Search

CN-122017586-A - Cell voltage drop prediction and screening method and system and battery pack

CN122017586ACN 122017586 ACN122017586 ACN 122017586ACN-122017586-A

Abstract

The invention discloses a method and a system for predicting and screening voltage drop of an electric core and a battery pack, and belongs to the technical field of lithium ion battery manufacturing. The method comprises the steps of S1, collecting initial voltage data and capacity-dividing data of a battery core, including an OCV3, an OCV4, a capacity A1 under 100% of SOC, a capacity A2 and a discharge time d1 when the capacity is divided off line, S2, determining reference parameters delta f (A) ', d1' and a pressure drop coefficient beta of a normal battery core based on big data statistical analysis, S3, constructing a pressure drop prediction model, calculating a predicted pressure drop value of the battery core after future time d, and S4, screening out the battery core with stable pressure drop by using u+/-3 sigma criterion based on the distribution of the predicted pressure drop value for packaging. The invention realizes accurate prediction of the future voltage drop trend of the battery cell, screens the battery cell with consistent voltage drop characteristics from the source, and thus improves the overall performance and reliability of the battery module and the battery pack.

Inventors

  • Du Baole
  • CAI GUIFAN
  • ZHONG YUXIANG
  • XIA JIALING

Assignees

  • 滁州国轩新能源动力有限公司

Dates

Publication Date
20260512
Application Date
20260204

Claims (10)

  1. 1. The cell voltage drop prediction and screening method is characterized by comprising the following steps of: s1, data acquisition, namely acquiring initial voltage data and capacity-dividing data of a cell to be tested; s2, determining reference parameters and coefficients, namely obtaining a reference capacity difference delta f (A) ', a reference discharge time d1' and a voltage drop coefficient beta by statistical analysis based on the data of the normal battery cells of the historical batch; s3, voltage drop prediction, namely calculating a predicted voltage drop value f (k) of a single cell to be detected after a future time d through a voltage drop prediction model according to the data acquired in the S1 and the reference parameters and coefficients determined in the S2; and S4, screening the battery cells, namely screening out the battery cells with the predicted voltage drop values in the stable interval for assembling the battery module based on the predicted voltage drop values f (k) of all the battery cells to be tested in the same batch.
  2. 2. The method according to claim 1, wherein the data acquisition in step S1 specifically comprises: s11, collecting voltage OCV3 of the battery cell at a first standing time point t3 and voltage OCV4 of the battery cell at a second standing time point t4, and calculating an initial K value based on a formula K= (OCV 3-OCV 4)/(t 4-t 3); S12, collecting discharge capacity A1 of the battery cell in a 100% SOC state in the capacity division process; s13, acquiring the capacity A2 of the battery cell in a specific SOC state when the battery cell is disconnected, and calculating a capacity difference delta f (A) =A1-A2; S14, collecting time d1 consumed by discharging the battery cell from 100% SOC to 0% SOC in the capacity-dividing process.
  3. 3. The method according to claim 1 or 2, wherein the determining of the reference parameters and coefficients in S2 specifically comprises: s21, selecting historical batch normal electric core data with the same model and the number not less than a preset threshold value; S22, carrying out u+/-3 sigma statistical analysis on the capacity difference delta f (A) and the discharge time d1 data of the historical batch of battery cells, and setting arithmetic average values of the residual normal data as a reference capacity difference delta f (A) 'and a reference discharge time d1' respectively after eliminating abnormal values; S23, placing historical batch of battery cells under a preset environmental condition, measuring voltage change of the battery cells according to fixed time intervals to obtain curve data of the voltage change along with time, carrying out u+/-3 sigma statistical analysis on the curve data, and obtaining a voltage drop coefficient beta based on voltage attenuation curve fitting of the residual normal battery cells after eliminating abnormal values.
  4. 4. The method of claim 3, wherein the pressure drop coefficient β is expressed as a function of time d, and fitting is performed by an exponential function β=a×e (b×d), where a and b are constants obtained by fitting.
  5. 5. The method of claim 1, wherein the voltage drop prediction model in S3 is f (K) =k- β [ (Δf (a) '. D1')/(Δf (a). D1) ]. D, where K is an initial K value of the cell to be measured, Δf (a) is a capacity difference of the cell to be measured, d1 is a discharge time of the cell to be measured, Δf (a) 'is a reference capacity difference, d1' is a reference discharge time, β is a voltage drop coefficient, and d is a future time to be predicted.
  6. 6. The method according to claim 1, wherein the cell screening step in S4 specifically includes calculating an arithmetic mean value u and a standard deviation σ of the predicted voltage drop values f (k) of all the cells to be tested in the same batch, and determining the cells with the predicted voltage drop values f (k) within the interval [ u-3σ, u+3σ ] as voltage drop stable cells for packaging.
  7. 7. A cell screening system based on the method of any one of claims 1-6, comprising: the data acquisition module is used for executing the steps S11 to S14 and acquiring initial performance data of the battery cell; The data processing and storing module is used for storing historical batch of electric core data, executing steps S21 to S23, and calculating and storing reference parameters delta f (A) ', d1' and a pressure drop coefficient beta; The prediction calculation module is used for calling the reference parameters and coefficients in the data processing and storage module, combining the to-be-detected cell data acquired by the data acquisition module, and calculating a predicted voltage drop value f (k) by using the voltage drop prediction model; And the screening decision module is used for executing the cell screening step of the step S4 and outputting a cell list meeting the package requirement.
  8. 8. A computer readable storage medium, characterized in that a computer program is stored on the medium, which computer program, when run, performs the method according to any one of claims 1 to 6.
  9. 9. A computer system comprising a processor, a storage medium having a computer program stored thereon, the processor reading from the storage medium and running the computer program to perform the method of any one of claims 1 to 6.
  10. 10. A battery pack is characterized by comprising battery modules assembled in series and/or parallel by screening out battery cells with stable predicted voltage drop by adopting the method of claims 1-6 or the system of claim 7.

Description

Cell voltage drop prediction and screening method and system and battery pack Technical Field The invention relates to the technical field of lithium ion battery manufacturing and quality control, in particular to a method and a system for predicting a long-term voltage drop trend and carrying out consistency screening based on an algorithm model after battery cell production is completed, and a battery pack. Background In the manufacturing process of the lithium ion battery, after the formation and capacity division processes of the battery cell, self-discharge can be generated due to incomplete stabilization of an internal chemical system, so that open circuit voltage (Open Circuit Voltage, OCV) is reduced along with standing time, and the phenomenon is called voltage drop. The excessive voltage drop of the battery cells or the inconsistent voltage drop rate of the battery cells after the battery cells are packaged can cause the single voltage difference (namely the phenomenon of 'high-low-high') of the battery module in the using or storing process, seriously affect the available capacity, the cycle life and the safety of the module, and even possibly cause the rejection of the whole package. The current common practice in the industry is to measure the voltage of a cell at two specific rest time points (commonly called OCV3 and OCV 4) before the cell is disconnected, and to screen abnormal cells with excessive voltage drop in a short period by calculating a K value (k= (OCV 3-OCV 4)/rest time difference). However, this method has the significant disadvantage that it only reflects the instant voltage drop rate of the cells during the test period, and cannot predict the voltage drop trend of the cells after subsequent long-term (e.g., weeks or months) storage, transportation, or even loading. Some cells with qualified initial K values may have serious voltage drop after long-term storage due to internal micro-short circuit, impurities or unstable interfaces, and have obvious difference with other cells in the same batch. The prior art cannot identify such cells with potential long-term voltage drop risks, thereby burying hidden danger for long-term consistency of the battery module. Therefore, a method for predicting the long-term voltage drop trend of the battery cells is needed, so as to more accurately screen the battery cells with consistent voltage drop characteristics and stable long-term performance before the battery cells are packaged. Disclosure of Invention In order to solve the existing problems, the invention provides a method and a system for predicting and screening voltage drop of a battery cell, and a battery pack, wherein the method comprises the following specific schemes: an electrical core voltage drop prediction and screening method based on an algorithm model comprises the following steps: s1, data acquisition, namely acquiring initial voltage data and capacity-dividing data of a cell to be tested; s2, determining reference parameters and coefficients, namely obtaining a reference capacity difference delta f (A) ', a reference discharge time d1' and a voltage drop coefficient beta by statistical analysis based on the data of the normal battery cells of the historical batch; s3, voltage drop prediction, namely calculating a predicted voltage drop value f (k) of a single cell to be detected after a future time d through a voltage drop prediction model according to the data acquired in the S1 and the reference parameters and coefficients determined in the S2; and S4, screening the battery cells, namely screening out the battery cells with the predicted voltage drop values in the stable interval for assembling the battery module based on the predicted voltage drop values f (k) of all the battery cells to be tested in the same batch. The invention establishes a complete method framework for predicting and screening the voltage drop of the battery cell. The method has the core progress that the screening of the battery cells is changed from 'static detection based on the current state' to 'dynamic evaluation based on model prediction', so that the battery cells which are qualified in initial test and have high risk of long-term voltage drop can be actively identified and removed, the problem of 'charge-up-down' caused by inconsistent voltage of the battery module at the later stage of use is fundamentally prevented, and the overall reliability and the service life of the battery pack are improved. Preferably, the data acquisition in step S1 specifically includes: S11, collecting voltage OCV3 of the battery cell at a first standing time point t3 and voltage OCV4 of the battery cell at a second standing time point t4, and calculating an initial K value based on a formula K= (OCV 3-OCV 4)/(t 4-t 3), wherein the step obtains a short-term self-discharge rate reference value of the battery cell. S12, collecting discharge capacity A1 of the battery cell in a 100% SOC state in the capacity divisi