Waste battery pack cascade utilization method, cell screening method and cascade utilization system

Abstract

This application relates to the field of battery technology, and particularly to a method for the cascade utilization of waste battery packs, a cell screening method, and a cascade utilization system. The method for the cascade utilization of waste battery packs includes: discharging waste battery packs to obtain discharged battery packs; disassembling the discharged battery packs to obtain multiple cells; detecting the cells using cell detection equipment to obtain detection data and operating parameters of cell detection events within a detection time period; determining a detection feature set based on the detection data and operating parameters; generating a target detection feature set corresponding to the detection feature set; generating cell screening results based on cell deviation parameters corresponding to the target detection feature set; grouping the cells that can be used in the cascade utilization from the cell screening results, installing insulating silicone sheets between the grouped cells, shaping the grouped and assembled cells, and then welding and assembling them to obtain a battery pack. This method significantly improves the accuracy of cell screening.

Description

Technical Field

[0001] This application belongs to the field of battery technology, and in particular relates to a method for the secondary utilization of waste battery packs, a method for screening battery cells, and a secondary utilization system. Background Technology

[0002] Secondary utilization is the core method for recycling spent power batteries. With the rapid development of new energy vehicles and energy storage industries, the amount of spent battery packs has increased significantly. Secondary utilization can not only improve the comprehensive utilization rate of battery resources and reduce dependence on mineral resource mining, but also effectively control the environmental costs of battery recycling and disposal, becoming a key link in the green development of the new energy industry. As the core step in the secondary utilization of spent battery packs, cell screening is a key process connecting battery pack dismantling and secondary battery pack reconstruction. Its screening accuracy directly determines the cycle life, safety, and performance stability of the secondary battery pack.

[0003] In existing technologies, the secondary utilization of waste battery packs involves collecting basic testing parameters of the cells and performing simple threshold judgments. However, in actual industrial sorting operations, the cell testing process is susceptible to interference from various external factors, such as fluctuations in power supply network voltage and abnormal CPU utilization caused by high-load operation of testing equipment. These factors can cause abnormal fluctuations in cell testing parameters that are not related to the cell's inherent performance. Existing screening methods cannot distinguish whether such fluctuations are caused by the cell's own aging or by external testing factors, which easily leads to misscreening. This may result in cells that meet the secondary utilization standards being classified as recycled, causing resource waste; or non-compliant cells being included in the secondary utilization system, posing safety hazards to subsequent secondary products and failing to meet the industrial demand for large-scale secondary utilization of waste batteries. Summary of the Invention

[0004] This application provides a method for the cascade utilization of waste battery packs, a cell screening method, and a cascade utilization system, which can solve the problem that misscreening is very likely to occur in the sorting operation in actual industrial sites, making it difficult to meet the industrial demand for large-scale cascade utilization of waste batteries.

[0005] In a first aspect, embodiments of this application provide a method for the tiered utilization of waste battery packs, including: Discharge the used battery pack to obtain a discharged battery pack; The discharge battery pack was disassembled to obtain multiple battery cells; The battery cell is tested using a battery cell testing device, and the detection data of the battery cell testing events and the operating parameters of the battery cell testing device are obtained within the testing time period. Based on the detection data and the operating parameters, a detection feature set is determined; wherein, the detection feature set includes detection statistical features corresponding to each detection time window in at least one detection time window, the detection statistical features indicate time-dependent and non-time-dependent covariates within the corresponding detection time window, and the detection time period includes at least one detection time window; Based on the detection feature set, a target detection feature set corresponding to the detection feature set is generated; wherein, the target detection feature set includes target detection features corresponding to each of the detection time windows; Based on the cell deviation parameters corresponding to the target detection feature set, a cell screening result is generated; wherein, the cell screening result is used to indicate whether the cell meets the cascade utilization standard within the detection time period; The cells that can be used in stages from the cell screening results are grouped together, insulating silicone sheets are installed between the grouped cells, and then the grouped and assembled cells are shaped, welded and assembled to obtain a battery pack.

[0006] The technical solutions described in this application embodiment have at least the following technical effects: The method for the cascade utilization of waste battery packs provided in this application involves sequentially discharging and disassembling the waste battery packs, acquiring cell testing data and equipment operating parameters throughout the entire testing process using cell testing equipment, constructing a detection feature set that includes time-dependent and non-time-dependent variables within the testing time window, and generating a corresponding target detection feature set. Cells are then screened based on the cell deviation parameters corresponding to the target detection feature set. Cells meeting the cascade utilization standards are then grouped together, and a standardized process involving the installation of insulating silicone sheets, shaping, and welding assembly between the grouped cells is used to create a cascade utilization battery pack. The screening process integrates multi-dimensional information from cell testing data and equipment operating status, combining the characteristics of dynamic changes in time-dependent variables and static stability of non-time-dependent variables. By conducting feature analysis, it is possible to distinguish whether abnormal fluctuations in cell testing parameters are caused by the degradation of the cell's own performance or by interference from the external testing environment and equipment operating status. This fundamentally reduces the problem of false screening caused by traditional single threshold judgment, significantly improves the accuracy of cell screening, effectively reduces the resource waste caused by the misclassification and recycling of qualified cells, and reduces the safety hazards caused by substandard cells flowing into the matching process. It ensures the performance consistency and reliability of cells used in the cascade utilization. Combined with standardized matching and assembly processes, it further improves the structural stability, charge and discharge efficiency, and cycle life of cascade utilization battery packs. Overall, it improves the resource utilization rate and core performance of waste battery packs in the cascade utilization process, and meets the development needs of the large-scale and standardized cascade utilization industry.

[0007] Secondly, embodiments of this application provide a method for screening battery cells in waste battery packs, including: The battery cell is tested using a battery cell testing device, and the detection data of the battery cell testing events and the operating parameters of the battery cell testing device are obtained within the testing time period. Based on the detection data and the operating parameters, a detection feature set is determined; wherein, the detection feature set includes detection statistical features corresponding to each detection time window in at least one detection time window, the detection statistical features indicate time-dependent and non-time-dependent covariates within the corresponding detection time window, and the detection time period includes at least one detection time window; Based on the detection feature set, a target detection feature set corresponding to the detection feature set is generated; wherein, the target detection feature set includes target detection features corresponding to each of the detection time windows; Based on the cell deviation parameters corresponding to the target detection feature set, a cell screening result is generated; wherein, the cell screening result is used to indicate whether the cell meets the cascade utilization standard within the detection time period.

[0008] Thirdly, embodiments of this application provide a waste battery pack cascade utilization system for implementing the method described in any one of the first aspects above.

[0009] It is understood that the beneficial effects of the second and third aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating the method for the tiered utilization of waste battery packs provided in the embodiments of this application; Figure 2 This is a schematic diagram of the implementation process of step S600 in the method for the cascade utilization of waste battery packs provided in the embodiments of this application. Detailed Implementation

[0012] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0013] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0014] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0015] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determination" or "if the described condition or event is detected" may be interpreted, depending on the context, as "once determination," "in response to determination," "once the described condition or event is detected," or "in response to the detection of the described condition or event."

[0016] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0017] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0018] In existing technologies, the cascade utilization process of waste battery packs typically involves simple discharge, disassembly, basic testing, and assembly. The cell screening stage only collects basic testing parameters such as cell voltage and internal resistance, and uses a single threshold comparison to determine whether a cell meets cascade utilization standards. It lacks comprehensive analysis of the entire cell testing process and the operating status of the testing equipment. In actual industrial sorting operations, the cell testing process is susceptible to interference from various external factors, such as fluctuations in power supply network voltage, abnormal CPU utilization and changes in sampling rate of testing channels caused by high-load operation of the testing equipment, and compatibility deviations between different types of cells and the sensitivity of the testing equipment. All of these can cause abnormal fluctuations in cell testing parameters that are not related to the cell's inherent performance. Existing screening methods lack the ability to mine the time-dimensional characteristics of test data and fail to distinguish between time-varying and static influencing factors during the testing process. They cannot effectively differentiate whether abnormal fluctuations in test parameters are caused by cell aging and performance degradation or by changes in the external testing environment and equipment operating status. This can easily lead to serious misscreening problems: on the one hand, cells that meet the performance standards for secondary use may be incorrectly classified into the material recycling category, resulting in a serious waste of battery resources and reducing the economic value of secondary use; on the other hand, cells that do not meet the actual performance standards may be included in the secondary use system. After being assembled into secondary battery packs, these packs are prone to problems such as poor cell consistency and low charge and discharge efficiency, and may even cause safety accidents such as thermal runaway and fire, posing significant safety hazards to the subsequent use of secondary battery packs.

[0019] To address the aforementioned issues, this application provides a method for the cascade utilization of waste battery packs, a cell screening method, and a cascade utilization system. In this method, after the waste battery pack is sequentially discharged and disassembled, cell testing data and equipment operating parameters throughout the entire testing process are obtained using cell testing equipment. A detection feature set containing time-dependent and non-time-dependent variables within the testing time window is constructed, and a corresponding target detection feature set is generated. Cell screening is performed based on the cell deviation parameters corresponding to the target detection feature set. Cells meeting the cascade utilization standards are then grouped together, and a cascade utilization battery pack is manufactured using standardized processes such as installing insulating silicone sheets between grouped cells, shaping, and welding assembly. The screening stage integrates multi-dimensional information from cell testing data and equipment operating status, combining the dynamic changes of time-dependent variables over time and the static stability of non-time-dependent variables to perform detection feature analysis. This analysis can distinguish whether abnormal fluctuations in cell testing parameters are caused by the cell's own performance degradation or by interference from the external testing environment and equipment operating status. It fundamentally reduces the misscreening problems caused by traditional single threshold judgment, significantly improves the accuracy of cell screening, effectively reduces the resource waste caused by misscrambling qualified cells for recycling, and reduces the safety hazards caused by substandard cells flowing into the matching process. It ensures the performance consistency and reliability of the cells used in the cascade utilization. Combined with standardized matching and assembly processes, it further improves the structural stability, charge and discharge efficiency, and cycle life of the cascade utilization battery pack. Overall, it improves the resource utilization rate and core performance of the cascade utilization of waste battery packs, and meets the development needs of the large-scale and standardized cascade utilization industry.

[0020] To better understand the method for the tiered utilization of waste battery packs provided in this application, the specific implementation process of the method for the tiered utilization of waste battery packs provided in this application will be described below by way of example.

[0021] Figure 1 This illustration shows a schematic flowchart of a method for the cascade utilization of waste battery packs provided in an embodiment of this application. The method for the cascade utilization of waste battery packs includes: S100 discharges the used battery pack to obtain a discharged battery pack.

[0022] It is understandable that the discharge process follows the core principles of safety priority, slow release, and complete discharge. The specific implementation depends on the type of used battery pack (e.g., ternary lithium battery pack, lithium iron phosphate battery pack), the remaining charge (SOC) status, and the battery pack health (SOH), flexibly selecting the appropriate discharge method. First, core parameters can be read through the battery pack's built-in BMS (Battery Management System). If the BMS is faulty, the total voltage of the battery pack and the voltage of individual cells can be measured using external testing equipment to determine the SOC. When the SOC is higher than 30%, the external load discharge method is preferred, using a resistive load (e.g., alloy resistor, cement resistor) that matches the battery pack's voltage level and rated capacity. The discharge current should be controlled at 0.1C-0.2C (C is the battery pack's rated capacity) to avoid a sudden rise in internal battery temperature and damage to the separator caused by high-current discharge. When the SOC is lower than 30%, a passive discharge method can be used, allowing the battery pack's internal equalization resistors to naturally release energy without additional external equipment. However, the surface temperature of the battery pack must be monitored in real time to ensure it does not exceed the safe threshold of 45℃. During the discharge process, key parameters such as the total voltage of the battery pack, the minimum voltage of individual cells, and the temperature of the battery pack casing can be continuously recorded through the data acquisition module. Dynamic stop conditions can be set: the discharge can be determined to be complete when the voltage of a single cell in the ternary lithium battery pack drops to 3.0V and the voltage of a single cell in the lithium iron phosphate battery pack drops to 2.5V, and the voltage recovery rate does not exceed 0.05V within 30 minutes.

[0023] The S200 disassembles the discharged battery pack to obtain multiple battery cells.

[0024] It is understandable that disassembly is a crucial step connecting discharge processing and cell testing and screening. Its goal is to disassemble the battery pack layer by layer into individual cells without damaging the internal structure, electrodes, or separators, and without affecting the electrochemical performance of the cells. Simultaneously, it aims to fully preserve the original performance state and traceability information of the cells, providing qualified basic units for subsequent accurate testing and grouping. A pre-treatment process is performed before disassembly: First, high-pressure airflow is used to clean dust, oil, electrolyte residue, and other impurities from the surface of the battery pack to prevent impurities from entering the cells or polluting the environment during disassembly. Then, visual inspection is conducted to check for abnormalities such as deformation, damage, bulging, and leakage in the battery pack casing. If leakage or severe damage to the casing is found, the battery pack must be placed in an anti-corrosion tray, and residual electrolyte must be cleaned with absorbent cotton before disassembly to prevent electrolyte corrosion of equipment or burns to operators. The disassembly process proceeds step-by-step according to the layered disassembly logic of the outer casing, module, and cell, with each step requiring specialized tools and operating procedures: The first step is to disassemble the battery pack outer casing. The appropriate disassembly tool is selected based on the casing's fixing method (a torque wrench is used for bolt-fixed types, and a laser cutter is used for weld-fixed types). During disassembly, the direction of force applied by the tool is controlled to minimize scratches to the internal battery modules or cells by casing fragments. The second step is to disassemble the internal modules. First, disconnect the connection lines between the module and the main positive and negative terminals of the battery pack, as well as the communication lines. Record the type, location, and wiring logic of the line interfaces. Then, remove the module from the outer casing by disassembling the fixing bracket. The module is usually wrapped with an insulating protective layer (such as PVC heat shrink tubing, insulation...). The first step involves carefully cutting along the seam with a utility knife to avoid damaging the internal battery cells. The third step is to disassemble the module down to the battery cells, removing the internal battery cell fasteners (such as cable ties, brackets, and thermal pads) and disconnecting the connecting pieces between the battery cells (for welded connections, use a laser spot welding separator or a low-temperature soldering iron for heating and separation; for bolted connections, disassemble directly). During the separation process, insulated tweezers can be used to hold the battery cells to avoid short circuits between the positive and negative terminals. After separation, each battery cell should be placed individually in a recessed area of ​​an anti-static tray. At the same time, traceability information such as the original battery pack number, module number, and location number of the battery cell should be marked with laser marking or labels to facilitate subsequent analysis and correlation between test results and the original battery pack information.

[0025] S300 uses a cell testing device to test cells and acquires the detection data of cell testing events and the operating parameters of the cell testing device within the testing period.

[0026] It is understood that the battery cell testing equipment includes a testing station module, a sensor acquisition module, a main control and computing module, a data storage module, a communication transmission module, and a power supply and voltage regulation module. The testing station module is equipped with adjustable elastic clamps adapted to different specifications of battery cells (cylindrical, square, and pouch). The clamps have built-in gold-plated probes that ensure stable contact with the battery cell tabs (contact resistance ≤1mΩ), and the probes have an elastic buffer structure (buffer stroke 0.5-1mm) to reduce damage to the battery cell tabs or shell due to excessive clamping pressure. At the same time, the bottom of the station integrates an anti-static tray and a temperature sensor to monitor the surface temperature of the battery cell in real time (detection range 20℃ to 85℃, accuracy ±0.5℃) to prevent abnormal temperature rise of the battery cell during the testing process. The sensor acquisition module, as the core unit for data acquisition, integrates multiple types of high-precision sensors: a voltage sensor (range 0-5V, accuracy ±0.01V) for acquiring cell open-circuit voltage and charge / discharge voltage curves; a current sensor (range 0-100A, accuracy ±0.1A) for capturing changes in charge / discharge current; and an internal resistance sensor (range 0-100mΩ, accuracy ±0.1mΩ) for rapidly measuring cell internal resistance using the AC impedance method. In addition, it is equipped with a sampling rate monitoring sensor and a power supply voltage monitoring sensor to capture operating parameters such as the sampling rate of the detection channel and the power supply status of the equipment. The main control and processing module can use a high-performance industrial-grade processor (such as an ARM Cortex-A9 quad-core processor with a main frequency of 1.8GHz), which can simultaneously process data from multiple sensors, perform preliminary processing such as data denoising, calibration, and format conversion, and simultaneously monitor the equipment's own hardware status (CPU utilization, memory usage, and communication link status). The data storage module is equipped with a local solid-state drive (capacity ≥ 1TB) and an SD card expansion interface, supporting real-time local storage of test data and operating parameters. The storage format adopts standardized JSON format and is associated with traceability information such as cell number, test timestamp, and test scheme number. The communication transmission module can integrate CAN bus, Ethernet, and 4G / 5G wireless communication interfaces, allowing flexible selection of data transmission methods according to application scenarios: high-speed wired transmission (transmission rate ≥ 100Mbps) with the backend server is achieved through Ethernet in industrial field scenarios; wireless transmission is achieved through 4G / 5G in mobile testing scenarios, enabling real-time data upload to the cloud or local management system; it also supports CAN bus communication between devices, facilitating clustered management of multi-station testing equipment. The power supply and voltage regulation module adopts a dual-power supply design (main power supply + backup battery). The main power supply supports wide voltage input (AC220V±10%) and has a built-in voltage regulation circuit and filter module to stabilize the input voltage to DC12V / 24V to cope with power grid fluctuations in industrial fields; the backup battery can maintain the operation of the core functions of the equipment when the main power supply is interrupted (battery life ≥ 30 minutes).

[0027] In one possible implementation, in step S300, the detection data includes one or more of the following: detection timestamp, detection event type, initial detection parameters of the battery cell, final detection parameters of the battery cell, and deviation value of the battery cell parameters; the operating parameters include one or more of the following: CPU utilization rate of the detection device, memory utilization rate of the device, sampling rate of the detection channel, battery cell detection sensitivity, power supply network status, and type of battery cell being detected. The power supply network status includes whether the network is in a stable power supply state or in a fluctuating power supply state.

[0028] It is understandable that the detection timestamp is accurate to the millisecond level (formatted as "YYYY-MM-DD HH:MM:SS.XXX"), used to mark the acquisition time of each detection data, providing a time series benchmark for subsequent time-dimensional trend analysis; the detection event type is used to distinguish the detection stage of the cell (such as resting, charging, discharging, internal resistance testing, etc.), so that data from different stages can be classified and analyzed; the initial detection parameters of the cell refer to the initial state data of the cell when it enters the detection process (such as initial open circuit voltage, initial internal resistance), and the final detection parameters of the cell refer to the state data at the end of the detection process (such as final open circuit voltage, discharge capacity). The comparison between the two can directly reflect the performance of the cell during the detection process; the cell parameter deviation value refers to the difference between the actual parameters of the cell and the preset standard parameters within the same detection stage (such as charging voltage deviation, discharge current deviation), used to quickly identify abnormal fluctuations during the detection process. The CPU utilization and memory utilization of the testing equipment reflect the equipment's load. If the CPU utilization consistently exceeds a certain threshold (e.g., 80%), it may cause delays in the response of testing commands, thereby affecting the accuracy of data acquisition. The sampling rate of the testing channel reflects the density of data acquisition, and its change can indicate whether there is a fault in the testing channel. The cell detection sensitivity is the response threshold of the testing equipment to changes in cell parameters (e.g., the response threshold for changes in internal resistance is 0.1mΩ), which is related to the cell type. The power supply network status is determined by the voltage monitoring module. A stable power supply status means that the voltage fluctuation amplitude is ≤ a preset threshold (e.g., ±5%), while a fluctuating power supply status means that the voltage fluctuation amplitude is > ±5%, which directly affects the power supply stability of the testing equipment. The cell type is used to distinguish the chemical system of the cell (e.g., ternary lithium is type 1, and lithium iron phosphate is type 2).

[0029] This setup, by simultaneously collecting cell testing data and equipment operating parameters, achieves dual data coverage of both cell performance and testing environment conditions. It addresses the technical pain point of existing technologies that rely solely on single-data collection of the cell, making it impossible to distinguish environmental interference. The multi-dimensional design of the testing data (timestamp, initial / final parameters, deviation values) comprehensively depicts the cell's performance state and changes, providing rich raw data for subsequent feature extraction. The targeted selection of operating parameters (CPU utilization, power supply status, etc.) accurately captures environmental interference factors during testing, providing a basis for distinguishing between cell performance degradation and external testing interference. The collaborative collection of both data ensures a correlation between cell performance and the testing environment, reducing the misjudgment of parameter fluctuations caused by high equipment load as cell performance anomalies due to isolated cell data analysis. This significantly improves data reliability and comprehensiveness, laying a solid foundation for the construction of subsequent testing feature sets and the accuracy of anomaly detection. It also adapts to the data analysis needs of large-scale testing scenarios, ensuring that the testing data for each cell truly reflects its actual performance.

[0030] S400, based on the detection data and operating parameters, determine the detection feature set; wherein, the detection feature set includes the detection statistical features corresponding to each detection time window in at least one detection time window, the detection statistical features indicate the time-dependent and non-time-dependent covariates within the corresponding detection time window, and the detection time period includes at least one detection time window.

[0031] It is understandable that the core technical goal of the detection feature set is to structure and dimensionally process the collected detection data and operating parameters, extract key features that reflect the fluctuations in cell performance and the influence of the detection environment, and provide a quantifiable and analyzable feature foundation for subsequent anomaly judgment. The logic of dividing the detection time window can balance analysis granularity and computational efficiency: a fixed duration window is usually adopted (e.g., 5 minutes / window, 10 minutes / window), and the window duration is dynamically adjusted according to the stage of the detection process (the window duration is shorter during the charging stage and longer during the resting stage). The detection statistical features corresponding to each detection time window are the aggregation and refinement of the original data within that window. Its core logic is to extract features by classifying them according to their attributes, that is, to distinguish between time-dependent and non-time-dependent covariates: time-dependent covariates refer to feature parameters that change dynamically with time within the detection time window, and their values ​​fluctuate regularly or randomly as the detection process progresses and the equipment status changes; non-time-dependent covariates refer to feature parameters that remain stable within the detection time window and do not change with time, and their values ​​are determined by the inherent attributes of the detection equipment, the basic information of the cell, and the initial detection settings, and remain constant throughout the window. The process of constructing the detection feature set is as follows: First, the detection data and running parameters within each detection time window are cleaned (missing values ​​and outliers are removed, and invalid data is filled in). Then, statistical indicators are calculated according to the type of covariate (such as calculating the mean and fluctuation value of time-dependent covariates, and directly extracting constant values ​​of non-time-dependent covariates). Finally, the detection statistical features corresponding to each window are formed, and the combination of the detection statistical features of all windows constitutes the complete detection feature set.

[0032] In one possible implementation, in step S400, the time-dependent covariates include one or more of the following: cell detection frequency, average fluctuation value of cell internal resistance, average CPU utilization rate of detection equipment, device memory growth rate, and change value of detection channel sampling rate; the non-time-dependent covariates include one or more of the following: cell detection sensitivity, cell type, and power supply network status.

[0033] It can be understood that the cell detection frequency refers to the number of times a cell is triggered for detection within a single detection time window (e.g., internal resistance detection times / minute), reflecting the intensity of the detection process. Its fluctuation can indicate the working status of the detection equipment. The average fluctuation value of the cell internal resistance refers to the ratio of the standard deviation to the mean of the internal resistance data from multiple tests within the window. The calculation formula is (internal resistance standard deviation / internal resistance mean) × 100%. It reflects the stability of the cell internal resistance and is a core indicator for assessing the health of the cell. The average CPU utilization rate of the detection equipment refers to the arithmetic mean of the CPU utilization rate within the window, reflecting the average load of the equipment within that window. The equipment memory growth rate refers to the increase in memory utilization rate within the window. The calculation formula is (end memory utilization rate - beginning memory utilization rate) / beginning memory utilization rate × 100%. The indicator shows the trend of memory usage changes; the change value of the detection channel sampling rate refers to the percentage difference between the actual sampling rate and the preset sampling rate within the window, calculated as (actual sampling rate - preset sampling rate) / preset sampling rate × 100%, used to determine whether the detection channel is working properly. Among the time-independent covariates, the cell detection sensitivity is an inherent parameter of the detection equipment, determined by the equipment hardware configuration (e.g., voltage detection sensitivity 0.001V), and has been calibrated and fixed before detection; the cell type is a basic attribute of the cell (e.g., ternary lithium, lithium iron phosphate), determined by manual input or automatic identification before detection begins; the power supply network status remains stable (stable or fluctuating) within a single detection time window. If a status switch occurs within the window, the window is re-divided, therefore it is a time-invariant parameter within the window.

[0034] This configuration, by dividing the detection time window, breaks down the continuous detection process into multiple independent analysis units. The operating conditions within each window are relatively stable, enabling precise capture of feature changes across different time periods and reducing the problem of local anomalies being masked by overall process analysis. Furthermore, the clear distinction between time-dependent and non-time-dependent covariates allows feature extraction to specifically characterize the impact of dynamic and static factors: time-dependent covariates focus on dynamic fluctuations during the detection process (such as fluctuations in cell internal resistance and equipment load changes), reflecting real-time changes in cell performance and dynamic interference from the detection environment; non-time-dependent covariates focus on stable factors during the detection process (such as cell type and detection sensitivity), providing a basic benchmark for feature analysis. This categorized extraction method makes the detection feature set structure clearer and more targeted. Subsequent analysis allows for differentiated judgment logic for the two types of covariates, significantly improving the accuracy of anomaly detection. Simultaneously, the structured feature set reduces the complexity of subsequent data processing, adapting to the efficient analysis needs of large-scale detection scenarios and providing core technical support for accurate cell screening.

[0035] S500 generates a target detection feature set corresponding to the detection feature set based on the detection feature set; wherein, the target detection feature set includes the target detection features corresponding to each detection time window.

[0036] It is understandable that the core technical objective of generating the target detection feature set is to perform redundancy removal, strong correlation, and optimal fusion processing on the initially constructed detection feature set, extract core features that can directly serve the cell screening and judgment, eliminate invalid and redundant feature information, reduce the computational complexity of subsequent anomaly judgment, and improve the correlation between features and screening criteria. This process follows the execution logic of data preprocessing-feature fusion-feature screening: First, data preprocessing is performed, and standardization is performed on time-dependent and non-time-dependent covariates in the detection feature set. Time-dependent covariates can be standardized using Z-score (converting the data into dimensionless data with a mean of 0 and a standard deviation of 1) to eliminate the dimensional differences of different parameters (such as the percentage of internal resistance fluctuation value and the percentage of CPU utilization rate can be directly compared); non-time-dependent covariates are normalized (mapping the data to the [0,1] interval) so that different types of stable parameters have a unified analytical scale. Subsequently, feature fusion is performed. For each detection time window, the time-dependent covariate features and non-time-dependent covariate features within that window are collaboratively fused. A weighted fusion algorithm is used, and the fusion coefficients are assigned according to the influence of features on the adaptability of battery cell tiered utilization (e.g., the weight of the average fluctuation value of battery cell internal resistance is 0.3, the weight of the detected battery cell type is 0.2, and the weight of the average CPU utilization rate of the detection equipment is 0.1). The fusion formula is target detection feature = Σ (single feature × corresponding weight), generating the initial fused features for each window. At the same time, feature interaction terms are introduced (e.g., average fluctuation value of battery cell internal resistance × power supply network state coefficient, with a stable power supply state coefficient of 1 and a fluctuation state coefficient of 1.2) to strengthen the correlation between key features and environmental factors. Finally, feature selection is performed. A combination of analysis of variance (ANOVA) and mutual information entropy can be used to calculate the correlation strength between each initial fused feature and whether the battery cell meets the tiered utilization standard. A correlation strength threshold is set (e.g., mutual information entropy ≥ 0.2). Features with correlation strength higher than the threshold are retained as the target detection features for that window, while redundant features with insufficient correlation strength (e.g., device memory growth rate features that are almost irrelevant to the selection results) are removed. Ultimately, the target detection features for all detection time windows are arranged in chronological order to form a complete target detection feature set. Each target detection feature is a core indicator that accurately reflects the combined impact of battery cell performance and the detection environment within that window.

[0037] S600 generates cell screening results based on the cell deviation parameters corresponding to the target detection feature set; the cell screening results are used to indicate whether the cells meet the cascade utilization standards within the detection period.

[0038] For example, the target detection features corresponding to each detection time window in the target detection feature set can be compared with the standard reference features for tiered utilization to obtain the cell deviation parameters corresponding to each detection time window. Then, based on multiple cell deviation parameters, cell screening results can be generated. Alternatively, the target detection features corresponding to each detection time window can be input into the detection evaluation module in the cell screening model to obtain the cell deviation parameters corresponding to the target detection features corresponding to each detection time window output by the detection evaluation module. Then, the cell deviation parameters corresponding to the target detection features corresponding to each detection time window can be input into the cell analysis module in the cell screening model to obtain the cell screening results output by the cell analysis module, and so on, but not limited to these.

[0039] In one possible implementation, please refer to Figure 2 In step S600, based on the cell deviation parameters corresponding to the target detection feature set, a cell screening result is generated, including: S610, compare the target detection features corresponding to each detection time window in the target detection feature set with the standard reference features for secondary utilization, and obtain the cell deviation parameters corresponding to each detection time window; wherein, the cell deviation parameters are used to indicate the degree of difference between the target detection features and the standard reference features for secondary utilization within the corresponding detection time window.

[0040] It is understandable that determining the reference characteristics for cascade utilization standards is a prerequisite for comparison. These characteristics are formulated based on relevant national standards (such as GB / T 34014-2023 "Requirements for Cascade Utilization of Automotive Power Batteries"), industry technical specifications, and specific application scenario requirements (such as backup power for communication base stations and power for low-speed vehicles). Different reference characteristic values ​​are set for different types of cells (ternary lithium and lithium iron phosphate) and different cascade application scenarios. For example, for lithium iron phosphate cells used as backup power for communication base stations, the reference characteristics for cascade utilization standards include an average internal resistance fluctuation value reference threshold ≤5%, an average CPU utilization rate reference threshold ≤60% for the testing equipment, and a stable power supply network status reference standard. For ternary lithium cells used for powering low-speed vehicles, the reference thresholds are: average internal resistance fluctuation value ≤3% and average CPU utilization rate ≤70%. The comparison process can employ relative deviation calculation. The formula for calculating the cell deviation parameter is: Cell Deviation Parameter = |(Target Detection Feature Value - Standard Reference Feature Value) / Standard Reference Feature Value × 100%|. This formula converts absolute differences into relative percentage differences, facilitating the comparison of deviation levels for different types of features (e.g., internal resistance fluctuation deviation and CPU utilization deviation can be uniformly measured as percentages). For example, if the average internal resistance fluctuation value of the target detection feature within a certain detection time window is 6%, and the standard reference feature value is 5%, then the corresponding cell deviation parameter for that window is |(6%-5%) / 5%×100%| = 20%. If the average CPU utilization rate of the detection equipment within that window is 65%, and the standard reference feature value is 60%, then the corresponding cell deviation parameter is |(65%-60%) / 60%×100%| ≈ 8.33%. Each target detection feature within a detection time window corresponds to a cell deviation parameter. This parameter quantifies the degree of difference between the overall performance of the cell and the tiered utilization standard within that window. The larger the deviation parameter, the more serious the deviation of the cell performance from the standard within that window.

[0041] The S620 generates cell screening results based on multiple cell deviation parameters.

[0042] For example, multiple cell deviation parameters can be classified to obtain time-varying deviation parameter groups corresponding to time-dependent covariates and non-time-varying deviation parameter groups corresponding to non-time-dependent covariates within the detection time window. Time-varying feature anomalies can be determined based on the time-varying deviation parameter groups, and non-time-varying feature anomalies can be determined based on the non-time-varying deviation parameter groups. Finally, cell screening results are generated based on the judgment results. Alternatively, the weights of each cell deviation parameter can be divided according to different usage scenarios, and a comprehensive deviation index can be calculated based on weighted summation. Cell screening results can be generated based on the comprehensive deviation index, and so on, but not limited to these methods.

[0043] This setup, by comparing the target detection features of each detection time window with the reference features of the tiered utilization standard one by one, achieves precise quantification of cell performance differences. The cell deviation parameters, in the form of objective data, clearly define the degree of deviation between the cell and the standard at different detection stages, reducing the ambiguity of traditional subjective judgment or single parameter comparison. At the same time, by comprehensively screening based on the deviation parameters of multiple detection time windows, rather than relying on data from a single window, the influence of accidental fluctuations during the detection process (such as instantaneous power supply interference and single detection errors) can be effectively eliminated, comprehensively capturing the overall trend and stability of cell performance, accurately distinguishing between cell performance degradation and parameter anomalies caused by external detection environment interference, and significantly improving the accuracy and reliability of cell screening results.

[0044] In one possible implementation, in step S620, a cell screening result is generated based on multiple cell deviation parameters, including: S621 classifies based on multiple cell deviation parameters to obtain time-varying deviation parameter groups corresponding to time-dependent covariates and non-time-varying deviation parameter groups corresponding to non-time-dependent covariates within the detection time window.

[0045] It is understandable that the core technical objective of this step is to perform reverse classification based on the covariate type corresponding to the deviation parameters, decomposing the comprehensive cell deviation parameters into two categories: time-varying deviations and non-time-varying deviations. This provides a classification basis for subsequent differential anomaly determination, and its classification logic is consistent with the definitions of time-dependent and non-time-dependent covariates in S400. A mapping relationship between deviation parameters and covariate types can be established: each cell deviation parameter is calculated from the corresponding target detection feature, and the target detection feature is the fusion result of time-dependent and non-time-dependent covariates. Therefore, by tracing the composition of the target detection feature, the contribution ratio of time-dependent and non-time-dependent covariates in each deviation parameter can be clarified. A contribution ratio threshold (e.g., 50%) is set. If the contribution ratio of the time-dependent covariate is ≥50%, the deviation parameter is classified into the time-varying deviation parameter group; if the contribution ratio of the non-time-dependent covariate is ≥50%, it is classified into the non-time-varying deviation parameter group. For example, if the average fluctuation of the cell internal resistance (time-dependent covariate) contributes 60% to a certain deviation parameter and the power supply network status (non-time-dependent covariate) contributes 40%, then this deviation parameter is classified into the time-varying deviation parameter group; if the cell type detected (non-time-dependent covariate) contributes 70% to a certain deviation parameter and the change in the sampling rate of the detection channel (time-dependent covariate) contributes 30%, then it is classified into the non-time-varying deviation parameter group.

[0046] S622, Time-varying feature anomaly determination is performed based on time-varying deviation parameter set, and non-time-varying feature anomaly determination is performed based on non-time-varying deviation parameter set.

[0047] It is understandable that the core technical objective of this step is to improve the accuracy of anomaly detection by adopting differentiated anomaly detection logic based on the different characteristics of time-varying and non-time-varying deviation parameter groups: time-varying deviation parameter groups reflect deviations caused by dynamic fluctuations, and it is necessary to focus on analyzing their changing trends in the time dimension; non-time-varying deviation parameter groups reflect deviations caused by static factors, and it is necessary to focus on static threshold verification and core parameter priority differentiation.

[0048] In one possible implementation, step S622 involves determining time-varying feature anomalies based on the time-varying deviation parameter set, including: S6221a performs time-dimensional trend analysis on the time-varying deviation parameter group, and calculates the dynamic change rate and cumulative deviation value of each parameter.

[0049] It is understandable that the core characteristic of time-varying deviation parameter sets is their dynamic change over time. Therefore, time-dimensional trend analysis is crucial for anomaly detection, and the dynamic change rate and cumulative deviation value are two core quantitative indicators characterizing this trend. The dynamic change rate reflects the speed of change of the deviation parameter. The calculation formula is: Dynamic Change Rate = (Current Window Deviation Parameter - Previous Window Deviation Parameter) / Time Interval × 100%, where the time interval is the duration of two adjacent detection time windows (e.g., 10 minutes). This indicator can quantify the rate of increase or decrease of the deviation parameter. If the dynamic change rate is consistently positive and large, it indicates that the deviation is rapidly expanding, and the cell performance may be deteriorating at an accelerated rate. For example, if the deviation parameter in the third window is 15%, the second window is 10%, and the time interval is 10 minutes, then the dynamic change rate = (15% - 10%) / 10 × 100% = 0.5% / minute, indicating that the deviation is increasing at a rate of 0.5% per minute. The cumulative deviation value reflects the cumulative degree of deviation parameters over the entire detection period. The calculation formula is: Cumulative Deviation Value = Σ(Deviation Parameter per Window - Deviation Threshold) × Window Weight. The deviation threshold is a preset upper limit for normal deviation (e.g., 20%), and the window weight is set according to the importance of the detection stage (e.g., 1.2 for the discharge stage and 0.8 for the resting stage). If the deviation parameter of a certain window is less than the deviation threshold, the cumulative value for that window is 0. The cumulative deviation value only adds the deviation exceeding the threshold, reflecting the total degree to which the cell performance deviates from the normal range. For example, if the detection period contains 5 windows, the deviation threshold is 20%, and the window weight is 1.0 for each window, with deviation parameters of 18%, 22%, 25%, 23%, and 19% respectively, then the cumulative deviation value = (22% - 20%) + (25% - 20%) + (23% - 20%) = 2% + 5% + 3% = 10%. During the calculation, the dynamic change rate needs to be calculated window by window in chronological order, and the cumulative deviation value is accumulated in real time.

[0050] S6222a, if the dynamic rate of change of any parameter is greater than the preset rate of change threshold, or the cumulative deviation of any parameter is greater than the preset cumulative threshold, it is marked as a time-varying feature anomaly.

[0051] It is understandable that the preset rate of change threshold and the preset cumulative threshold are pre-set values. They can be manually input, obtained from the cell database, etc., but are not limited to these methods. The cell database refers to a database containing the preset rate of change threshold and the preset cumulative threshold. This data can be obtained through laboratory experiments, on-site measurements and monitoring, and past experience. After acquisition, the collected data is organized, classified, and archived, useful information and patterns are extracted, and the relevant data is then saved into the database to form the cell database.

[0052] This setup employs a dual-indicator judgment logic of dynamic change rate and cumulative deviation value to address the dynamic characteristics of time-varying deviation parameter groups. The dynamic change rate accurately captures the rapid deterioration trend of cell performance, reducing the potential risk of deviations not exceeding thresholds but deteriorating too rapidly (e.g., a deviation parameter of 18% in a certain window does not exceed the 20% threshold, but the dynamic change rate is 0.6% / minute, far exceeding the 0.3% / minute threshold, indicating that the cell may rapidly exceed the threshold in subsequent windows). The cumulative deviation value reflects the continuous accumulation effect of deviations, reducing the missed detection of multiple slight exceedances of the threshold in a single window (e.g., multiple window deviation parameters are 21%-23%, all not exceeding the maximum threshold of 25%, but the cumulative deviation value reaches 30%, indicating insufficient cell performance stability). The synergistic judgment of the dual indicators improves the sensitivity of identifying rapid deterioration anomalies and achieves comprehensive coverage of continuously accumulating anomalies, significantly improving the accuracy of time-varying feature anomaly judgment and providing a reliable dynamic anomaly basis for the generation of subsequent screening results.

[0053] In one possible implementation, step S622 involves determining time-invariant characteristic anomalies based on the time-invariant deviation parameter set, including: S6221b performs static threshold verification on the time-invariant deviation parameter group; wherein, the time-invariant deviation parameter group includes core time-invariant deviation parameters and non-core time-invariant deviation parameters.

[0054] It is understandable that the core characteristic of time-invariant deviation parameter groups is static stability. Their deviations mainly stem from the inherent properties of the testing equipment and the compatibility issues with the basic information of the battery cells. Therefore, static threshold verification is the core judgment method, and the distinction between core and non-core parameters reflects the priority judgment logic of key factors. Core time-invariant deviation parameters refer to parameters that play a decisive role in the adaptability of battery cells for secondary use. Their deviations directly affect the safety and core performance of the battery cells. For example, they affect the detection of battery cell type (if the battery cell type does not match the secondary application scenario, such as using a ternary lithium battery cell in a base station backup power scenario requiring lithium iron phosphate batteries, it cannot be used even if other parameters are normal), and the battery cell detection sensitivity (if the sensitivity deviation is too large, it will cause all detection data to be distorted, and subsequent analysis will be meaningless). Non-core time-invariant deviation parameters refer to parameters that have a smaller impact on the adaptability of battery cells for secondary use. Their slight deviations can be compensated for by subsequent grouping or process adjustments. For example, the power supply network status (if it is fluctuating during testing, but the deviation is small, and the core performance of the battery cell is normal, the impact can be avoided by optimizing the power supply adaptation design of the secondary products). Before static threshold verification, differentiated thresholds need to be set for core parameters and non-core parameters: the thresholds for core time-varying deviation parameters are set more strictly (e.g., the threshold for cell type compatibility deviation is 0%, meaning a perfect match is required; the threshold for cell detection sensitivity deviation is 5%), while the thresholds for non-core parameters can be appropriately relaxed (e.g., the threshold for deviation caused by power supply network fluctuations is 15%). During the verification process, the core and non-core parameters in the time-varying deviation parameter group are first extracted, and then the deviation value of each parameter is compared with the corresponding preset threshold. The names of parameters exceeding the threshold and the degree of deviation are recorded.

[0055] S6222b: If at least one parameter value in the core time-invariant deviation parameters is greater than the preset core single parameter threshold, it is marked as a time-invariant feature anomaly; if only the non-core time-invariant deviation parameters have parameter values ​​greater than the preset non-core single parameter threshold, and the proportion of parameters exceeding the threshold is greater than the preset proportion threshold, it is marked as a time-invariant feature anomaly.

[0056] It is understandable that deviations in core parameters can prevent battery cells from meeting the basic requirements for secondary use. For example, if the cell type is mismatched with the application scenario, even if other parameters are qualified, it cannot be used to avoid safety hazards or substandard performance after subsequent pairing. For non-core, non-time-varying deviation parameters, a comprehensive judgment method based on the percentage of parameters exceeding the threshold is adopted. First, the ratio of the number of parameters exceeding the threshold to the total number of non-core parameters (the percentage exceeding the threshold) is calculated. Then, this percentage is compared with a preset percentage threshold (e.g., 30%). If the percentage is greater than the threshold, it indicates that the deviation of non-core parameters is relatively common and may indirectly affect the overall performance of the battery cell, and it is marked as abnormal. If the percentage is less than or equal to the threshold, it indicates that only a few non-core parameters have slight deviations, and the impact on the secondary use of the battery cell is small, so it is not marked as abnormal. For example, if there are 5 non-core, non-time-varying deviation parameters, and 2 of them exceed the threshold, with a percentage exceeding the threshold of 40%, if the preset percentage threshold is 30%, it is marked as a non-time-varying characteristic abnormality; if only 1 parameter exceeds the threshold, with a percentage of 20%, it is not marked as abnormal.

[0057] This setup ensures the basic safety and performance baseline for secondary utilization through a veto system for core parameters, avoiding risks to subsequent products due to deviations in core parameters. The method of determining the proportion of non-core parameters reflects the flexibility of the selection process, reducing the possibility of qualified cells being misjudged due to slight deviations in individual minor parameters, thus improving resource utilization. This differentiated judgment logic not only meets the actual needs of secondary utilization (core performance prioritized, minor performance can be compromised) but also balances the dual goals of safety and resource conservation, significantly improving the accuracy and rationality of non-time-varying characteristic anomaly detection.

[0058] S623 generates cell screening results based on time-varying characteristic anomaly determination results and non-time-varying characteristic anomaly determination results.

[0059] It is understandable that the core technical objective of this step is to integrate the judgment results of both time-varying and non-time-varying anomalies to form the final comprehensive screening result. Its judgment logic reflects the principle of risk priority and classification decision-making, that is, to distinguish different risk levels based on the combination of anomaly types, and then make corresponding screening decisions.

[0060] This setup categorizes multiple cell deviation parameters based on the essential attributes of time-dependent and non-time-dependent covariates, forming time-varying and non-time-varying deviation parameter groups and conducting differentiated anomaly detection. Finally, the results from both types of detection are merged to generate cell screening results. This approach accurately matches the characteristic patterns of different types of deviation parameters, solving the problems of insufficient targeting and masking of abnormal features caused by mixed analysis of various deviation parameters. Specifically, the dynamic characteristic anomaly detection of the time-varying deviation parameter group accurately captures the dynamic degradation trend of cell performance during the testing process, while the static characteristic anomaly detection of the non-time-varying deviation parameter group effectively verifies the adaptability of the cell's basic attributes and testing conditions. The two types of detection have different focuses and complementary coverage, achieving comprehensive analysis of different... Comprehensive identification of abnormal risks; and by integrating the results of two types of judgments for comprehensive screening, the one-sidedness of single-dimensional judgment is reduced. It can accurately distinguish parameter anomalies caused by different reasons such as the degradation of the cell's own performance, dynamic interference in the testing environment, and incompatibility of the basic testing conditions. This improves the accuracy of cell screening from the root, effectively reducing the problems of misscreening and missed screening. It not only reduces the waste of resources caused by the misjudgment of qualified cells, but also reduces the safety hazards of substandard cells entering the cascade utilization stage. At the same time, the structured judgment logic is adapted to the data analysis needs of large-scale cell testing, providing core technical support for the standardized and efficient implementation of the cascade utilization of waste batteries, and further ensuring the performance consistency and reliability of subsequent cascade battery packs.

[0061] In one possible implementation, in step S623, based on the time-varying characteristic anomaly determination results and the non-time-varying characteristic anomaly determination results, a cell screening result is generated, including: If a cell is simultaneously marked as having time-varying characteristic anomalies and non-time-varying characteristic anomalies, or if it is marked as having time-varying characteristic anomalies and the dynamic change rate of any parameter exceeds the threshold range, the cell screening result indicates that the cell does not meet the cascade utilization standard; if it is only non-time-varying characteristic anomalies and the parameter exceeding the threshold is a non-core, non-time-varying deviation parameter, or if there are no anomaly markers, the cell screening result indicates that the cell meets the cascade utilization standard.

[0062] It is understandable that scenarios marked as having both types of anomalies indicate a dual risk for the battery cell, with issues in both dynamic performance and static foundation. After being put into secondary use, it may not only fail rapidly but also cause safety accidents, thus it is directly deemed non-compliant. Scenarios with only time-varying characteristic anomalies and a dynamic change rate exceeding the threshold range indicate that the battery cell performance deteriorates too quickly. Even if the current deviation has not reached an extreme value, it may quickly exceed the safe range during subsequent use. For example, a dynamic change rate exceeding the threshold by 20% indicates that core parameters such as the cell's internal resistance are fluctuating rapidly, thus it is deemed non-compliant. Scenarios with only non-time-varying characteristic anomalies and the parameters exceeding the threshold are non-core parameters indicate that minor issues with the battery cell's basic compatibility do not affect core performance and can be compensated for through subsequent process adjustments. For example, deviations caused by power supply network fluctuations can be addressed by adding a voltage regulator module in the secondary battery pack design, thus it is deemed compliant. Scenarios with no anomaly markings indicate that the battery cell's dynamic performance is stable and its static foundation is compatible, fully meeting the secondary use standards, thus it is deemed compliant.

[0063] It should be noted that there are also scenarios where the time-varying characteristic anomaly is only marked, but the dynamic change rate of any parameter does not exceed the preset change rate threshold. In this scenario, the time-varying characteristic anomaly is not caused by accelerated degradation of the cell performance, but rather by the anomaly marker triggered by the cumulative deviation value of the time-varying deviation parameter exceeding the preset cumulative threshold. This means that although the cell's performance parameters deviated slightly from the tiered utilization standard multiple times during the entire testing process, forming a cumulative effect of deviation, it reflects that the overall performance stability of the cell is slightly poor. However, the rate of change of the core performance parameters is always within a controllable range, without a sudden or accelerated degradation trend, which is fundamentally different from the rapid deterioration characteristic of the dynamic change rate exceeding the threshold. For example, the average fluctuation value of the cell's internal resistance is controlled within the preset threshold of 0.2% / minute in each testing time window, but the cumulative deviation value of the entire testing cycle exceeds the preset cumulative threshold of 30%. This indicates that although the cell's internal resistance deviated slightly from the standard multiple times during the testing process, the overall fluctuation rate was slow, without sudden or accelerated fluctuations. The core performance of the cell did not show rapid degradation characteristics, but only insufficient long-term stability. For this scenario, a differentiated assessment can be made manually based on the target tiered application scenario of the battery cell: If the battery cell is intended for core tiered application scenarios with high requirements for battery cell performance stability, cycle life, and operational reliability, such as grid energy storage, communication base station backup power, and new energy low-speed vehicle power, these scenarios have stringent requirements for the long-term operational performance of the battery cell. Even if the battery cell does not show an accelerated degradation trend, the cumulative effect of performance deviation will lead to problems such as decreased consistency of tiered battery packs, reduced charging and discharging efficiency, and shortened cycle life during long-term use, and may even affect the operational safety of the entire battery pack. Therefore, it is judged as not meeting the tiered utilization standard. If the battery cell is intended for non-core tiered application scenarios with lower requirements for battery cell performance stability and cycle life, such as small household energy storage, emergency lighting, portable power supplies, and toy power supply, these scenarios have a higher tolerance for battery cell performance errors, and the usage scenarios are mostly intermittent and short-term. The cumulative effect of battery cell deviation will not have a significant impact on the performance and safety. At the same time, such deviations can be effectively compensated for by subsequent grouping optimization (such as grouping cells with the same type of cumulative deviation into one group) and precise equalization control of the battery management system (BMS).

[0064] There are also scenarios where only time-invariant characteristics are abnormal and the core time-invariant deviation parameter is the parameter exceeding the threshold. Such defects cannot be compensated for by subsequent grouping optimization, process adjustment, equipment adaptation, etc. Therefore, there is no need to make differentiated judgments based on the application scenario, and they are directly judged as not meeting the tiered utilization standard.

[0065] This setup, with its direct rejection logic for dual abnormal scenarios, minimizes the inflow of high-risk cells into the secondary utilization stage, ensuring the safety and reliability of subsequent secondary products. On the other hand, the detailed differentiation of single abnormal scenarios reduces the need for a one-size-fits-all approach to judgment, strictly controlling the risk of rapid deterioration of time-varying characteristics while flexibly handling minor anomalies of non-core, non-time-varying characteristics, thus improving resource utilization.

[0066] The S700 process groups the cells that can be reused from the cell screening results, installs insulating silicone sheets between the grouped cells, shapes the grouped cells, and then welds and assembles them to obtain a battery pack.

[0067] Understandably, the battery pack matching process follows the principles of prioritizing homogeneity and parameter matching. First, qualified cells are graded according to core parameters such as chemical system (ternary lithium / lithium iron phosphate), nominal capacity, nominal voltage, internal resistance, remaining capacity (SOH), and cycle life. Cells in the same group must belong to the same grade, and deviations in key parameters must be controlled within preset ranges (e.g., capacity deviation ≤5%, internal resistance deviation ≤10%, SOH deviation ≤8%). The number of cells in each group is determined based on the power and capacity requirements of the battery pack for secondary use. For example, a battery pack for home energy storage requires 16 series and 8 parallel cells, with a total voltage of 57.6V and a total capacity of 100Ah. Therefore, matching cells are selected and combined in series and parallel according to this quantity. During the assembly process, insulating silicone sheets are installed between adjacent cells. These sheets have three functions: first, electrical insulation to prevent short circuits caused by contact between the cell tabs and the casing; second, cushioning and shock absorption to absorb vibrations and shocks during battery pack use, protecting the cells from mechanical damage; and third, heat conduction and dissipation to transfer the heat generated by the cells during operation to the battery pack's heat dissipation structure, reducing localized overheating. During installation, the silicone sheets completely cover the contact surfaces between the cells, without misalignment or wrinkles, and have a uniform thickness (typically 1-3mm), conforming to the cell dimensions. After assembly, the cells undergo shaping. A specialized shaping fixture is used to press the assembled cell groups to preset dimensions (e.g., length 300mm, width 200mm, height 150mm) to ensure a compact structure and consistent dimensions for easy subsequent assembly. During shaping, the fixture pressure is controlled (e.g., 0.2-0.3MPa) to avoid excessive pressure that could deform the cells or damage their internal structure. After shaping, welding is performed using laser welding to connect the battery cell's tabs to the connecting pieces. Welding parameters need to be adjusted according to the material of the battery cell's tabs (e.g., nickel or aluminum sheets) (e.g., laser power 100-150W, welding speed 5-10mm / s) to ensure a strong weld joint with low resistance and no incomplete welds. After welding, a tensile test (tensile force ≥ 5N) and a resistance test (joint resistance ≤ 5mΩ) are performed to check for welding defects. Finally, the entire assembly is performed, assembling the welded battery cells into the battery pack casing, installing the BMS (Battery Management System), heat dissipation module, explosion-proof valve, and other accessories, connecting the communication and power supply lines, and sealing the casing. An airtightness test (0.3MPa pressure, pressure held for 30 minutes without leakage), a charge / discharge test (0.3C charging, 0.5C discharging, 3 cycles, capacity retention ≥ 95%), and a safety test (overcharge, over-discharge, and short-circuit protection tests) are conducted. Only after passing these tests is the final battery pack for reuse obtained.

[0068] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0069] Corresponding to the method for the tiered utilization of waste battery packs described in the above embodiments, this application also provides a method for screening battery cells in waste battery packs, including: The battery cell is tested using a battery cell testing device, and the testing data of the battery cell testing events and the operating parameters of the battery cell testing device are obtained within the testing period. Based on the detection data and operating parameters, a detection feature set is determined; wherein, the detection feature set includes detection statistical features corresponding to each detection time window in at least one detection time window, the detection statistical features indicate the time-dependent and non-time-dependent covariates within the corresponding detection time window, and the detection time period includes at least one detection time window; Based on the detection feature set, a target detection feature set corresponding to the detection feature set is generated; wherein, the target detection feature set includes the target detection features corresponding to each detection time window; Based on the cell deviation parameters corresponding to the target detection feature set, cell screening results are generated; the cell screening results are used to indicate whether the cells meet the cascade utilization standards within the detection period.

[0070] It should be noted that the above-mentioned waste battery pack cell screening method is based on the same concept as some embodiments of the method in this application. For details on its specific functions and technical effects, please refer to some method embodiments, which will not be repeated here.

[0071] This application also provides a waste battery pack cascade utilization system for implementing the waste battery pack cascade utilization method described in any of the above embodiments.

[0072] It is understood that the waste battery pack cascade utilization system includes cell testing equipment. The cell testing equipment includes a testing station module, a sensor acquisition module, a main control and computing module, a data storage module, a communication transmission module, and a power supply and voltage regulation module. The testing station module is equipped with adjustable elastic clamps adapted to different specifications of cells (cylindrical, square, and pouch). The clamps have built-in gold-plated probes for stable contact with the cell tabs (contact resistance ≤1mΩ), and the probes have an elastic buffer structure (buffer stroke 0.5-1mm) to reduce damage to the cell tabs or shell due to excessive clamping pressure. At the same time, the bottom of the station integrates an anti-static tray and a temperature sensor to monitor the cell surface temperature in real time (detection range 20℃ to 85℃, accuracy ±0.5℃) to prevent abnormal temperature rise of the cell during the testing process. The sensor acquisition module, as the core unit for data acquisition, integrates multiple types of high-precision sensors: a voltage sensor (range 0-5V, accuracy ±0.01V) for acquiring cell open-circuit voltage and charge / discharge voltage curves; a current sensor (range 0-100A, accuracy ±0.1A) for capturing changes in charge / discharge current; and an internal resistance sensor (range 0-100mΩ, accuracy ±0.1mΩ) for rapidly measuring cell internal resistance using the AC impedance method. In addition, it is equipped with a sampling rate monitoring sensor and a power supply voltage monitoring sensor to capture operating parameters such as the sampling rate of the detection channel and the power supply status of the equipment. The main control and processing module can use a high-performance industrial-grade processor (such as an ARM Cortex-A9 quad-core processor with a main frequency of 1.8GHz), which can simultaneously process data from multiple sensors, perform preliminary processing such as data denoising, calibration, and format conversion, and simultaneously monitor the equipment's own hardware status (CPU utilization, memory usage, and communication link status). The data storage module is equipped with a local solid-state drive (capacity ≥ 1TB) and an SD card expansion interface, supporting real-time local storage of test data and operating parameters. The storage format adopts standardized JSON format and is associated with traceability information such as cell number, test timestamp, and test scheme number. The communication transmission module can integrate CAN bus, Ethernet, and 4G / 5G wireless communication interfaces, allowing flexible selection of data transmission methods according to application scenarios: high-speed wired transmission (transmission rate ≥ 100Mbps) with the backend server is achieved through Ethernet in industrial field scenarios; wireless transmission is achieved through 4G / 5G in mobile testing scenarios, enabling real-time data upload to the cloud or local management system; it also supports CAN bus communication between devices, facilitating clustered management of multi-station testing equipment. The power supply and voltage regulation module adopts a dual-power supply design (main power supply + backup battery). The main power supply supports wide voltage input (AC220V±10%) and has a built-in voltage regulation circuit and filter module to stabilize the input voltage to DC12V / 24V to cope with power grid fluctuations in industrial fields; the backup battery can maintain the operation of the core functions of the equipment when the main power supply is interrupted (battery life ≥ 30 minutes).

[0073] In some embodiments, the waste battery pack reuse system may also include a discharge device for discharging the waste battery pack. For example, the discharge device may be a resistive load (such as an alloy resistor or a cement resistor) that matches the voltage level and rated capacity of the battery pack, and the discharge current may be controlled to be 0.1C-0.2C (C is the rated capacity of the battery pack).

[0074] In some embodiments, the waste battery pack reuse system may further include a battery pack dismantling device for dismantling the waste battery pack to obtain battery cells. For example, the battery pack dismantling device may include a casing dismantling device (such as a wrench, laser cutter, etc.), a wiring dismantling device, and a module dismantling device. The appropriate disassembly tool can be selected based on the casing fixing method (a torque wrench is used for bolt-fixed types, and a laser cutter is used for weld-fixed types). During disassembly, control the direction of the tool's force to reduce scratches to the internal battery module or cells by casing fragments. The second step is to disassemble the internal module. First, disconnect the connection lines and communication lines between the module and the main positive and negative terminals of the battery pack. Record the type, location, and wiring logic of the line interfaces. Then, remove the module from the casing by disassembling the fixing bracket. The module is usually wrapped with an insulating protective layer (such as PVC heat shrink tubing or insulating cloth). Use a utility knife to carefully cut along the seam to avoid damaging the internal cells of the module. The third step is to disassemble the module down to the cells. Remove the cell fixing components inside the module (such as cable ties, brackets, and thermal pads), and disconnect the connecting pieces between the cells (for welded connections, use a laser spot welding separator or a low-temperature soldering iron for heating and separation; for bolted connections, disassemble directly).

[0075] In some real-world examples, the waste battery pack reuse system may also include an assembly device for assembling the selected battery cells to obtain a battery pack. For example, the assembly device may include a welding device, an assembly device, etc. The welding device is used to weld the circuitry, and the assembly device is used to assemble the welded battery cell molds into the battery pack casing.

[0076] The specific implementation process of the above-mentioned device can be found in some of the method embodiments, and will not be repeated here.

[0077] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for the tiered utilization of waste battery packs, characterized in that, The method includes: Discharge the used battery pack to obtain a discharged battery pack; The discharge battery pack was disassembled to obtain multiple battery cells; The battery cell is tested using a battery cell testing device, and the detection data of the battery cell testing events and the operating parameters of the battery cell testing device are obtained within the testing time period. Based on the detection data and the operating parameters, a detection feature set is determined; wherein, the detection feature set includes detection statistical features corresponding to each detection time window in at least one detection time window, the detection statistical features indicate time-dependent and non-time-dependent covariates within the corresponding detection time window, and the detection time period includes at least one detection time window; Based on the detection feature set, a target detection feature set corresponding to the detection feature set is generated; wherein, the target detection feature set includes target detection features corresponding to each of the detection time windows; Based on the cell deviation parameters corresponding to the target detection feature set, a cell screening result is generated; wherein, the cell screening result is used to indicate whether the cell meets the cascade utilization standard within the detection time period; The cells that can be used in stages from the cell screening results are grouped together, insulating silicone sheets are installed between the grouped cells, and then the grouped and assembled cells are shaped, welded and assembled to obtain a battery pack.

2. The method for the tiered utilization of waste battery packs as described in claim 1, characterized in that, The time-dependent covariates include one or more of the following: cell detection frequency, average fluctuation value of cell internal resistance, average CPU utilization rate of detection equipment, equipment memory growth rate, and change value of detection channel sampling rate; the non-time-dependent covariates include one or more of the following: cell detection sensitivity, cell type, and power supply network status.

3. The method for the tiered utilization of waste battery packs as described in claim 2, characterized in that, The detection data includes one or more of the following: detection timestamp, detection event type, initial detection parameters of the battery cell, final detection parameters of the battery cell, and deviation value of the battery cell parameters; the operating parameters include one or more of the following: CPU utilization rate of the detection device, memory utilization rate of the device, sampling rate of the detection channel, detection sensitivity of the battery cell, status of the power supply network, and type of the detected battery cell. The status of the power supply network includes whether the network is in a stable power supply state or in a fluctuating power supply state.

4. The method for the tiered utilization of waste battery packs as described in claim 1, characterized in that, The step of generating cell screening results based on the cell deviation parameters corresponding to the target detection feature set includes: The target detection features corresponding to each detection time window in the target detection feature set are compared with the secondary utilization standard reference features to obtain the cell deviation parameters corresponding to each detection time window; wherein, the cell deviation parameters are used to indicate the degree of difference between the target detection features and the secondary utilization standard reference features within the corresponding detection time window; Based on multiple cell deviation parameters, cell screening results are generated.

5. The method for the tiered utilization of waste battery packs as described in claim 4, characterized in that, The generation of cell screening results based on multiple cell deviation parameters includes: Based on the classification of multiple cell deviation parameters, the time-varying deviation parameter group corresponding to the time-dependent covariate and the non-time-varying deviation parameter group corresponding to the non-time-dependent covariate are obtained within the detection time window. Time-varying feature anomaly determination is performed based on the time-varying deviation parameter set, and non-time-varying feature anomaly determination is performed based on the non-time-varying deviation parameter set. Based on the time-varying anomaly determination results and the non-time-varying anomaly determination results, cell screening results are generated.

6. The method for the tiered utilization of waste battery packs as described in claim 5, characterized in that, The determination of time-varying feature anomalies based on the time-varying deviation parameter set includes: Perform time-dimensional trend analysis on the time-varying deviation parameter group to calculate the dynamic change rate and cumulative deviation value of each parameter; If the dynamic rate of change of any parameter is greater than a preset rate of change threshold, or the cumulative deviation of any parameter is greater than a preset cumulative threshold, it is marked as a time-varying feature anomaly.

7. The method for the tiered utilization of waste battery packs as described in claim 6, characterized in that, The determination of non-time-varying feature anomalies based on the non-time-varying deviation parameter set includes: Static threshold verification is performed on the time-invariant deviation parameter group; wherein, the time-invariant deviation parameter group includes core time-invariant deviation parameters and non-core time-invariant deviation parameters; If at least one of the core time-invariant deviation parameters has a value greater than the preset core single parameter threshold, it is marked as a time-invariant feature anomaly; if only one of the non-core time-invariant deviation parameters has a value greater than the preset non-core single parameter threshold, and the proportion of parameters exceeding the threshold is greater than the preset proportion threshold, it is marked as a time-invariant feature anomaly.

8. The method for the tiered utilization of waste battery packs as described in claim 7, characterized in that, The generation of cell screening results based on time-varying feature anomaly determination results and non-time-varying feature anomaly determination results includes: If a cell is simultaneously marked as having both time-varying and non-time-varying characteristics, or if it is marked as having time-varying characteristics and the dynamic change rate of any parameter exceeds the threshold range, the cell screening result indicates that the cell does not meet the cascade utilization standard; if it is only having non-time-varying characteristics and the parameter exceeding the threshold is the non-core non-time-varying deviation parameter, or if there is no abnormality marker, the cell screening result indicates that the cell meets the cascade utilization standard.

9. A method for screening battery cells in waste battery packs, characterized in that, include: The battery cell is tested using a battery cell testing device, and the detection data of the battery cell testing events and the operating parameters of the battery cell testing device are obtained within the testing time period. Based on the detection data and the operating parameters, a detection feature set is determined; wherein, the detection feature set includes detection statistical features corresponding to each detection time window in at least one detection time window, the detection statistical features indicate time-dependent and non-time-dependent covariates within the corresponding detection time window, and the detection time period includes at least one detection time window; Based on the detection feature set, a target detection feature set corresponding to the detection feature set is generated; wherein, the target detection feature set includes target detection features corresponding to each of the detection time windows; Based on the cell deviation parameters corresponding to the target detection feature set, a cell screening result is generated; wherein, the cell screening result is used to indicate whether the cell meets the cascade utilization standard within the detection time period.

10. A system for the cascade utilization of waste battery packs, characterized in that, This method is used to implement the cascade utilization method of waste battery packs as described in any one of claims 1 to 8.

Images