter parameters. This filtering process aims to effectively suppress background noise while preserving as much useful information in the signal as possible.

[0100] As a preferred implementation, specific frequency components are weighted and protected during the filtering process. This means that in the adaptive filtering algorithm, frequency components related to electrochemical anomalies are given higher weights, thereby suppressing background noise while ensuring that these key anomaly-related components are not mistakenly filtered out or excessively attenuated. This weighted protection mechanism is crucial for the early detection of cell anomalies.

[0101] Finally, based on specific frequency components and retained anomalous correlation components, and combined with the physical location information of the cells, crosstalk between adjacent cells is identified and canceled to obtain the processed data. Electromagnetic coupling or thermal transfer between adjacent cells can cause signal crosstalk, affecting the accuracy of individual cell data. By utilizing the physical location information of the cells, a crosstalk model can be established, and by combining the identified specific frequency components and retained anomalous correlation components, these crosstalks can be accurately estimated and canceled, thereby obtaining cleaner and more accurate processed data.

[0102] Optionally, the step of performing spectral analysis on the voltage and current in the processed data to extract the fluctuation components within a specific frequency range includes:

[0103] Obtain the charging and discharging stages, local temperature, and aging status of the battery cell;

[0104] Adjust the frequency range and resolution of the spectrum analysis according to the charging and discharging stage, local temperature and aging state;

[0105] Based on the adjusted resolution, specific frequency components related to electrochemical anomalies are identified and extracted within the adjusted frequency range; the fluctuation component is a set of spectral fluctuations extracted within the specific frequency range, and the specific frequency component is a subset of frequency points or frequency bands related to electrochemical anomalies within the fluctuation component.

[0106] Specifically, before performing spectrum analysis, it is necessary to obtain the cell's charge / discharge stage, local temperature, and aging state. The charge / discharge stage refers to the cell's current operating mode, such as charging, discharging, or resting. Different charge / discharge stages result in varying levels of electrochemical reaction activity and noise characteristics within the cell. Local temperature refers to the real-time temperature of the environment surrounding the cell. Temperature changes affect the cell's impedance characteristics and electrochemical reaction rate, thus altering the spectral performance of abnormal signals. Aging state refers to the degree of performance degradation of the cell during use; aged cells may exhibit different abnormal spectral characteristics compared to new cells.

[0107] Furthermore, the frequency range and resolution of the spectral analysis are dynamically adjusted based on the acquired charge / discharge stages, local temperatures, and aging conditions. For example, certain types of electrochemical anomalies may be more pronounced in specific high-frequency or low-frequency regions during certain charge / discharge stages or at specific temperatures. In such cases, the frequency range can be narrowed accordingly, and the resolution in that region can be increased to capture the anomalous signal more precisely. Conversely, under other operating conditions, if noise in certain frequency ranges is known to be significant and unrelated to the anomaly, the resolution in that range can be appropriately reduced or the noise excluded from the analysis.

[0108] Therefore, based on the adjusted resolution, specific frequency components related to electrochemical anomalies are identified and extracted within the adjusted frequency range. The fluctuation component can be understood as a set of spectral fluctuations extracted within a specific frequency range, containing signals with all amplitude and frequency variations within that range. The specific frequency component refers to a subset of frequency points or bands within the fluctuation component that are strongly correlated with electrochemical anomalies. These frequency points or bands are determined through pre-analysis or model training and can effectively characterize abnormal electrochemical activity within the battery cell.

[0109] Furthermore, this application does not apply a fixed frequency range and resolution to all operational stages of the spectrum analysis. Instead, it dynamically adjusts the frequency range and resolution of the spectrum analysis based on the cell's charging / discharging stage, local temperature, and aging state. This approach is necessary because anomalous disturbances manifest differently in the frequency domain at different process stages. Using a uniform spectrum analysis strategy might result in anomaly-related frequency bands being masked by noise or diluted by ineffective frequency bands. By adjusting the spectrum analysis parameters according to the stage, the spectral fluctuation components related to anomaly evolution can be extracted more accurately, thereby improving the ability to detect early anomalies.

[0110] Optionally, the steps for constructing electrochemical state characteristics include:

[0111] Obtain the charging and discharging current of the battery cell, local temperature, operating current of adjacent battery cells, and physical distance;

[0112] The internal resistance is corrected based on the charging and discharging current and local temperature to eliminate the influence of current and temperature on the internal resistance.

[0113] Based on the operating current and physical distance of adjacent cells, the electromagnetic coupling interference intensity is calculated, and the electromagnetic coupling interference intensity is subtracted from the corrected internal resistance to obtain the internal resistance after interference is deducted.

[0114] Perform time-series analysis on the internal resistance after removing disturbances to extract short-term rate of change and long-term drift trend;

[0115] By integrating short-term change rates, long-term drift trends, and specific frequency components, electrochemical state characteristics are constructed.

[0116] Specifically, to more accurately assess the electrochemical state of a battery cell, it is necessary to acquire key parameters affecting internal resistance measurements and electrochemical behavior in real time. Charge and discharge currents directly influence the ohmic and concentration polarization of the cell, while local temperature significantly affects the cell's reaction kinetics and ion transport rates. The operating currents of adjacent cells generate electromagnetic fields, causing electromagnetic coupling interference to the measurement signals of the target cell; the physical distance determines the strength of this coupling interference. These parameters can be acquired through a sensor network integrated into the formation and capacity testing system, such as current sensors, temperature sensors, and pre-defined cell layout information.

[0117] Internal resistance is a crucial indicator of a battery cell's health, but its measured value fluctuates significantly with changes in charge / discharge current and local temperature. To obtain a true internal resistance that reflects the cell's intrinsic condition, these influences need to be corrected. This correction process can be based on pre-established cell models or empirical lookup tables that describe the functional relationship between internal resistance and current and temperature. By substituting real-time measured charge / discharge current and local temperature into the model or lookup table, the deviation in internal resistance caused by these factors can be calculated and subtracted from the original measured value, thus obtaining an equivalent internal resistance value under standard conditions.

[0118] In practical applications, the cells in a formation-capacity system are typically closely packed. The currents generated by adjacent cells during charging and discharging create a changing magnetic field, which in turn induces an electromotive force or current in the measurement circuit of the target cell via electromagnetic induction, thus interfering with the measurement of internal resistance. The intensity of this electromagnetic coupling interference is directly proportional to the operating current of adjacent cells and inversely proportional to the physical distance between them. By establishing an electromagnetic coupling model and combining real-time acquired operating currents of adjacent cells with known physical distances, the intensity of electromagnetic coupling interference can be accurately calculated. Subsequently, this interference intensity is subtracted from the internal resistance after current and temperature correction to further refine the internal resistance data and make it more accurately reflect the electrochemical characteristics of the cell itself.

[0119] Internal resistance data, after multiple corrections and interference removal, can more accurately reflect the health status of the battery cell. To uncover early signs of anomalies from this data, time-series analysis is necessary. Various methods can be employed for time-series analysis, such as moving averages, exponential smoothing, and Kalman filtering, to extract the short-term rate of change and long-term drift trend of internal resistance. The short-term rate of change characterizes transient responses or sudden anomalies within a short period, while the long-term drift trend reflects gradual anomalies such as slow aging, capacity decay, or changes in internal structure. These time-series characteristics are crucial for identifying different types of electrochemical anomalies.

[0120] Therefore, by fusing the short-term rate of change and long-term drift trend extracted from the time-series analysis of internal resistance with the fluctuation components (i.e., specific frequency components) extracted within the aforementioned specific frequency range, a comprehensive and robust electrochemical state characteristic can be constructed. This fusion can be achieved through methods such as weighted averaging, eigenvector concatenation, and input to machine learning models. Specific frequency components can capture the dynamic characteristics and microstructural changes of the electrochemical reactions within the cell, while the changing trend of internal resistance reflects the macroscopic impedance characteristics and aging state. Combining the two allows for the characterization of the cell's electrochemical state from multiple dimensions and scales, thus forming a comprehensive characteristic that is more sensitive to and accurate in detecting anomalies.

[0121] In this application, the internal resistance is not directly determined using the original measured value. Instead, the internal resistance is first corrected based on the charging / discharging current and local temperature. Then, the electromagnetic coupling interference intensity is calculated by combining the operating current of adjacent cells and the physical distance. The electromagnetic coupling interference intensity is then subtracted from the corrected internal resistance to obtain internal resistance data that better reflects the true state of the cell itself. Based on this, time-series analysis is performed on the internal resistance data to extract the short-term rate of change and long-term drift trend, and then fused with specific frequency components. The necessity of setting up this processing chain lies in the fact that relying solely on the original internal resistance value in the formation and capacity testing process is easily affected by changes in operating conditions and disturbances from neighboring cells. The internal resistance trend after subtraction using current, temperature, and electromagnetic coupling is more conducive to identifying the intrinsic electrochemical state changes corresponding to the true source of the anomaly.

[0122] Optionally, the step of matching electrochemical state characteristics with electrochemical anomalous patterns in an anomalous pattern library includes:

[0123] Based on the charging and discharging stage, local temperature, and aging state of the battery cell, adjust the matching threshold and matching weight of each abnormal mode in the abnormal mode library.

[0124] Electrochemical state characteristics are compared in parallel with multiple electrochemical anomaly patterns, and a confidence score is calculated for each matching result;

[0125] When the matching results of multiple abnormal patterns fall into the overlapping area of ​​normal fluctuations and abnormal signals, local cross-validation is initiated. By analyzing the operating data and heat transfer trends of adjacent cells, the possibility of the current cell being abnormal is assessed, and the local cross-validation results are obtained.

[0126] Based on the adjusted matching threshold, matching weight, confidence score, and local cross-validation results, the system outputs a judgment result on whether the cell matches the electrochemical abnormal mode, and updates the judgment of the abnormal starting cell according to the judgment result.

[0127] Specifically, during electrochemical anomaly mode matching, the matching thresholds and weights of each preset anomaly mode in the anomaly mode library are dynamically adjusted based on real-time operating conditions such as the cell's charge / discharge stage, local temperature, and aging status. The charge / discharge stage refers to the cell's current charging, discharging, or resting state; local temperature refers to the real-time temperature of the environment surrounding the cell; and aging status refers to indicators such as the cell's cycle life and capacity decay. These parameters significantly affect the cell's electrochemical behavior and the manifestation of its anomaly characteristics. Therefore, dynamically adjusting the matching thresholds and weights makes the matching process more adaptive. The matching threshold is the critical value used to determine whether an electrochemical state characteristic matches an anomaly mode, while the matching weights characterize the importance of different characteristics in the matching process.

[0128] Furthermore, the constructed electrochemical state characteristics are compared in parallel with multiple electrochemical anomaly patterns in the anomaly pattern library. This means that the electrochemical state characteristics of a cell can be compared with multiple potential anomaly patterns simultaneously, rather than sequentially or individually. For each comparison result, a confidence score is calculated, which quantifies the similarity or matching degree between the current electrochemical state characteristics and a specific anomaly pattern. The higher the confidence score, the better the matching degree.

[0129] Furthermore, when the matching results of multiple abnormal patterns fall into the overlapping region of normal fluctuations and abnormal signals—that is, when the matching results are unclear and difficult to directly determine as normal or abnormal—local cross-validation is initiated. This local cross-validation analyzes the operating data and heat transfer trends of adjacent cells to assist in judging the probability of an anomaly of the current cell. For example, if the matching result of a cell is in an ambiguous region, but its adjacent cells also show similar or related abnormal signs and have a clear heat transfer trend, this can enhance the judgment of the current cell's anomaly. Thus, a local cross-validation result can be obtained for further decision support.

[0130] Finally, based on the adjusted matching threshold, matching weight, confidence score, and local cross-validation results, a comprehensive judgment is output regarding whether the cell matches the electrochemical anomaly pattern. This judgment result can be used to update the identification of cells initiating anomalies, ensuring more accurate and reliable identification of such cells.

[0131] When the matching results of multiple abnormal patterns fall into the overlapping area of ​​normal fluctuations and abnormal signals, this application does not directly rely on a single matching score to make the final judgment. Instead, it initiates local cross-validation. By analyzing the operating data of adjacent cells, local heat transfer trends, and electrochemical state characteristics of adjacent cells, it assists in determining whether the current cell is the true starting point of the anomaly. The necessity of setting up this local cross-validation mechanism lies in the fact that, in the scenario of high-density arrays with capacity-deployment formation, neighboring cells may exhibit characteristic signals similar to those of the anomalous source cell due to heat transfer or electromagnetic coupling. Without local cross-validation, the affected cells are easily misjudged as the anomalous source. Through local cross-validation, the initial source and subsequent affected objects can be further distinguished based on the initial judgment of the abnormal pattern.

[0132] Optionally, the steps of comparing electrochemical state characteristics with multiple electrochemical anomaly patterns in parallel and calculating a confidence score for each match include:

[0133] The feature weights of each abnormal mode are adjusted according to the charging and discharging stage, local temperature and aging state of the battery cell; the feature weights are coefficient parameters used to characterize the contribution of each feature to the matching of abnormal modes.

[0134] Based on the adjusted feature weights of each anomalous mode, the initial matching degree between the electrochemical state characteristics and each anomalous mode is calculated.

[0135] When the initial matching degree reaches the preset matching degree, the local heat transfer trend and the electrochemical state characteristics of adjacent cells are introduced for auxiliary judgment to obtain the auxiliary judgment result.

[0136] Based on the auxiliary judgment results, the initial matching degree is corrected to obtain the corrected matching degree;

[0137] Based on the corrected matching degree, a confidence score is generated for each matching result.

[0138] Specifically, when performing parallel comparisons of electrochemical state characteristics with electrochemical anomaly modes in the anomaly mode library and calculating confidence scores, the feature weights of each anomaly mode are first dynamically adjusted based on real-time operating conditions such as the cell's current charge / discharge stage, local temperature, and aging status. Feature weights can be understood as coefficients representing the contribution of different electrochemical state characteristics (e.g., voltage fluctuation components, current fluctuation components, internal resistance change rate, etc.) to the matching of a specific anomaly mode during the matching process. For example, in high-temperature environments, the feature weights related to thermal runaway may be increased; during the charging stage, the feature weights related to overcharging may be strengthened. This approach makes the matching process more adaptive and accurate.

[0139] Furthermore, based on the adjusted feature weights of each anomalous mode, the initial matching degree between the current electrochemical state characteristics and each anomalous mode in the anomalous mode library is calculated. The initial matching degree can be calculated using various algorithms, such as methods based on Euclidean distance, cosine similarity, or support vector machines, to quantify the similarity between the current cell state and known anomalous modes.

[0140] When the initial matching degree reaches a preset matching degree, to improve the accuracy and robustness of the judgment, local heat transfer trends and electrochemical state characteristics of adjacent cells are introduced for auxiliary judgment, thereby obtaining auxiliary judgment results. Local heat transfer trends may include the cell surface temperature change rate and temperature difference with adjacent cells, used to assess the risk of abnormal heat diffusion. Electrochemical state characteristics of adjacent cells may include data such as voltage, current, and internal resistance of adjacent cells, used to determine whether the anomaly is regional or diffuse. For example, if a cell has a high initial matching degree, and its adjacent cells also exhibit similar abnormal trends or significant local temperature increases, the auxiliary judgment result will further strengthen the possibility that this cell is the initiating cell of the anomaly.

[0141] Therefore, based on the auxiliary judgment results, the initial matching degree is corrected to obtain the corrected matching degree. The correction process can be a weighted average, logical judgment, or machine learning model-based correction, aiming to incorporate effective information from the auxiliary judgment results into the initial matching degree to eliminate potential misjudgments or improve the ability to identify true anomalies. For example, if the auxiliary judgment results strongly support the occurrence of an anomaly, the initial matching degree will be corrected upwards; conversely, if the auxiliary judgment results indicate a low probability of an anomaly, the initial matching degree may be corrected downwards.

[0142] Finally, based on the corrected matching degree, a confidence score is generated for each matching result. The confidence score is the result of normalizing or mapping the corrected matching degree using a specific function, and is used to intuitively represent the degree of matching and reliability between the current cell state and a specific anomaly pattern. This score will serve as one of the key criteria for determining the cell that initiated the anomaly.

[0143] Optionally, the initial matching degree is corrected based on the auxiliary judgment result. The steps to obtain the corrected matching degree include:

[0144] The charging and discharging stages, local temperature, and aging status of the battery cell are retrieved, and the corresponding abnormal mode type to be corrected for the battery cell is obtained.

[0145] Based on the type of abnormal mode to be corrected, the corresponding correction strategy is selected from the correction rule base; the correction strategy includes the weight configuration of heat transfer characteristics and the weight configuration of electrochemical signal correlation.

[0146] Adjust the weight configuration in the correction strategy according to the charging and discharging stage, local temperature and aging state;

[0147] Based on the adjusted correction strategy, the correction strength of the auxiliary judgment result on the initial matching degree is calculated;

[0148] The corrected strength is applied to the initial matching degree to obtain the corrected matching degree.

[0149] Specifically, in the process of correcting the initial matching degree, it is first necessary to access real-time operating condition information such as the cell's charging and discharging stages, local temperature, and aging status, and to obtain the abnormal mode type that the current cell is initially identified as needing correction. This information forms the basis for fine-tuning. The abnormal mode type to be corrected can refer to specific electrochemical anomalies indicated by the initial matching results, such as internal short circuits, overcharging, over-discharging, and electrolyte decomposition.

[0150] Furthermore, based on the acquired anomaly mode type to be corrected, the system selects the most suitable correction strategy from a pre-established correction rule base. The correction rule base can be a database containing various correction algorithms, parameter configurations, and weight settings, each strategy optimized for a specific anomaly mode. Specifically, the correction strategy includes weight configurations for heat transfer characteristics and electrochemical signal correlation. The weight configuration for heat transfer characteristics refers to the relative importance coefficient assigned to local heat transfer trend information (e.g., temperature rise rate, temperature gradient, etc.) during the correction process; the weight configuration for electrochemical signal correlation refers to the relative importance coefficient assigned to the electrochemical state characteristics of adjacent cells (e.g., synchronous changes in voltage, current, impedance, etc.). For example, for anomaly modes that may lead to drastic temperature rises, the weight configuration for heat transfer characteristics may be set higher.

[0151] Based on this, the weighting of the selected correction strategy is dynamically adjusted according to the current charging and discharging stage, local temperature, and aging state of the battery cell. For example, when the battery cell is in a high-rate discharge stage, its internal heat generation is already high, so the weight of heat transfer characteristics may be appropriately reduced to avoid misjudgment; while when the battery cell is at a high degree of aging, its sensitivity to external disturbances may be enhanced, and the weight of electrochemical signal correlation may be increased.

[0152] Subsequently, based on the dynamically adjusted correction strategy, the correction strength of the auxiliary judgment result on the initial matching degree is calculated. The correction strength is a quantitative value that characterizes the degree to which the auxiliary judgment result improves or reduces the initial matching degree. For example, if the auxiliary judgment result strongly supports the occurrence of an anomaly, the correction strength is positive and relatively large; conversely, if the auxiliary judgment result shows a low probability of an anomaly, the correction strength is negative.

[0153] Finally, the calculated correction strength is applied to the initial matching degree to obtain the corrected matching degree. This corrected matching degree will more accurately reflect the true abnormal state of the cell because it comprehensively considers the abnormal mode type, cell operating condition, and the dynamic influence of auxiliary judgment results.

[0154] Therefore, this application establishes an anomaly identification and control chain specifically for formation and capacity testing scenarios. First, high-reliability operational data is obtained through staged adaptive acquisition and real-time calibration under strong electromagnetic interference. Then, electrochemical state characteristics are constructed by jointly analyzing spectral fluctuation components and internal resistance change trends. Local cross-validation is then used to identify the cell initiating the anomaly. Subsequently, based on the proximity influence degree formed by the combined effects of thermal transfer and electromagnetic coupling, graded protection control is implemented for neighboring cells. Finally, while suppressing the spread of anomalies, continuous formation and capacity testing operation of unaffected cells is maintained. Through this chain, this application achieves not general battery monitoring, but rather anomaly source identification and production line-level localized response control specifically for formation and capacity testing processes.

[0155] This application also discloses a real-time data acquisition and monitoring system for a formation and filling system, used to perform real-time data acquisition and monitoring of the formation and filling system, combined with... Figure 3 As shown, the real-time data acquisition and monitoring system 1 for the formulation and compatibilization system includes:

[0156] The abnormal cell determination module 11 is used to collect the operating data of each cell in the formation stage or capacity testing stage, and determine the cell that is the source of the abnormality based on the operating data. The cell that is the source of the abnormality is the cell that first meets the determination criteria after matching the electrochemical state characteristics with the electrochemical abnormality mode and combining the local cross-validation results and confidence score within a preset time window. This cell is used to characterize the cell that is the source of the abnormal event.

[0157] The stop command issuing module 12 is used to issue a stop command to the corresponding control unit based on the abnormal starting cell, so as to stop the charging and discharging operation of the abnormal starting cell;

[0158] The proximity influence acquisition module 13 is used to acquire the proximity influence degree of the neighboring cells of the abnormal starting cell; the proximity influence degree is used to characterize the combined effect of the thermal transfer and electromagnetic coupling of the abnormal starting cell on the neighboring cells.

[0159] The neighboring cell stop module 14 is used to issue a stop command to the neighboring cells that meet the influence determination conditions according to the degree of proximity influence, so as to stop the charging and discharging operation of the neighboring cells.

[0160] The heat dissipation activation module 15 is used to activate enhanced heat dissipation measures for the local area where the abnormal starting cell is located;

[0161] The charge / discharge strategy adjustment module 16 is used to adjust the charge / discharge strategy of the unaffected cells according to the current state of the unaffected cells.

[0162] This application proposes a real-time data acquisition and monitoring system for a formation and capacity testing system, aiming to address the problems of untimely identification of abnormal cells, inaccurate location of abnormal sources, and insufficient assessment of the scope of abnormal impact in existing formation and capacity testing systems under high-density cell arrangement conditions. During the formation and capacity testing process, once an abnormality occurs in a single cell, its abnormal state will not only be reflected in its own operating parameters but may also affect neighboring cells through heat transfer and electromagnetic coupling, causing neighboring cells to also experience temperature rise, impedance fluctuations, or abnormal charging / discharging states within a short period. If the system makes a rough judgment based solely on local parameter changes of a single cell, it is easy to misidentify the affected cell as the abnormal source cell, or to adopt overly crude control methods such as shutting down the entire pallet due to the inability to clearly distinguish between the abnormal source and the affected area, thereby affecting production continuity and system utilization efficiency. Therefore, this application forms a hierarchical response system consisting of anomaly identification, anomaly isolation, impact assessment, local protection, and capacity compensation through the coordinated cooperation of the abnormal cell determination module, the stop command issuance module, the proximity impact acquisition module, the proximity cell stop module, the heat dissipation measure activation module, and the charge and discharge strategy adjustment module, so as to balance the timeliness of anomaly handling, the accuracy of control, and the maintenance of production efficiency.

[0163] The above embodiments have already described the abnormal initiation cells, proximity influence, electrochemical anomaly mode matching threshold, confidence score, and related technical terms, which will not be repeated here. The following focuses on the implementation methods of each module and the necessity of setting them up.

[0164] The abnormal cell determination module is used to collect the operating data of each cell during the formation or capacity testing stage, and to determine the cell that initiated the abnormality based on the operating data. Setting up this module is necessary because all subsequent control actions of the system are based on the identification result of the cell that initiated the abnormality. If the source cell that first became abnormal cannot be identified among multiple cells, the subsequent stop command issuance, proximity influence calculation, and activation of local heat dissipation measures will lack a basis, easily leading to inaccurate control targets or chaotic response sequences. This abnormal cell determination module can be implemented as a software processing unit integrated into a central monitoring server, which includes a data acquisition interface, a data preprocessing unit, an electrochemical abnormality pattern matching unit, and a confidence score calculation unit. Through this module, the collected operating data can be processed first, and then the processed data can be compared with the electrochemical abnormality patterns in the abnormality pattern library to determine the cell that initiated the abnormality. Alternatively, this module can also adopt a distributed structure, located in the capacity testing cabinet's lower-level computer or a dedicated hardware acceleration unit, such as using an FPGA or ASIC chip for real-time processing. The necessity of this setup lies in the fact that in a high-density battery cell layout environment, the amount of data is large and the sampling frequency is high. If all of it relies on centralized processing by the host computer, it may increase the response delay. However, high-speed identification at the front end can shorten the time for determining abnormal battery cells.

[0165] The stop command issuance module is used to issue stop commands to the corresponding control unit based on the abnormal starting cell, thereby stopping the charging and discharging operation of the abnormal starting cell. The necessity of this module lies in the fact that once an abnormal starting cell is identified, if its charging and discharging path is not promptly cut off, the abnormal cell may continue to heat up and its internal resistance may change during continuous charging and discharging, even inducing more serious abnormal expansion. Therefore, a dedicated stop command issuance module is needed to quickly convert the abnormal identification result into actual control actions. This module can be implemented as a communication interface and logic processing unit in the central control system, used to convert the output result of the abnormal cell determination module into standard control commands and send them to the corresponding charging and discharging control unit via industrial Ethernet, CAN bus, or other control communication methods. As one implementation, this module can also be connected to the relay array, contactor array, or solid-state switch array corresponding to each cell's charging and discharging circuit, achieving rapid physical isolation of the abnormal starting cell by controlling these actuators. This physical isolation method is necessary because software current limiting or parameter modification alone may not be sufficient to quickly cut off the abnormal energy exchange path in the early stages of an abnormality, while physical disconnection is more conducive to suppressing the spread of the abnormality.

[0166] The proximity impact acquisition module is used to obtain the proximity impact degree of the neighboring cells of the cell that initiated the anomaly. Setting up this module is necessary because stopping the charging and discharging of the cell that initiated the anomaly does not mean the risk has been eliminated. The heat accumulation, current disturbance, and local electromagnetic coupling formed by the cell before and after the anomaly may continue to affect neighboring cells for a short period. Without quantitative analysis of the degree of this impact, it is impossible to determine which neighboring cells have entered a high-risk state, and subsequent preventative control cannot be targeted. This proximity impact acquisition module can be implemented as a calculation module in a data analysis server. By running a heat conduction model and an electromagnetic coupling model, combined with parameters such as the real-time temperature, current change rate, physical distance, and arrangement relationship of the cell that initiated the anomaly, it calculates the temperature rise trend and induced disturbance intensity of the neighboring cells, thus obtaining the proximity impact degree. Alternatively, additional temperature and electromagnetic field sensors can be deployed in the cell array to directly measure the effects of heat transfer and electromagnetic coupling, and then the proximity impact degree can be formed based on the measurement results. The necessity of setting up this type of direct measurement method lies in the fact that, in some high-density arrangement scenarios, relying solely on model calculations may be limited by the accuracy of boundary conditions, while the actual measurement method is conducive to improving the reliability of the results obtained from the proximity influence degree.

[0167] The neighboring cell stop module is used to issue stop commands to neighboring cells that meet the impact judgment conditions based on the degree of proximity influence, thereby stopping the charging and discharging operations of neighboring cells. Setting up this module is necessary because neighboring cells do not necessarily all need to be stopped; selective control is required based on the degree of impact. Directly stopping all neighboring cells after the occurrence of an abnormal cell, while improving safety redundancy, would significantly reduce production efficiency. Conversely, failing to take any control measures for neighboring cells could allow already significantly affected cells to continue operating, increasing the risk of abnormality propagation. Therefore, a neighboring cell stop module is needed to achieve graded protection based on the degree of proximity influence. This module can be implemented as a control logic unit that works in conjunction with the stop command issuance module, internally storing the impact judgment condition threshold and graded stop strategies. When the proximity influence output by the proximity influence acquisition module exceeds a preset threshold, this module can issue a stop command to the charging and discharging control unit of the corresponding neighboring cell, or issue load reduction commands, current limiting commands, etc., based on the magnitude of the proximity influence. In one implementation, this module can also be implemented as a separate hardware control board to receive the proximity impact calculation results and directly control the charging and discharging state of neighboring cells. This setup facilitates rapid preventative protection actions in the event of localized system anomalies.

[0168] The heat dissipation activation module is used to initiate enhanced heat dissipation measures in the localized area where the abnormally originating battery cell is located. The necessity of this module lies in the fact that even after the abnormally originating battery cell has stopped charging and discharging, heat may still accumulate in its body and surrounding area. If enhanced heat dissipation is not implemented in this localized area, the localized thermal field created by the abnormally originating battery cell may continue to spread to neighboring areas, causing some neighboring battery cells to experience abnormal temperature rises even without directly participating in the anomaly. Therefore, after isolating the anomaly source, thermal management of its localized area is still required. This heat dissipation activation module can be implemented as a control unit linked to the capacity cabinet's heat dissipation system. Upon receiving the location information of the abnormally originating battery cell, it activates local fans, liquid cooling branches, spray devices, or directional air delivery devices to enhance heat dissipation in the abnormal area. Alternatively, this module can also correspond to an independent localized heat dissipation device, such as an independently controlled micro-fan array or a localized liquid cooling unit. The necessity of implementing localized enhanced heat dissipation rather than a unified overall enhanced heat dissipation is that it can suppress the spread of abnormal heat while reducing ineffective intervention in unaffected areas, thus balancing heat dissipation efficiency and system energy consumption control.

[0169] The charge / discharge strategy adjustment module is used to adjust the charge / discharge strategies of unaffected battery cells based on their current status. This module is necessary because after the abnormal starting cell and some adjacent cells are stopped from charging / discharging, the number of available cells in the system decreases. If the charge / discharge strategies of the remaining unaffected cells are not reconfigured, the original formation and capacity utilization plan may not be completed as expected, thus affecting production cycle time and capacity utilization. Therefore, a charge / discharge strategy adjustment module is needed to compensate for the unaffected cells. This module can be implemented as an optimization algorithm module in a central monitoring server. Internally, it includes a cell health status assessment model, a remaining capacity prediction model, and production planning scheduling logic. Based on the real-time status, remaining capacity, health level, and current production requirements of the unaffected cells, it dynamically adjusts their charge / discharge current, voltage, charge / discharge duration, or grouping task allocation method to compensate for the capacity loss caused by the abnormal cell shutdown as much as possible. As one implementation method, this module can also be deployed in a distributed manner in each charge / discharge control unit, with the central system issuing optimization targets and each control unit autonomously adjusting the corresponding cell's operating strategy. The necessity of this setting is to maintain the continuity of system operation after anomaly handling, and to avoid a complete conflict between anomaly handling and production efficiency.

[0170] The real-time data acquisition and monitoring system for the formation and capacity testing system proposed in this application forms a closed-loop processing system of "anomaly identification—anomaly source isolation—impact assessment—affected cell prevention and protection—local heat dissipation suppression—unaffected cell strategy compensation" through the sequential connection and functional coordination between the abnormal cell judgment module, stop command issuance module, proximity impact acquisition module, proximity cell stop module, heat dissipation measure activation module, and charge / discharge strategy adjustment module. This system is necessary because in high-density cell layout environments, abnormal events are usually not isolated occurrences or isolated problems, but involve anomaly source location, impact range identification, heat diffusion control, and the continuity of production capacity maintenance. Without the abnormal cell judgment module, the anomaly source cannot be clearly identified; without the stop command issuance module, the initiating cell of the anomaly cannot be isolated in time; without the proximity impact acquisition module and the proximity cell stop module, it is difficult to provide graded protection for affected cells; without the heat dissipation measure activation module, local thermal risks cannot be suppressed in time; and without the charge / discharge strategy adjustment module, the system cannot maintain production efficiency after an anomaly occurs. Therefore, through the coordinated setup of the above modules, this application can improve the accuracy of anomaly identification and control while minimizing the impact on the overall batching and filling process, thereby balancing safety and production continuity.

[0171] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for real-time acquisition and monitoring of data of a formation and dispensing system, characterized in that, include: Collect the operating data of each cell during the formation stage or capacity testing stage, and determine the cell that started the abnormality based on the operating data; Within a preset time window, electrochemical state characteristics are constructed based on the spectral fluctuation components and internal resistance change trends extracted from the running data. The electrochemical state characteristics are matched with electrochemical abnormal modes, and the cell that first meets the judgment criteria after combining local cross-validation results and confidence scores is determined as the abnormal initiating cell. The abnormal initiating cell is used as the initial source cell characterizing the abnormal event. Based on the abnormal starting cell, a stop command is sent to the corresponding control unit to stop the charging and discharging operation of the abnormal starting cell; Obtain the proximity influence degree of the neighboring cells of the abnormal starting cell; the proximity influence degree is used to characterize the combined effect of the thermal transfer and electromagnetic coupling of the abnormal starting cell on the neighboring cells. Based on the proximity influence degree, a stop command is issued to the neighboring cells that meet the influence determination conditions to stop the charging and discharging operation of the neighboring cells. Enhanced heat dissipation measures were initiated for the local area where the abnormal starting cell was located. Adjust the charging and discharging strategies for unaffected cells based on their current status. 2.The data real-time acquisition and monitoring method of the formation and dispensing system according to claim 1, characterized in that, The steps of collecting operational data from each cell during the formation or capacity testing phase, and determining the cell at the onset of abnormality based on the operational data, include: The voltage, current and internal resistance of each cell are collected during the formation or capacity testing stages as operating data. Adaptive filtering is performed on the running data to obtain the processed data; Spectral analysis is performed on the voltage and current in the processed data to extract the fluctuation components within a specific frequency range; Based on the fluctuation components and the changing trend of internal resistance in the processed data, electrochemical state characteristics are constructed. The electrochemical state characteristics are matched with electrochemical anomaly patterns in the anomaly pattern library; Cells that match the electrochemical anomaly pattern and meet the threshold conditions are identified as anomalous initiation cells.

3. The method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 2, characterized in that, The steps of collecting the voltage, current, and internal resistance of each cell during the formation or capacity testing stage as operational data include: Based on the charging and discharging stage of the battery cell and the local heat dissipation status, the gain and bandwidth are adjusted: the gain and bandwidth of the analog signal conditioning circuit are adjusted to adapt to different signal strengths and noise levels. Perform measurement calibration: Inject calibration signals into the measurement loop and measure the response to evaluate and compensate for measurement deviations caused by electromagnetic interference in real time and achieve calibration; After completing gain and bandwidth adjustment and measurement calibration, the voltage, current and internal resistance of each cell are collected to obtain raw data; Preliminary digital filtering is performed on the raw data to remove high-frequency random noise, resulting in preprocessed data. By combining the physical location information of the battery cell with the operating status of surrounding equipment, the preprocessed data is compensated and corrected, and the operating data is output.

4. The method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 2, characterized in that, The step of performing adaptive filtering on the running data to obtain the processed data includes: The charging and discharging stages of the battery cell, local temperature, and the operating status of adjacent battery cells are obtained to form operating condition information and local environmental noise characteristics; Based on the operating condition information and local environmental noise characteristics, adjust the adaptive filter parameters; Analyze the spectral characteristics of the battery cell signal to identify specific frequency components associated with electrochemical anomalies; Based on the adjusted adaptive filter parameters, adaptive filtering is performed on the running data; During the filtering process, the specific frequency components are weighted and protected to retain abnormally correlated components while suppressing background noise; Based on the specific frequency components and the retained abnormal correlation components, the crosstalk between adjacent cells is identified and canceled by combining the physical location information of the cells, resulting in processed data.

5. The method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 2, characterized in that, The step of performing spectral analysis on the voltage and current in the processed data to extract the fluctuation components within a specific frequency range includes: Obtain the charging and discharging stages, local temperature, and aging status of the battery cell; Adjust the frequency range and resolution of the spectrum analysis based on the charging / discharging stage, local temperature, and aging state. Based on the adjusted resolution, specific frequency components related to electrochemical anomalies are identified and extracted within the adjusted frequency range; the fluctuation components are a set of spectral fluctuations extracted within the specific frequency range, and the specific frequency components are a subset of frequency points or frequency bands related to electrochemical anomalies in the fluctuation components.

6. The method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 5, characterized in that, The steps for constructing electrochemical state characteristics include: Obtain the charging and discharging current of the battery cell, local temperature, operating current of adjacent battery cells, and physical distance; The internal resistance is corrected based on the charging and discharging current and local temperature to eliminate the influence of current and temperature on the internal resistance. Based on the operating current and physical distance of adjacent cells, the electromagnetic coupling interference intensity is calculated, and the electromagnetic coupling interference intensity is subtracted from the corrected internal resistance to obtain the internal resistance after interference deduction. Perform time-series analysis on the internal resistance after removing disturbances to extract short-term rate of change and long-term drift trend; The short-term rate of change, long-term drift trend, and specific frequency components are integrated to construct electrochemical state characteristics.

7. The method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 2, characterized in that, The step of matching the electrochemical state characteristics with electrochemical anomaly patterns in the anomaly pattern library includes: Based on the charging and discharging stage, local temperature, and aging state of the battery cell, adjust the matching threshold and matching weight of each abnormal mode in the abnormal mode library. Electrochemical state characteristics are compared in parallel with multiple electrochemical anomaly patterns, and a confidence score is calculated for each matching result; When the matching results of multiple abnormal patterns fall into the overlapping area of ​​normal fluctuations and abnormal signals, local cross-validation is initiated. By analyzing the operating data and heat transfer trends of adjacent cells, the possibility of the current cell being abnormal is assessed, and the local cross-validation results are obtained. Based on the adjusted matching threshold, matching weight, confidence score, and local cross-validation results, the system outputs a judgment result on whether the cell matches the electrochemical abnormal mode, and updates the determination of the abnormal starting cell according to the judgment result.

8. The method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 7, characterized in that, The step of comparing electrochemical state characteristics with multiple electrochemical anomaly patterns in parallel and calculating a confidence score for each matching result includes: The feature weights of each abnormal mode are adjusted according to the charging and discharging stage, local temperature and aging state of the battery cell; the feature weights are coefficient parameters used to characterize the degree of contribution of each feature to the matching of abnormal modes. Based on the adjusted feature weights of each anomalous mode, the initial matching degree between the electrochemical state characteristics and each anomalous mode is calculated. When the initial matching degree reaches the preset matching degree, the local heat transfer trend and the electrochemical state characteristics of adjacent cells are introduced for auxiliary judgment to obtain the auxiliary judgment result. Based on the auxiliary judgment result, the initial matching degree is corrected to obtain the corrected matching degree; Based on the corrected matching degree, a confidence score is generated for each matching result.

9. A method for real-time data acquisition and monitoring of a formulation and capacity control system according to claim 8, characterized in that, The step of correcting the initial matching degree based on the auxiliary judgment result to obtain the corrected matching degree includes: The charging and discharging stages, local temperature, and aging status of the battery cell are retrieved, and the corresponding abnormal mode type to be corrected for the battery cell is obtained. Based on the type of abnormal mode to be corrected, a corresponding correction strategy is selected from the correction rule base; the correction strategy includes the weight configuration of heat transfer characteristics and the weight configuration of electrochemical signal correlation. The weight configuration in the correction strategy is adjusted according to the charging and discharging stage, local temperature, and aging state. Based on the adjusted correction strategy, the correction strength of the auxiliary judgment result on the initial matching degree is calculated; The modified strength is applied to the initial matching degree to obtain the modified matching degree.

10. A real-time data acquisition and monitoring system for a formation and capacity system, used to perform real-time data acquisition and monitoring of the formation and capacity system, characterized in that, include: The abnormal cell determination module is used to collect the operating data of each cell during the formation stage or capacity testing stage, and determine the abnormal starting cell based on the operating data. Within a preset time window, electrochemical state characteristics are constructed based on the spectral fluctuation components and internal resistance change trends extracted from the running data. The electrochemical state characteristics are matched with electrochemical abnormal modes, and the cell that first meets the judgment criteria after combining local cross-validation results and confidence scores is determined as the abnormal initiating cell. The abnormal initiating cell is used as the initial source cell characterizing the abnormal event. The stop command issuing module is used to issue a stop command to the corresponding control unit based on the abnormal starting cell, so as to stop the charging and discharging operation of the abnormal starting cell; The proximity influence acquisition module is used to acquire the proximity influence degree of the neighboring cells of the abnormal starting cell; the proximity influence degree is used to characterize the combined degree of thermal transfer influence and electromagnetic coupling influence of the abnormal starting cell on the neighboring cells. The neighboring cell stop module is used to issue a stop command to the neighboring cells that meet the influence determination conditions based on the proximity influence degree, so as to stop the charging and discharging operation of the neighboring cells. A heat dissipation activation module is used to activate enhanced heat dissipation measures for the local area where the abnormal starting cell is located; The charge / discharge strategy adjustment module is used to adjust the charge / discharge strategy of unaffected cells based on their current state.

Images