Cell voltage drop prediction and screening method and system, and battery pack

Abstract

The invention discloses a cell voltage drop prediction and screening method and system, and a battery pack, and belongs to the technical field of lithium ion battery manufacturing. The method comprises the following steps: S1, acquiring initial voltage data and capacity grading data of a battery cell, including OCV3, OCV4, capacity A1 under 100% SOC, capacity A2 during capacity grading offline and discharge time d1; s2, based on big data statistical analysis, determining reference parameters delta f (A) 'and d1 'and a voltage drop coefficient beta of the normal battery cell; s3, constructing a voltage drop prediction model, and calculating a predicted voltage drop value of the battery cell after the future time d; and S4, based on the distribution of the predicted voltage drop values, screening out the battery cells with stable voltage drop by using a u + / -3 sigma criterion for packaging. According to the invention, accurate prediction of the future voltage drop trend of the battery cells is realized, and the battery cells with consistent voltage drop characteristics are screened from the source, so that the overall performance and reliability of the battery module and the battery pack are improved.

Description

Technical Field

[0001] This invention relates to the field of lithium-ion battery manufacturing and quality control technology, and in particular to a method and system for predicting the long-term voltage drop trend of a battery cell based on an algorithm model and performing consistency screening after the cell production is completed, as well as a battery pack. Background Technology

[0002] During the manufacturing process of lithium-ion batteries, after the formation and capacity testing processes, the cells undergo self-discharge due to the incomplete instability of their internal chemical systems. This causes the open circuit voltage (OCV) to decrease over time, a phenomenon known as voltage drop. Cells with excessive voltage drop, or cells with inconsistent voltage drop rates after being assembled into a battery pack, can lead to individual cell voltage differences (i.e., the "high charge, low discharge" phenomenon) during use or storage, severely affecting the module's usable capacity, cycle life, and safety, and may even result in the entire battery pack being scrapped.

[0003] The current industry practice is to measure the voltage of a battery cell at two specific resting time points (usually called OCV3 and OCV4) before it comes off the production line, and then use the calculated K value (K=(OCV3-OCV4) / resting time difference) to screen for abnormal cells with excessive voltage drops in a short period. However, this method has a significant drawback: it only reflects the instantaneous voltage drop rate of the cell during the test period and cannot predict the voltage drop trend of the cell after long-term storage (e.g., weeks or months), transportation, or even after it is installed in a vehicle. Some cells with initially acceptable K values ​​may experience increased voltage drops after long-term storage due to internal micro-short circuits, impurities, or interface instability, resulting in significant differences from other cells in the same batch. Existing technology cannot identify these cells with potential long-term voltage drop risks, thus creating a hidden danger for the long-term consistency of battery modules.

[0004] Therefore, there is an urgent need for a method that can predict the long-term voltage drop trend of battery cells, so as to more accurately select battery cells with consistent voltage drop characteristics and stable long-term performance before battery cell assembly. Summary of the Invention

[0005] To address the existing problems, this invention provides a method and system for predicting and screening battery cell voltage drop, as well as a battery pack. The specific solution is as follows:

[0006] A cell voltage drop prediction and screening method based on an algorithm model includes the following steps:

[0007] S1. Data Acquisition: Acquire the initial voltage data and capacity assessment data of the battery cell under test;

[0008] S2. Determination of reference parameters and coefficients: Based on the data of normal cells in historical batches, statistical analysis is used to obtain the reference capacity difference Δf(A)', reference discharge time d1', and voltage drop coefficient β;

[0009] S3, Voltage Drop Prediction: For a single cell under test, based on the data collected in S1 and the reference parameters and coefficients determined in S2, the predicted voltage drop value f(k) after a future time d is calculated using the voltage drop prediction model.

[0010] S4. Cell Screening: Based on the predicted voltage drop value f(k) of all cells to be tested in the same batch, cells with predicted voltage drop values ​​within the stable range are screened for battery module assembly.

[0011] This invention establishes a complete methodological framework for predicting and screening battery cell voltage drop. Its core advancement lies in transforming cell screening from "static detection based on the current state" to "dynamic evaluation based on model prediction." This allows for the proactive identification and elimination of cells that initially pass testing but have a high risk of long-term voltage drop, fundamentally preventing the "high charge, low discharge" problem caused by voltage inconsistency in the later stages of battery module use, and improving the overall reliability and lifespan of the battery pack.

[0012] Preferably, the data acquisition in step S1 specifically includes:

[0013] S11. Collect the voltage OCV3 of the battery cell at the first resting time point t3 and the voltage OCV4 at the second resting time point t4, and calculate the initial K value based on the formula K=(OCV3-OCV4) / (t4-t3); this step obtains the short-term self-discharge rate reference value of the battery cell.

[0014] S12. Collect the discharge capacity A1 of the battery cell at 100% SOC during the capacity grading process;

[0015] S13. Collect the capacity A2 of the cell under a specific SOC state when it is off the production line for capacity testing, and calculate the capacity difference Δf(A) = A1 - A2; where the capacity difference Δf(A) can reflect the capacity decay of the cell during the capacity testing and storage process, and is related to the internal stability of the cell.

[0016] S14. Collect the time d1 consumed by the battery cell to discharge from 100% SOC to 0% SOC during the capacity grading process; where the discharge time d1 can indirectly reflect the internal resistance and polarization characteristics of the battery cell.

[0017] This approach clarifies the specific, measurable raw data items (K value, A1, A2, d1) necessary for constructing the predictive model. This step firmly establishes the invention's foundation on standard testing procedures in battery cell manufacturing, ensuring the method's operability and repeatability. The selected parameters (K value, capacity difference Δf(A), discharge time d1) comprehensively characterize the cell's state from three key dimensions: short-term self-discharge, internal stability, and internal resistance polarization, providing a comprehensive and multidimensional data foundation for subsequent accurate predictions.

[0018] Preferably, the determination of the reference parameters and coefficients in S2 specifically includes:

[0019] S21. Select normal battery cell data from historical batches of the same model and in quantities not less than the preset threshold.

[0020] S22. Perform u±3σ statistical analysis on the capacity difference Δf(A) and discharge time d1 data of historical batches of battery cells respectively. After removing outliers, set the arithmetic mean of the remaining normal data as the reference capacity difference Δf(A)' and reference discharge time d1' respectively.

[0021] S23. Place historical batches of battery cells under preset environmental conditions, measure their voltage changes at fixed time intervals, obtain voltage change curve data over time, perform u±3σ statistical analysis on the curve data, remove outliers, and obtain the voltage drop coefficient β based on the voltage decay curve of the remaining normal battery cells.

[0022] This scheme discloses a scientific method for determining the core parameters of the model (Δf(A)', d1', β). By employing large-sample statistics (e.g., ≥5000) and outlier removal (u±3σ), the established "normal cell" benchmark is ensured to have high representativeness and purity. In particular, by fitting the voltage drop coefficient β through long-term tracking experiments, the model can truly reflect the objective physical laws governing the evolution of cell voltage drop over time, rather than simple assumptions. This is the fundamental guarantee of prediction accuracy.

[0023] Preferably, the pressure drop coefficient β is expressed as a function related to time d, and is fitted by an exponential function β = a * e^(b*d), where a and b are constants obtained from the fitting.

[0024] This scheme further specifies that the voltage drop coefficient β is in the form of an exponential function. This quantitative description acknowledges and utilizes the universal characteristic of the nonlinear decay of the cell voltage drop rate over time (typically changing rapidly in the early stages and gradually stabilizing in the later stages). Using an exponential model for fitting, compared to a simple linear assumption, can more accurately describe the long-term voltage drop process, thus making the long-term prediction results more reliable and accurate.

[0025] Preferably, the voltage drop prediction model in S3 is: f(k) = K - β * [ (Δf(A)' * d1') / (Δf(A) * d1) ] * d; where K is the initial K value of the cell under test, Δf(A) is the capacity difference of the cell under test, d1 is the discharge time of the cell under test, Δf(A)' is the reference capacity difference, d1' is the reference discharge time, β is the voltage drop coefficient, and d is the future time to be predicted.

[0026] This solution discloses the specific mathematical expression of the voltage drop prediction model. This model not only considers the initial K value and general rules (β*d), but also uses this coefficient to dynamically correct the deviation of individual cells relative to a normal benchmark. If the cell's Δf(A) or d1 is abnormal, this coefficient automatically amplifies its future predicted voltage drop, achieving a personalized and refined assessment of the cell's long-term risk.

[0027] Preferably, the cell screening step in S4 specifically involves: calculating the arithmetic mean u and standard deviation σ of the predicted voltage drop values ​​f(k) of all cells to be tested in the same batch; and determining the cells whose predicted voltage drop values ​​f(k) are within the interval [u-3σ, u+3σ] as stable voltage drop cells for use in packaging.

[0028] This scheme specifies concrete statistical criteria for consistency screening based on prediction results. Using u±3σ as the screening interval, this method can identify cell groups with a high concentration (99.73% confidence interval) in predicted voltage drop values. Battery modules assembled in this way exhibit extremely high voltage consistency among their internal cells after long-term storage, thus maximizing the uniformity and stability of module performance. This is an objective and efficient screening method based on statistical principles.

[0029] The present invention also discloses a cell screening system based on any of the methods described above, comprising:

[0030] The data acquisition module is used to execute steps S11 to S14 to obtain the initial performance data of the battery cell;

[0031] The data processing and storage module is used to store historical batch cell data and execute steps S21 to S23 to calculate and store the reference parameters Δf(A)', d1' and voltage drop coefficient β;

[0032] The prediction calculation module is used to call the reference parameters and coefficients in the data processing and storage module, and combine them with the data of the battery cell under test obtained by the data acquisition module to calculate the predicted voltage drop value f(k) using the voltage drop prediction model.

[0033] The screening decision module is used to execute the cell screening step in step S4 and output a list of cells that meet the packaging requirements.

[0034] This solution visualizes the entire method as a hardware and software integrated cell screening system. Its benefits lie in clarifying the industrial implementation model of this invention, protecting the dedicated equipment or system integration solutions for achieving automated data acquisition, processing, prediction, and decision-making, and helping enterprises build technological barriers and achieve intelligent upgrades to their production lines.

[0035] The present invention also discloses a battery pack, comprising a battery module assembled by using any of the methods or systems described above to screen out cells with predicted stable voltage drop, and assembling them in series and / or in parallel.

[0036] The present invention also discloses a computer-readable storage medium and a computer system, wherein the computer-readable storage medium stores a computer program, and the computer program, when executed, performs the method described in any of the preceding claims. A computer system includes a processor and a storage medium, wherein the storage medium stores a computer program, and the processor reads from the storage medium and runs the computer program to perform the method described in any of the preceding claims.

[0037] The beneficial effects of this invention are as follows:

[0038] 1. Achieving a leap from "detecting the current state" to "predicting future trends": This invention constructs a mathematical model to predict the long-term voltage drop of the battery cell using its initial performance parameters, thus solving the problem that existing technologies cannot provide early warnings of long-term voltage drop risks in battery cells.

[0039] 2. More comprehensive and accurate screening dimensions: The model comprehensively considers multiple parameters such as initial K value, capacity decay, and discharge time, and introduces normal statistical benchmarks for comparison, so that the screened cells are not only qualified in short-term performance, but also more reliable in long-term voltage drop consistency.

[0040] 3. Improved overall quality and reliability of the battery pack: It ensures the consistency of the voltage drop characteristics of the battery cells from the source, effectively prevents the occurrence of "high charge, low discharge" phenomenon at the module end, reduces the module scrap rate, and improves the cycle life and safety performance of the battery pack.

[0041] 4. Easy to achieve automated integration: The steps of this method are clear, and the required data are all routine test data in the cell manufacturing process. It is easy to integrate with existing manufacturing execution systems (MES) or data acquisition and monitoring control systems (SCADA) to achieve automated and intelligent cell screening and grading. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of the method of the present invention;

[0044] Figure 2 This is the normal voltage drop curve of the L600 cell in the example;

[0045] Figure 3 This is a verification curve showing the degree of conformity of the prediction model in the example. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] The core of this invention lies in establishing a performance benchmark and voltage drop law model for normal battery cells by analyzing a large amount of historical data of battery cells, and using this model in combination with the initial data of the battery cell under test to make personalized predictions of its future voltage drop value, and finally performing consistency screening based on the prediction results.

[0048] The following describes the implementation of the present invention in detail using the L600 battery cell as an example.

[0049] like Figure 1 Step S2: Determination of baseline parameters and coefficients (model building stage)

[0050] First, collect complete data on no fewer than 5,000 historical batches of the same model of battery cells that have proven to have stable performance in long-term use. Then process this data:

[0051] Determination of the capacity difference benchmark Δf(A)': Extract the A1 and A2 data of these cells and calculate their respective Δf(A). Perform u±3σ statistical analysis on all Δf(A) values ​​and remove outliers outside the range. Calculate the arithmetic mean of all remaining normal Δf(A) values ​​and define it as the Δf(A)' of this cell model.

[0052] Determination of discharge time reference d1': Similarly, perform u±3σ statistical analysis on the d1 data of the above cells, remove outliers, calculate the arithmetic mean of the remaining normal d1 values, and define it as d1'.

[0053] Determination of the voltage drop coefficient β: These historically normal battery cells were placed in a standard workshop environment (e.g., 25°C), and their open-circuit voltage was measured periodically (e.g., weekly) for several months. The voltage data at each time point was analyzed using u±3σ, and data from cells with abnormal voltage decay were removed, resulting in a "family of normal battery cell voltage decay curves". Mathematical fitting of this family of curves revealed that the voltage drop rate decays exponentially with time, thus yielding an expression for the voltage drop coefficient β. For example, as shown... Figure 2As shown, the fitting result is β=0.8736e^(-0.012d), where d is in days. This β function characterizes the general law of voltage drop variation over time for this type of normal battery cell.

[0054] Steps S1, S3, and S4: Prediction and screening of the batch to be tested (model application stage)

[0055] For the newly produced batch of L600 battery cells, perform the following operations:

[0056] S1. For each cell, collect its OCV3 and OCV4, calculate the initial K value; record its capacity grading data A1, A2, and d1, and calculate Δf(A) = A1 - A2.

[0057] S3. For each cell, substitute the measured K, Δf(A), d1, and the determined model parameters Δf(A)', d1' and β function into the voltage drop prediction model: f(k) = K - β(d) * [ (Δf(A)' * d1') / (Δf(A) * d1) ]* d.

[0058] like Figure 3 The graph shows the validation curves for the degree of conformity of the prediction model.

[0059] For example, to predict the voltage drop of a battery cell after 60 days of storage, d=60. First, calculate β(60)=0.8736e^(-0.012*60)≈0.8736*0.487≈0.425. Then, based on the specific Δf(A) and d1 of the battery cell, calculate the deviation coefficient, and finally obtain the predicted value f(k).

[0060] S4. Calculate the average u and standard deviation σ of the f(k) values ​​of all cells in this batch over the same predicted time d (e.g., 60 days). Set the screening interval to [u-3σ, u+3σ]. Compare the f(k) values ​​of all cells with this interval. Cells whose f(k) falls within this interval are identified as "stable voltage drop cells" and marked as qualified products, which can be used for subsequent module assembly. Cells whose f(k) falls outside this interval are identified as cells with abnormal predicted voltage drop and are isolated or reworked.

[0061] The present invention also discloses a battery pack, comprising a battery module assembled by using any of the methods or systems described above to screen out cells with predicted stable voltage drop, and assembling them in series and / or in parallel.

[0062] The present invention also discloses a computer-readable storage medium and a computer system, wherein the computer-readable storage medium stores a computer program, and the computer program, when executed, performs the method described in any of the preceding claims. A computer system includes a processor and a storage medium, wherein the storage medium stores a computer program, and the processor reads from the storage medium and runs the computer program to perform the method described in any of the preceding claims.

[0063] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

[0064] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor such that the processor can read and write information to / from the storage medium. In an alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and storage medium may reside as discrete components in the user terminal.

[0065] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0066] Although the present invention 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; and these 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 the present invention.

Claims

1. A method for predicting and screening battery cell voltage drop, characterized in that, Includes the following steps: S1. Data Acquisition: Acquire the initial voltage data and capacity assessment data of the battery cell under test; S2. Determination of reference parameters and coefficients: Based on the data of normal cells in historical batches, statistical analysis is used to obtain the reference capacity difference Δf(A)', reference discharge time d1', and voltage drop coefficient β; S3, Voltage Drop Prediction: For a single cell under test, based on the data collected in S1 and the reference parameters and coefficients determined in S2, the predicted voltage drop value f(k) after a future time d is calculated using the voltage drop prediction model. S4. Cell Screening: Based on the predicted voltage drop value f(k) of all cells to be tested in the same batch, cells with predicted voltage drop values ​​within the stable range are selected for battery module assembly.

2. The method according to claim 1, characterized in that, The data collection in step S1 specifically includes: S11. Collect the voltage OCV3 of the battery cell at the first resting time point t3 and the voltage OCV4 at the second resting time point t4, and calculate the initial K value based on the formula K=(OCV3-OCV4) / (t4-t3); S12. Collect the discharge capacity A1 of the battery cell at 100% SOC during the capacity grading process; S13. Collect the capacity A2 of the battery cell under a specific SOC state when it is decommissioned and calculate the capacity difference Δf(A)=A1-A2; S14. Collect the time d1 consumed by the battery cell to discharge from 100% SOC to 0% SOC during the capacity grading process.

3. The method according to claim 1 or 2, characterized in that, The determination of the reference parameters and coefficients in S2 specifically includes: S21. Select normal battery cell data from historical batches of the same model and in quantities not less than the preset threshold. S22. Perform u±3σ statistical analysis on the capacity difference Δf(A) and discharge time d1 data of historical batches of battery cells respectively. After removing outliers, set the arithmetic mean of the remaining normal data as the reference capacity difference Δf(A)' and reference discharge time d1' respectively. S23. Place historical batches of battery cells under preset environmental conditions, measure their voltage changes at fixed time intervals, obtain voltage change curve data over time, perform u±3σ statistical analysis on the curve data, remove outliers, and obtain the voltage drop coefficient β based on the voltage decay curve of the remaining normal battery cells.

4. The method according to claim 3, characterized in that: The pressure drop coefficient β is expressed as a function related to time d, and is fitted by an exponential function β = a * e^(b*d), where a and b are constants obtained from the fitting.

5. The method according to claim 1, characterized in that, The voltage drop prediction model in S3 is: f(k) = K -β * [ (Δf(A)' * d1') / (Δf(A) * d1) ] * d; where K is the initial K value of the cell under test, Δf(A) is the capacity difference of the cell under test, d1 is the discharge time of the cell under test, Δf(A)' is the reference capacity difference, d1' is the reference discharge time, β is the voltage drop coefficient, and d is the future time to be predicted.

6. The method according to claim 1, characterized in that, The cell screening step in S4 is as follows: calculate the arithmetic mean u and standard deviation σ of the predicted voltage drop value f(k) of all cells to be tested in the same batch, and determine the cells whose predicted voltage drop value f(k) is in the interval [u-3σ, u+3σ] as stable voltage drop cells for use in packaging.

7. A cell screening system based on the method of any one of claims 1-6, characterized in that, include: The data acquisition module is used to execute steps S11 to S14 to obtain the initial performance data of the battery cell; The data processing and storage module is used to store historical batch cell data and execute steps S21 to S23 to calculate and store the reference parameters Δf(A)', d1' and voltage drop coefficient β; The prediction calculation module is used to call the reference parameters and coefficients in the data processing and storage module, and combine them with the data of the battery cell under test obtained by the data acquisition module to calculate the predicted voltage drop value f(k) using the voltage drop prediction model. The screening decision module is used to execute the cell screening step in step S4 and output a list of cells that meet the packaging requirements.

8. A computer-readable storage medium, characterized in that: The medium contains a computer program, which, when run, performs the method as described in any one of claims 1 to 6.

9. A computer system, characterized in that: It includes a processor and a storage medium, on which a computer program is stored, and the processor reads from the storage medium and runs the computer program to perform the method as described in any one of claims 1 to 6.

10. A battery pack, characterized in that: This includes battery modules assembled by using the methods described in claims 1-6 or the system described in claim 7 to screen out cells with predicted stable voltage drops and assemble them in series and / or in parallel.

Images